Jid 2016 37 issue 1 by Hogrefe

Volume 37 / Number 1 / 2016

Journal of

Individual Differences Journal of Individual Differences

Editor-in-Chief André Beacuducel Associate Editors Philip J. Corr Sam Gosling Jürgen Hennig Philipp Y. Herzberg Aljoscha Neubauer Thomas Rammsayer Willibald Ruch Stefan Schmukle Astrid Schütz Andrzej Sekowski Jutta Stahl Martin Vocarek

jid_55-0-41-3_58-60_positiv.indd 3

09.12.2015 11:14:30

Hogrefe OpenMind Open Access Publishing? It’s Your Choice! Your Road to Open Access Authors of papers accepted for publication in any Hogrefe journal can now choose to have their paper published as an open access article as part of the Hogrefe OpenMind program. This means that anyone, anywhere in the world will – without charge – be able to read, search, link, send, and use the article for noncommercial purposes, in accordance with the internationally recognized Creative Commons licensing standards.

The Choice is Yours 1. Open Access Publication: The final “version of record” of the article is published online with full open access. It is freely available online to anyone in electronic form. (It will also be published in the print version of the journal.) 2. Traditional Publishing Model: Your article is published in the traditional manner, available worldwide to journal subscribers online and in print and to anyone by “pay per view”. Whichever you choose, your article will be peer-reviewed, professionally produced, and published both in print and in electronic versions of the journal. Every article will be given a DOI and registered with CrossRef.

www.hogrefe.com

How Does Hogrefe‘s Open Access Program Work? After submission to the journal, your article will undergo exactly the same steps, no matter which publishing option you choose: peer-review, copy-editing, typesetting, data preparation, online reference linking, printing, hosting, and archiving. In the traditional publishing model, the publication process (including all the services that ensure the scientific and formal quality of your paper) is financed via subscriptions to the journal. Open access publication, by contrast, is financed by means of a one-time article fee (€ 2,500 or US $3,000) payable by you the author, or by your research institute or funding body. Once the article has been accepted for publication, it’s your choice – open access publication or the traditional model. We have an open mind!

Journal of

Individual Differences Volume 37, No. 1, 2016

Editor-in-Chief

Prof. Dr. Andre´ Beauducel, Institut fu¨r Psychologie, Rheinische Friedrich-Wilhelms-Universita¨t, Kaiser-Karl-Ring 9, D-53111 Bonn, Germany, Tel. +49 228 734-151, E-mail beauducel@uni-bonn.de

Associate Editors

Philip J. Corr, UK Sam Gosling, USA Ju¨rgen Hennig, Germany Philipp Y. Herzberg, Germany Aljoscha Neubauer, Austria Thomas Rammsayer, Switzerland

Willibald Ruch, Switzerland Stefan Schmukle, Germany Astrid Schu¨tz, Germany Andrzej Sekowski, Poland Jutta Stahl, Germany Martin Vocarek, Austria

Editorial Board

Philipp L. Ackerman, USA Jens Asendorpf, Germany Jose´ Bermudez, Spain Peter Borkenau, Germany John Brebner, Australia Burkhard Brocke, Germany Ian Deary, UK Richard Depue, USA Richard Ebstein, Israel Aiden P. Gregg, UK Hartmut Ha¨cker, Germany Willem B. Hofstee, The Netherlands Klaus Kubinger, Austria

Bernd Marcus, Germany Robert R. McCrae, USA Carolyn C. Morf, Switzerland Pierre Mormede, France Jaak Panksepp, USA Kurt Pawlik, Germany Robert Plomin, UK Rainer Riemann, Germany Kurt Stapf, Germany Bob Stelmack, Canada Gerhard Stemmler, Germany Jan Strelau, Poland

Publisher

Hogrefe Publishing, Merkelstr. 3, D-37085 Go¨ttingen, Germany, Tel. +49 551 99950-0, Fax +49 551 99950-111, E-mail publishing@hogrefe.com, Web http://www.hogrefe.com North America: Hogrefe Publishing, 38 Chauncy Street, Suite 1002, Boston, MA 02111, USA, Phone (866) 823-4726, Fax (617) 354-6875, E-mail publishing@hogrefe.com

Production

Regina Pinks-Freybott, Hogrefe Publishing, Merkelstr. 3, D-37085 Go¨ttingen, Germany, Tel. +49 551 99950-0, Fax +49 551 99950-111, E-mail journalsproduction@hogrefe.com

Subscriptions

Hogrefe Publishing, Herbert-Quandt-Str. 4, D-37081 Go¨ttingen, Germany, Tel. +49 551 50688-900, Fax +49 551 50688-998

Advertising/Inserts

Melanie Beck, Hogrefe Publishing, Merkelstr. 3, D-37085 Go¨ttingen, Germany, Tel. +49 551 99950-423, Fax +49 551 99950-111, E-mail marketing@hogrefe.com

ISSN

ISSN-L 1614-0001, ISSN-Print 1614-0001, ISSN-Online 2151-2299

Ó 2016 Hogrefe Publishing. This journal as well as the individual contributions and illustrations contained within it are protected under international copyright law. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, microﬁlming, recording or otherwise, without prior written permission from the publisher. All rights, including translation rights, reserved.

Publication

Published in 4 issues per annual volume.

Subscription Prices

Calendar year subscriptions only. Rates for 2016: Institutions US $324.00/1 238.00/GBP 190.00; Individuals US $159.00/1 114.00/GBP 91.00. Postage and handling: US $16.00/1 12.00/GBP 10.00. Single copies: US $81.00/1 59.50/GBP 47.50 + postage and handling.

Payment

Payment may be made by check, international money order, or credit card, to Hogrefe Publishing, Merkelstr. 3, D-37085 Go¨ttingen, Germany, or, for North American customers, to Hogrefe Publishing, 38 Chauncy Street, Suite 1002, Boston, MA 02111, USA.

Abstracting Services

Abstracted/indexed in Current Contents/Social and Behavioral Sciences (CC/S&BS), Social Sciences Citation Index (SSCI), Scopus, PsycINFO (APA), and PSYNDEX (ZPID).

Electronic Full-Text

The full-text of the Journal of Individual Differences is available online at http://econtent.hogrefe.com and also in PsycARTICLES. Impact Factor (2014): 1.222

Journal of Individual Differences 2016; Vol. 37(1)

Ó 2016 Hogrefe Publishing

Contents Original Articles

Ó 2016 Hogrefe Publishing

Gymnasts and Orienteers Display Better Mental Rotation Performance Than Nonathletes Mirko Schmidt, Fabienne Egger, Mario Kieliger, Benjamin Rubeli, and Julia Schu¨ler

Conscientiousness Is the Most Powerful Noncognitive Predictor of School Achievement in Adolescents Barbara Dumfart and Aljoscha C. Neubauer

Perception of Emotional Expressions in Adults: The Role of Temperament and Mood Chit Yuen Yi, Matthew W. E. Murry, and Amy L. Gentzler

The Relationships Between the Dark Triad, the Moral Judgment Level, and the Students’ Disciplinary Choice: Self-Selection, Indoctrination, or Both? Annika Krick, Stephanie Tresp, Mirijam Vatter, Antonia Ludwig, Michael Wihlenda, and Martin Rettenberger

Social Support, Emotional Intelligence, and Posttraumatic Stress Disorder Symptoms: A Mediation Analysis Nicole L. Hofman, Austin M. Hahn, Christine K. Tirabassi, and Raluca M. Gaher

Rank-Order Consistency and Proﬁle Stability of Self- and InformantReports of Personal Values in Comparison to Personality Traits Henrik Dobewall and Toivo Aavik

Personality of Clown Doctors: An Exploratory Study Alberto Dionigi

Validation and Revision of a German Version of the Balanced Measure of Psychological Needs Scale Andreas B. Neubauer and Andreas Voss

Journal of Individual Differences 2016; Vol. 37(1)

Original Article

Gymnasts and Orienteers Display Better Mental Rotation Performance Than Nonathletes Mirko Schmidt, Fabienne Egger, Mario Kieliger, Benjamin Rubeli, and Julia Schüler Institute of Sport Science, University of Bern, Switzerland Abstract. The aim of this study was to examine whether athletes differ from nonathletes regarding their mental rotation performance. Furthermore, it investigated whether athletes doing sports requiring distinguishable levels of mental rotation (orienteering, gymnastics, running), as well as varying with respect to having an egocentric (gymnastics) or an allocentric perspective (orienteering), differ from each other. Therefore, the Mental Rotations Test (MRT) was carried out with 20 orienteers, 20 gymnasts, 20 runners, and 20 nonathletes. The results indicate large differences in mental rotation performance, with those actively doing sports outperforming the nonathletes. Analyses for the specific groups showed that orienteers and gymnasts differed from the nonathletes, whereas endurance runners did not. Contrary to expectations, the mental rotation performance of gymnasts did not differ from that of orienteers. This study also revealed gender differences in favor of men. Implications regarding a differentiated view of the connection between specific sports and mental rotation performance are discussed. Keywords: cognition, physical exercise, spatial ability, gender differences

Regular physical exercise is associated with many beneficial effects on physical and mental health, such as better general and health-related quality of life, better functional capacity, and better mood states (Penedo & Dahn, 2005). Aside from both these preventive and therapeutic effects, cognitive functions also appear to benefit from acute and chronic sports (Chang, Labban, Gapin, & Etnier, 2012; Hillman, Erickson, & Kramer, 2008). It remains unclear, however, which specific physical activity affects which specific cognitive skill. Physical activities with higher cognitive demands may be assumed to have a stronger influence on cognitive skills than those, which make lower or fewer cognitive demands (Pesce, 2012). For example, Pesce, Crova, Cereatti, Casella, and Bellucci (2009) tested the influence of either an aerobic circuit training lesson or a team games lesson on memory performance in preadolescents. The two conditions were characterized by similar exercise intensities but differed in their cognitive complexity (and therefore in their cognitive demands). Memory performance was significantly better after team games than after aerobic circuit training. One significant cognitive skill, along with others, is mental rotation, that is, the ability to mentally manipulate two- or three-dimensional objects, whereby these objects may be rotated in any direction or translated in space (Shepard & Metzler, 1971). When the mental rotation paradigm is applied in research, subjects usually have to judge whether a couple of (mostly three-dimensional) objects Ó 2016 Hogrefe Publishing

presented in various orientations are identical to a specific target object. Apart from the classical cube figures used by Shepard and Metzler (1971), alphanumeric characters (e.g., Cooper & Shepard, 1973), images of human faces (e.g., Valentine & Bruce, 1988), body parts (e.g., Petit, Pegna, Mayer, & Hauert, 2003), or even entire bodies (e.g., Jola & Mast, 2005) are used as objects to be rotated. In this context, two types of mental transformations can be differentiated: object-based transformations are mental rotations from an allocentric point of view, in which an object is rotated and the observer’s point of view stays fixed. Perspective transformations are mental rotations from an egocentric point of view, in which an object stays fixed and the observer’s point of view rotates in relation to the object or the environment (Zacks, Mires, Tversky, & Hazeltine, 2002). Mental rotation, however, is an important and relevant construct for spatial ability and problemsolving strategies (Geary, Saults, Liu, & Hoard, 2000) as well as for specific mathematical and scientific competencies (Hegarty & Kozhevnikov, 1999; Peters, Chisholm, & Laeng, 1995). Studies consistently found gender differences in favor of male subjects (Voyer, Voyer, & Bryden, 1995), which are attributed not only to biological, but also to sociocultural differences (Maeda & Yoon, 2012). The ability to mentally rotate objects and the ability to rotate objects using motor commands or to navigate through one’s environment appear to be interrelated. For example, studies in developmental psychology Journal of Individual Differences 2016; Vol. 37(1):1–7 DOI: 10.1027/1614-0001/a000180

M. Schmidt et al.: Physical Activity and Mental Rotation

indicating strong associations between motor development and mental rotation performance in children (Frick & Möhring, 2013; Rakison & Woodward, 2008) corroborate studies that show children with impaired motor function displaying impaired mental rotation performance (Jansen, Schmelter, Kasten, & Heil, 2011). Furthermore, several neuroimaging studies have been able to demonstrate that the same areas of the brain are active when carrying out mental rotations as during physical activities involving the rotation of objects (e.g., Draganski et al., 2004; Jordan, Heinze, Lutz, Kanowski, & Jäncke, 2001). However, it appears that the areas that are active during egocentric and allocentric transformations are not the same ones, whereby differences are found particularly at the level of the posterior parietal cortex (Pelgrims, Andres, & Olivier, 2009). These differences suggest that allocentric and egocentric sports may affect mental rotation skills differently. Taken together, considering how important mental rotation skills are to succeed in numerous activities and professions (Hegarty & Waller, 2005), and consulting empirical evidence both from developmental and neuroimaging studies, it is astonishing that the connection between different sports activities and mental rotation performance has received so little attention until now. However, the few existing studies that do look at the relation between general physical activity and mental rotation performance indicate promising positive connections (Jansen, Lange, & Heil, 2011; Jansen & Lehmann, 2013; Jansen, Lehmann, & Van Doren, 2012; Jansen & Pietsch, 2010; Jansen, Titze, & Heil, 2009; Moreau, Clerc, Mansy-Dannay, & Guerrien, 2012; Moreau, MansyDannay, Clerc, & Guerrien, 2011; Ozel, Larue, & Molinaro, 2002). The sports investigated have included juggling, football, gymnastics, and a range of different martial arts, such as fencing, judo, and wrestling, as well as physical activities such as running. Acute and chronic physical activity have been found to be positively related to mental rotation performance. In all studies, those who actively performed sports displayed a significantly higher mental rotation performance than subjects who were inactive. The question which types of sport are particularly effective at training mental rotation skills has only been marginally investigated so far, since most studies have only compared a single sport with a group of nonathletes. To the best of our knowledge, only four studies (Jansen, Lange, et al., 2011; Jansen & Lehmann, 2013; Moreau et al., 2011, 2012) have compared different types of sport, and these uniformly report that cognitively demanding sports involving mental rotation exert a greater influence. In their intervention studies, Jansen, Lange, et al. (2011) compared the coordinative sport of juggling with a strength-training program; Moreau et al. (2012) compared wrestlers with endurance athletes. Only the study by Moreau et al. (2011) and the one by Jansen and Lehmann (2013) have compared two different types of sport and investigated in detail the specific effects of various sports on mental rotation performance. Moreau et al. (2011) showed that elite combat athletes (fencing, judo, and wrestling) displayed better mental rotation performance than elite runners. Since the daily practice time did not differ between the two groups, the better mental rotation Journal of Individual Differences 2016; Vol. 37(1):1–7

performance was explained by the repeated use of mental and physical rotations in the practice of combat sports as compared with running, which entails only small degrees of mental and physical rotations. Building upon this, a recent study by Jansen and Lehmann (2013) compared three groups, namely soccer players, gymnasts, and nonathletes, in an object-based mental rotation task consisting of human postures and cube figures. They expected athletes whose sport involves mental and physical rotations to display a better object-based mental rotation performance than nonathletes, and the object-based mental rotation performance to differ between the two types of sport. The rationale for these hypotheses was that soccer players are thought to perceive objects mostly from an allocentric point of view, whereas gymnasts mostly train their own body transformations, perceiving the environment from an egocentric point of view. Results showed that the gymnasts displayed a better performance than nonathletes in the objectbased mental rotation task. Nevertheless, contrary to the expectations, gymnasts and soccer players did not differ in their mental rotation performance. This well-designed study is limited, however, by the fact that, despite including two different types of sport with varying engagements in mental manipulation of objects (allocentric vs. egocentric transformations), a sports group that trains physically but lacks the systematic involvement of mental rotation in its respective sport activity, for example runners, is missing. With respect to the latter study, one might speculate that there may be some sports that differ more from gymnastics concerning the types of mental transformations and the general mental rotation engagement than soccer does. For example, orienteering is a sport in which the allocentric perspective plays a central role when navigating through one’s surroundings. For successful orientation, mostly in the woods, orienteers have to adopt an allocentric perspective to locate themselves as an object on the coordinate system of the map. Therefore, while exercising, they train their mental rotation skills, which in turn should lead to enhanced performance especially in object-based mental rotation tasks – since to solve them one has to adopt an allocentric perspective. To the best of our knowledge, no study has ever compared orienteers with other athletes with or without mental rotation engagement, or even tested whether they differ from nonathletes. However, any differences that might be found would help to determine which specific effects different sports may have on mental rotation skills. The present study therefore examines whether different sports that make various mental rotation demands differ in terms of the resulting mental rotation performance. Thus, the mental rotation performance of athletes in a sport with a high degree of egocentric transformations (gymnastics), a sport with a high degree of allocentric transformations (orienteering), and an endurance sport without mental rotation demands (running) are compared with each other, and in addition with the performance of a group of nonathletes. The conjecture is that (1) people actively involved in sports will have a better mental rotation performance than nonathletes. That (2) each individual sport will lead to a better mental rotation performance than in nonathletes, Ó 2016 Hogrefe Publishing

M. Schmidt et al.: Physical Activity and Mental Rotation

irrespective of the mental-spatial demands made by the sport. (3) Furthermore, athletes involved in sports which require mental rotations in order to be carried out successfully (gymnastics and orienteering) will have a better mental rotation performance than athletes doing a sport that does not require mental rotation (running). (4) Finally, it is surmised that orienteers (allocentric transformations) will do better in an object-based mental rotation test focusing on allocentric rotations than gymnasts (egocentric transformations).

Method Participants A total of 80 undergraduate students (50% men in each group, Mage = 25.73, SD = 4.63) took part in the study, having been recruited with the alleged explanation of wanting to compare mental rotation performance between people studying different subjects at university. The group of nonathletes (n = 20) consisted of people who had spent less than 30 min a day doing at least moderate-intensity physical activity on 5 or more days of the week in the 7 days before the survey (Sproston & Primatesta, 2003). The group of athletes (n = 60) consisted of people involved in the three sports orienteering (n = 20), gymnastics (n = 20), and running (n = 20), who had to have been participating in the respective sport for at least 2 years and who were doing it at least twice a week. The four groups (orienteers, gymnasts, runners, nonathletes) did not differ in terms of their ages, F(3, 72) = 2.22, p = .088, g2p = .10; see Table 1, but did with respect to the number of training sessions per week, F(3, 72) = 24.60, p < .0005, g2p = .49, the minutes spent per training session, F(3, 72) = 44.62, p < .0005, g2p = .64, and the years spent practicing their respective sport, F(3, 72) = 26.02, p < .0005, g2p = .51. As expected, Tukey-HSD post hoc tests revealed that all three sport groups train more times per week than the nonathletes (p < .0005), but that they do not differ between each other (p > .999). Tukey-HSD post hoc tests for the minutes spent per training session showed that, apart from the runners and the orienteers (p = .179), all groups differ in the duration of their training session (ps < .004). Tukey-HSD post hoc tests for the years spent doing the sport showed that all three sports groups spent more years practicing their sport than nonathletes (p < .0005), but also that gymnasts started their sports career earlier than runners (p = .001). The data contained no missing values, nor were

any univariate outliers found using Grubbs’ test (Z = 2.58, p > .05).

Measures Mental Rotation Set ‘‘A’’ of the Mental Rotations Test (MRT-A) by Peters, Laeng, et al. (1995) was used to assess mental rotation performance. This test was originally developed by Vandenberg and Kuse (1978), using objects made up of cubes, as designed by Shepard and Metzler (1971). Overall, 24 tasks need to be solved in the MRT-A, whereby in each case the object has to be mentally rotated and compared with four other objects made up of cubes. Two of the four stimuli are identical to the target figure. A task is only considered to have been solved if both the correct answers are identified. This means that the maximum possible score is 24 points. The reliability and validity of the test have been demonstrated (Geiser, Lehmann, & Eid, 2006).

Physical Activity Physical activity was examined using the ‘‘Sportaktivität’’ (Sports Activity) subscale of the ‘‘Bewegungs- und Sportaktivität’’ questionnaire (BSA; Fuchs, 2012) in which participants are asked about their regular sports activities over the past 4 weeks, the frequency with which they carried out these activities, and the duration of each session.

Procedure First, the MRT-A was carried out to determine the mental rotation performance. Next, participants completed the questionnaire on their current sports behavior and gave details of their age and sex. The entire procedure lasted 15 min and was carried out as an individual test in a quiet room. All subjects gave their informed consent before taking part and were able to discontinue the study at any time.

Statistical Analyses In order to test whether athletes (all three sports groups together) and nonathletes differ from each other with respect to their MRT-A performance, a univariate analysis

Table 1. Mean scores (and standard deviations in parentheses) for all study variables by group Variable

Orienteers (n = 20)

MRT score

13.80 (4.35)

Age Training sessions per week Minutes per training session Years spent doing sport

26.20 2.75 59.50 10.80

Ó 2016 Hogrefe Publishing

(5.36) (1.68) (8.87) (5.80)

Gymnasts (n = 20) 13.65 (5.16) 27.15 2.65 130.00 14.20

(6.10) (0.93) (44.84) (4.58)

Runners (n = 20)

Nonathletes (n = 20)

12.10 (3.57)

9.35 (3.83)

25.95 2.85 80.75 8.35

(3.12) (0.88) (26.12) (5.06)

23.60 0.35 22.00 1.38

(2.50) (0.49) (29.49) (3.22)

Journal of Individual Differences 2016; Vol. 37(1):1–7

M. Schmidt et al.: Physical Activity and Mental Rotation

Figure 1. Mean scores of MRT-A for the four groups studied, with standard error of the mean.

of variance (ANOVA) was conducted with the dependent variable MRT-A test score and the independent variable activity group (athletes, nonathletes) and gender (male, female). To test all three remaining study hypotheses, a univariate analysis of variance (ANOVA) was performed with the dependent variable MRT-A test score and the independent variables group (orienteers, gymnasts, runners, and nonathletes) and gender (male, female). Gender was considered as a factor because men are known to perform better in object-based mental rotation tests than women (e.g., Voyer et al., 1995). Statistical Package for Social Sciences (SPSS 21.0) was used for all analyses. Post hoc comparisons using Tukey’s HSD test were used to compare specific groups with one another. A significance level of .05 (two-tailed) was set for all tests. The effect size was calculated and interpreted using Cohen’s (1988) definition of small, medium, and large effect sizes (g2p = .01, .06, .14).

Results Preliminary Analyses The overall ANOVA with the dependent variable MRT-A test score and the independent variables group and gender revealed a significant overall effect, indicating that groups and/or genders differ with respect to their mental rotation performance, F(7, 72) = 4.07, p = .001, g2p = .284. Whereas there was a significant main effect for the factors group, F(3, 72) = 5.22, p = .003, g2p = .179, and gender, F(1, 72) = 11.39, p = .001, g2p = .137, there was no significant interaction between group and gender, F(3, 72) = .49, p = .693, g2p = .020. The mean sum score of the male subjects was M = 13.75 (SD = 4.46), whereas the mean sum score of the female subjects was M = 10.68 (SD = 4.16), meaning that this effect can be described as being large. Thus, the uniform distribution of the sexes within the different groups proved to be the central control variable in the present study design.

Journal of Individual Differences 2016; Vol. 37(1):1–7

Primary Analyses Figure 1 shows the mean scores of the four test groups in the MRT-A. The first hypothesis postulated a difference between athletes and nonathletes. Athletically active subjects (M = 13.18, SD = 4.40) displayed significantly higher MRT-A scores, F(1, 76) = 13.56, p < .0005, g2p = .151, than their nonathletic counterparts (M = 9.35, SD = 3.83). The reported effect size can be described as large. The second hypothesis postulated that each different sport group would display a better mental rotation performance than the group of nonathletes. As already indicated in the preliminary analyses, there is a significant difference between the mean MRT-A scores for the four groups investigated, F(3, 72) = 5.22, p = .003, g2p = .179. The effect size suggests a large effect. The post hoc comparison using Tukey’s HSD test shows that orienteers (p = .008) and gymnasts (p = .011) have significantly better mental rotation skills than nonathletes. Runners do not differ significantly from nonathletes in terms of their mental rotation performance (p = .184). Hypothesis 3 suggested that those sports involving higher levels of mental rotation (gymnastics and orienteering) would be associated with a better mental rotation performance than sports without mental rotation (running). Although on a descriptive level the gymnasts (p = .661) and orienteers (p = .592) displayed higher means than the runners (see Table 1), this difference is not statistically significant. Contrary to the fourth hypothesis, gymnasts (egocentric transformations) do not differ from orienteers (allocentric transformations) with respect to their mental rotation performance (p = .999). Since it is not only the level of the mental rotation demands made by a specific sport that could influence mental rotation performance, but also the time spent doing a specific sport and the intensity – in addition to the four hypotheses tested – another ANOVA was conducted including the variables training sessions per week, minutes per training session, and years spent doing the sport as additional factors. The results show that none of the

Ó 2016 Hogrefe Publishing

M. Schmidt et al.: Physical Activity and Mental Rotation

variables included exerted a significant main effect on the dependent variable ( ps > .134), nor did they interact significantly with one of the other factors ( ps > .09). Furthermore, the main effects for the factors group, F(3, 57) = 2.45, p = .033, g2p = .167, and gender, F(1, 57) = 8.18, p = .006, g2p = .113, did not differ notably when these variables were included in the model.

Discussion The main findings of the present study were that with respect to their mental rotation performance (a) people actively doing sports outperformed nonathletes; (b) orienteers and gymnasts outperformed nonathletes; (c) contrary to expectations, gymnasts did not differ from orienteers; (d) men outperformed women. Since findings (a) and (d) reflect the existing literature on differences in mental rotation performance between athletes and nonathletes (Jansen et al., 2012; Ozel et al., 2002) and between men and women (Maeda & Yoon, 2012; Voyer et al., 1995), respectively, these results will not be discussed any further. Sports do not only differ in terms of their physical demands (such as aerobic and anaerobic activity, strength and coordination tasks) though, but also in terms of the demands they make on mental rotation skills (thereby promoting these). It can therefore be assumed that not all sports will affect mental rotation performance in the same way or to the same extent. This is supported by experimental studies demonstrating greater effects on subject’s mental rotation performance for sports involving coordination and mental rotation than those with just few cognitive demands (Jansen, Lange, et al., 2011; Moreau et al., 2012). However, to date only two cross-sectional studies have compared two sports groups with different mental rotation demands with each other and revealed that elite combat athletes display better mental rotation performance than elite runners (Moreau et al., 2011) or that soccer players do not differ from gymnasts (Jansen & Lehmann, 2013). The present study is therefore the first to compare the three different sports gymnastics (egocentric transformations), orienteering (allocentric transformations), and running (without mental rotations) with a group of nonathletes. Interestingly, besides gymnasts also orienteers display a higher mental rotation performance than nonathletes. The finding that athletes with lots of experience in allocentric transformations (orienteers) and those with lots of experience in egocentric transformations (gymnasts) did not differ in their mental rotation performance is in line with the results of Jansen and Lehmann (2013). It may indicate that the connection between sports activity and mental rotation performance is primarily explained by the general level of mental rotation demands made by a specific type of sport rather than the different (ego- or allocentric) perspective inherent in the sport. Nevertheless, it also has to be said that in our sample with equal numbers of training sessions per week, the mean training duration is much longer in gymnasts than in orienteers. On the one hand, this Ó 2016 Hogrefe Publishing

reflects the natural setting of the respective sports, on the other hand this could have affected the results of the present study. Further studies aimed at discovering the differential effects of ego- or allocentric sports could ensure that this possible confounding factor does not differ between the groups being compared. Besides the small sample size, one limitation of the present study is certainly its cross-sectional design, which prevents any causal conclusions from being drawn. Whether people with good mental rotation skills tend to choose sports that make high demands on their mental rotation performance, or whether sports with high levels of object rotations are in fact able to influence mental rotation performance, cannot be answered by this study. However, training and experimental studies do certainly suggest that this is the case (Jansen, Lange, et al., 2011; Jansen et al., 2009; Jansen & Pietsch, 2010; Moreau et al., 2011, 2012). More intervention studies are required that compare different sports with one another and with nonathletes. To shed more detailed light on the causal connections between different sports activities and their effects on mental rotation performance, future studies could include the underlying physiological mechanisms involved in different sports activity to a greater extent. Neurophysiological measures, such as functional imaging technologies (fMRT), could directly measure the different neuronal activities and how these change over time (e.g., Voelcker-Rehage, Godde, & Staudinger, 2011). Another important limitation is that no additional cognitive ability, for example participant’s overall intelligence or cognitive processing speed, was measured. Considering the study conducted by Jansen and Lehmann (2013), which showed that male and female soccer players differ in their processing speed, one might argue that the differences found in the present study may have been due to differences in precisely this variable. Thus, processing speed is a cognitive ability that deserves more attention in future studies investigating mental rotation differences. Nevertheless, the present study revealed not only gymnasts, but also orienteers to have better mental rotation performance than nonathletes. This result supports the claim that specific sports may have specific effects on specific cognitive skills.

References Chang, Y. K., Labban, J. D., Gapin, J. I., & Etnier, J. L. (2012). The effects of acute exercise on cognitive performance: A meta-analysis. Brain Research, 1453, 87–101. doi: 10.1016/ j.brainres.2012.02.068 Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. Cooper, L. A., & Shepard, R. N. (1973). The time required to prepare for a rotated stimulus. Memory & Cognition, 1, 246–250. doi: 10.3758/BF03198104 Draganski, B., Gaser, C., Busch, V., Schuierer, G., Bogdahn, U., & May, A. (2004). Neuroplasticity: Changes in grey matter induced by training – newly honed juggling skills show up as a transient feature on a brain-imaging scan. Nature, 427, 311–312. doi: 10.1038/427311a Journal of Individual Differences 2016; Vol. 37(1):1–7

M. Schmidt et al.: Physical Activity and Mental Rotation

Frick, A., & Möhring, W. (2013). Mental object rotation and motor development in 8- and 10-month-old infants. Journal of Experimental Child Psychology, 115, 708–720. doi: 10.1016/j.jecp.2013.04.001 Fuchs, R. (2012). Messung der Bewegungs- und Sportaktivität: Der BSA-Fragebogen [Measurement of physical activity and exercise: The BSA-questionnaire]. Retrieved from http:// www.sport.uni-freiburg.de/institut/Arbeitsbereiche/psychologie Geary, D. C., Saults, S. J., Liu, F., & Hoard, M. K. (2000). Sex differences in spatial cognition, computational fluency, and arithmetical reasoning. Journal of Experimental Child Psychology, 77, 337–353. doi: 10.1006/jecp.2000.2594 Geiser, C., Lehmann, W., & Eid, M. (2006). Separating ‘‘rotators’’ from ‘‘nonrotators’’ in the Mental Rotations Test: A multigroup latent class analysis. Multivariate Behavioral Research, 41, 261–293. doi: 10.1207/s15327906mbr4103_2 Hegarty, M., & Kozhevnikov, M. (1999). Types of visual-spatial representations and mathematical problem solving. Journal of Educational Psychology, 91, 684–689. doi: 10.1037/00220663.91.4.684 Hegarty, M., & Waller, D. A. (2005). Individual differences in spatial abilities. In P. Shah & A. Miyake (Eds.), The Cambridge Handbook of Visuospatial Thinking (pp. 121–169). New York, NY: Cambridge University Press. Hillman, C. H., Erickson, K. I., & Kramer, A. F. (2008). Be smart, exercise your heart: Exercise effects on brain and cognition. Nature Reviews Neuroscience, 9, 58–65. doi: 10.1038/nrn2298 Jansen, P., Lange, L. F., & Heil, M. (2011). The influence of juggling on mental rotation performance in children. Biomedical Human Kinetics, 3, 18–22. doi: 10.2478/ v10101-011-0005-6 Jansen, P., & Lehmann, J. (2013). Mental rotation performance in soccer players and gymnasts in an object-based mental rotation task. Advances in Cognitive Psychology, 9, 92–98. doi: 10.2478/v10053-008-0135-8 Jansen, P., Lehmann, J., & Van Doren, J. (2012). Mental rotation performance in male soccer players. PLoS One, 7, e48620. doi: 10.1371/journal.pone.0048620 Jansen, P., & Pietsch, S. (2010). Physical activity improves mental rotation performance. Creative Education, 1, 58–61. doi: 10.4236/ce.2010.11009 Jansen, P., Schmelter, A., Kasten, L., & Heil, M. (2011). Impaired mental rotation performance in overweight children. Appetite, 56, 766–769. doi: 10.1016/j.appet.2011. 02.021 Jansen, P., Titze, C., & Heil, M. (2009). The influence of juggling on mental rotation performance. International Journal of Sport Psychology, 40, 351–359. Jola, C., & Mast, F. W. (2005). Mental object rotation and egocentric body rotation: Two dissociable processes? Spatial Cognition and Computation, 5, 217–237. Jordan, K., Heinze, H. J., Lutz, K., Kanowski, M., & Jäncke, L. (2001). Cortical activations during the mental rotation of different visual objects. NeuroImage, 13, 143–152. doi: 10.1006/nimg.2000.0677 Maeda, Y., & Yoon, S. Y. (2012). A meta-analysis of gender differences in mental rotation ability measured by the Purdue Spatial Visualization Test: Visualization of rotations. Educational Psychology Review, 25, 69–94. doi: 10.1007/ s10648-012-9215-x Moreau, D., Clerc, J., Mansy-Dannay, A., & Guerrien, A. (2012). Enhancing spatial ability through sport practice: Evidence for an effect of motor training on mental rotation performance. Journal of Individual Differences, 33, 83–88. doi: 10.1027/1614-0001/a000075

Journal of Individual Differences 2016; Vol. 37(1):1–7

Moreau, D., Mansy-Dannay, A., Clerc, J., & Guerrien, A. (2011). Spatial ability and motor performance: Assessing mental rotation processes in elite and novice athletes. International Journal of Sport Psychology, 42, 525–547. Ozel, S., Larue, J., & Molinaro, C. (2002). Relation between sport activity and mental rotation: Comparison of three groups of subjects. Perceptual and Motor Skills, 95, 1141–1154. doi: 10.2466/pms.95.8.1141-1154 Pelgrims, B., Andres, M., & Olivier, E. (2009). Double dissociation between motor and visual imagery in the posterior parietal cortex. Cerebral Cortex, 19, 2298–2307. doi: 10.1093/cercor/bhn248 Penedo, F. J., & Dahn, J. R. (2005). Exercise and well-being: A review of mental and physical health benefits associated with physical activity. Current Opinion in Psychiatry, 18, 189–193. doi: 10.1097/00001504-200503000-00013 Pesce, C. (2012). Shifting the focus from quantitative to qualitative exercise characteristics in exercise and cognition research. Journal of Sport & Exercise Psychology, 34, 766–786. Pesce, C., Crova, C., Cereatti, L., Casella, R., & Bellucci, M. (2009). Physical activity and mental performance in preadolescents: Effects of acute exercise on free-recall memory. Mental Health and Physical Activity, 2, 16–22. doi: 10.1016/ j.mhpa.2009.02.001 Peters, M., Chisholm, P., & Laeng, B. (1995). Spatial ability, student gender and academic performance. Journal of Engineering Education, 84, 60–73. doi: 10.1002/j.21689830.1995.tb00148.x Peters, M., Laeng, B., Latham, K., Jackson, M., Zaiyouna, R., & Richardson, C. (1995). A redrawn Vandenberg and Kuse mental rotations test: Different version and factors that affect performance. Brain and Cognition, 28, 39–58. doi: 10.1006/ brcg.1995.1032 Petit, L. S., Pegna, A. J., Mayer, E., & Hauert, C.-A. (2003). Representation of anatomical constraints in motor imagery: Mental rotation of a body segment. Brain and Cognition, 51, 95–101. doi: 10.1016/S0278-2626(02)00526-2 Rakison, D. H., & Woodward, A. L. (2008). New perspectives on the effects of action on perceptual and cognitive development. Developmental Psychology, 44, 1209–1213. doi: 10.1037/a0012999 Shepard, R. N., & Metzler, J. (1971). Mental rotation of threedimensional objects. Science, 171, 701–703. doi: 10.1126/ science.171.3972.701 Sproston, K., & Primatesta, P. (2003). Health Survey for England 2002: The health of children and young people. London, UK: The Stationery Office. Valentine, T., & Bruce, V. (1988). Mental rotation of faces. Memory & Cognition, 16, 556–566. doi: 10.3758/BF03197057 Vandenberg, S. G., & Kuse, A. P. (1978). Mental rotations, a group test of three-dimensional spatial visualization. Perceptual and Motor Skills, 47, 599–604. doi: 10.2466/ pms.1978.47.2.599 Voelcker-Rehage, C., Godde, B., & Staudinger, U. M. (2011). Cardiovascular and coordination training differentially improve cognitive performance and neural processing in older adults. Frontiers in Human Neuroscience, 5, 26. doi: 10.3389/fnhum.2011.00026 Voyer, D., Voyer, S., & Bryden, M. P. (1995). Magnitude of sex differences in spatial abilities: A meta-analysis and consideration of critical variables. Psychological Bulletin, 117, 250–270. doi: 10.1037/0033-2909.117.2.250 Zacks, J. M., Mires, J., Tversky, B., & Hazeltine, E. (2002). Mental spatial transformation of objects and perspective. Spatial Cognition and Computation, 2, 315–332.

Ó 2016 Hogrefe Publishing

M. Schmidt et al.: Physical Activity and Mental Rotation

Date of acceptance: February 26, 2015 Published online: February 29, 2016

Mirko Schmidt Institute of Sport Science University of Bern Bremgartenstrasse 145 3012 Bern Switzerland Tel. +41 31 631 83 52 E-mail mirko.schmidt@ispw.unibe.ch

Ă&#x201C; 2016 Hogrefe Publishing

Journal of Individual Differences 2016; Vol. 37(1):1â&#x20AC;&#x201C;7

Original Article

Conscientiousness Is the Most Powerful Noncognitive Predictor of School Achievement in Adolescents Barbara Dumfart1 and Aljoscha C. Neubauer2 1

Academy of Lower Austria, Sankt Pölten, Austria, Department of Psychology, University of Graz, Austria

Abstract. Much research has demonstrated that intelligence and conscientiousness have a high impact on individual school achievement. To figure out if other noncognitive traits have incremental validity over intelligence and conscientiousness, we conducted a study on 498 eighthgrade students from general secondary schools in Austria. Hierarchical regressions for three criteria (GPA, science, and languages) were performed, including intelligence, the Big Five, self-discipline, grit, self-efficacy, intrinsic-extrinsic motivation, and test anxiety. Intelligence and conscientiousness alone accounted for approximately 40% in the variance of school achievement. For languages and GPA, no other personality and motivational predictors could explain additional variance; in science subjects, only self-discipline added incremental variance. We conclude that – in addition to intelligence as powerful cognitive predictor – conscientiousness is the crucial noncognitive predictor for school achievement and should be focused on when supporting students in improving their performance. Keywords: school achievement, intelligence, Big Five, motivation, anxiety

Individual academic achievement in adolescents is determined by various factors. Across different studies, one of the most important predictors is intelligence. Depending on the measure used, the average correlation between intelligence and school grades is about 0.5 (cf. Gustafsson & Undheim, 1996; Laidra, Pullmann, & Allik, 2007). Some studies show even higher relationships; in a 5-year prospective longitudinal study of about 70.000 English children (Deary, Strand, Smith, & Fernandes, 2007), a correlation of 0.81 between the g factor of intelligence and a latent trait of educational achievement was observed. Especially in lower grades and nonselective comprehensive schools, intelligence explains the largest amount of variance in school achievement. At higher levels of education, such as the tertiary educational system (college, university), intelligence is not as important anymore in the prediction of achievement (Chamorro-Premuzic & Furnham, 2005; Jensen, 1980). This might go back to the restriction of range because intelligence has already served as selection criterion for the admission to the higher educational track (Boekaerts, 1995). Intelligence generally seems to be more important for achievement in science subjects than in languages. In science, logical analysis plays a great role, whereas in arts, especially in languages, traits like social confidence are essential (Furnham, Rinaldelli-Tabaton, & ChamorroPremuzic, 2011). Even if intelligence is doubtlessly an important factor, there are many other variables which should be considered Journal of Individual Differences 2016; Vol. 37(1):8–15 DOI: 10.1027/1614-0001/a000182

in the prediction of school achievement. Especially when it comes to educational consulting, it is important to focus not only on the rather unmalleable trait intelligence, but also on intrapersonal strengths like personality and motivational variables. Noncognitive variables might be easier to train and more sensitive to intervention (Stankov, Lee, Luo, & Hogan, 2012). One of the central noncognitive variables to predict school achievement is conscientiousness. In a meta-analysis it has been shown that conscientiousness is the most consistent and stable personality predictor for academic achievement (Poropat, 2009). It combines various traits which are crucial for successful learning: for example, self-discipline, ambition, persistence, diligence, and dutifulness. The narrow traits of conscientiousness can predict academic achievement better than the broad trait (Paunonen & Ashton, 2001). Duckworth and Seligman (2005) found out that self-discipline accounted for more than twice as much variance as intelligence in school achievement and learning behavior of eighth-grade students. However, this result could be partly due to the fact that the study was conducted in a selective school and the consequential range restriction of intelligence. A conceptually related trait, which has lately been researched, is grit, defined as the ‘‘perseverance and passion for long-term-goals’’ (Duckworth, Peterson, Matthews, & Kelly, 2007, p. 1). Grit integrates aspects of achievement striving, self-control, and consistency of interests and encourages the realization of existing talents in an individual. Duckworth et al. (2007) conducted several studies in 2016 Hogrefe Publishing

B. Dumfart & A. C. Neubauer: Prediction of School Achievement in Adolescents

high-achieving persons and found out that grit was related to educational attainment and career stability. Openness and related traits, such as Typical Intellectual Engagement or Intellectual Curiosity, have also turned out to be important for academic achievement. Students who enjoy spending time with cognitively demanding tasks are more likely to perform well in school (Goff & Ackerman, 1992; Poropat, 2009; von Stumm, Hell, & Chamorro-Premuzic, 2011). On the other hand, extraversion turned out as a trait whose impact on academic achievement changes with age. While extraversion might be beneficial for school performance in earlier years of formal education, in higher levels of education it is negatively correlated with grades. This could be traced back to the changing – from social to formal – character of school across different levels of education (O’Connor & Paunonen, 2007). A similar pattern is found for agreeableness: in primary school, it seems to have relatively high impact on achievement, whereas it does not play a role in later years of education (Laidra et al., 2007). Most studies have found a negative relation between neuroticism and academic achievement (Poropat, 2009). If a student is very anxious, this might interfere with his attention to academic tasks and lead to poorer performance (De Raad & Schouwenburg, 1996). A trait which specifically addresses this issue is test anxiety. It has been shown that test anxiety is negatively correlated with academic achievement at different educational levels (Hembree, 1988; Rindermann & Neubauer, 2001). In addition to the discussed personality traits, we also considered motivational variables like self-efficacy and intrinsic versus extrinsic school motivation. Self-efficacy is defined as ‘‘beliefs in one’s capabilities to mobilize the motivation, cognitive resources, and courses of action needed to meet given situational demands’’ (Wood & Bandura, 1989, p. 408). This trait plays a role in various life areas; among others, it is associated with job search success (Saks, 2006), career success (Abele & Spurk, 2009), and academic achievement (Caprara, Vecchione, Alessandri, & Barbaranelli, 2011). Intrinsic-extrinsic motivation is described in the framework of Deci and Ryan’s (1985) self-determination theory. According to this theory, motivation is not only two-dimensional, but extrinsic motivation can be divided again into the components external regulation, introjected regulation, and identified regulation. Intrinsic regulation represents the highest level of self-determination, external regulation the lowest. Selfdetermined motivation turned out to be positively related to educational outcomes (Deci, Vallerand, Pelletier, & Ryan, 1991). Most studies on the prediction of academic achievement can be found in higher education (cf. Poropat, 2009). But if we want to support students early in improving their achievement to open them new possibilities for future

career, more research in earlier years of education is needed. In the present study, we include all of the described variables to find the best set of predictors for the school achievement of 13–14-year-old adolescents. In contrast to most of the studies mentioned above, which only focus on certain personality or motivational traits, our aim was to consider almost all variables that are often examined in the context of school achievement. It could be possible that some of the discussed variables turn out as redundant when simultaneously assessed with other (similar) traits.1 We did not only include various predictors of school achievement simultaneously, but also examined their impact on different criteria of school achievement, namely GPA, and science versus languages. In this, we tested the predictive power of the broad personality traits of the big five but, additionally, also whether the – in this context most relevant narrow personality traits – allow for an incremental prediction of the various school achievement indicators. The results of this study should allow us to conclude which personality and motivational traits are most important for students who want to improve their achievement and should, therefore, be assessed in counseling contexts. First, the relationships between school achievement and the selected personality and motivational variables shall be studied. Based on previous findings, we expect substantial relationships of the broad big five factors conscientiousness and openness with school achievement (Laidra et al., 2007; Poropat, 2009). Furthermore, we assume relationships with the narrow traits self-discipline, grit, test anxiety, selfefficacy, and intrinsic (versus extrinsic) motivation (Caprara et al., 2011; Deci et al., 1991; Duckworth et al., 2007; Duckworth & Seligman, 2005; Hembree, 1988). Second, the relative importance of the personality and motivation predictors compared to intelligence should be examined. Due to the characteristics of our sample that is not restricted in intellectual range, intelligence should have the highest impact on school achievement (ChamorroPremuzic & Furnham, 2005; Jensen, 1980). In this, we will also explore the question if any of the selected personality and motivational variables shows incremental validity above the hitherto most potent noncognitive predictor of school achievement, that is, conscientiousness. On the basis of the findings of Paunonen and Ashton (2001) as well as Lounsbury, Sundstrom, Loveland, and Gibson (2003) we expect all narrow traits to potentially improve the prediction. Third, similarities and discrepancies in the prediction of different criteria of school achievement – grade point average (GPA), science, and languages – shall be assessed. Based on the findings of Furnham et al. (2011), intelligence should be most important for science, whereas personality and motivational traits should be most important for languages.

Due to economic reasons, we could not include each relevant construct. In smaller pilot studies we found out that, for example, typical intellectual engagement could not predict school achievement in a sample of average-achieving adolescents. Therefore, this trait was omitted for the main study presented here.

2016 Hogrefe Publishing

Journal of Individual Differences 2016; Vol. 37(1):8–15

B. Dumfart & A. C. Neubauer: Prediction of School Achievement in Adolescents

Method Participants We tested 498 students from general secondary schools. This school track is more work-oriented (in contrast to academic secondary schools that prepare for university) and attended by approximately two-thirds of Austrian adolescents (Freudenthaler, Spinath, & Neubauer, 2008). General secondary schools are open for all children independent of previous school achievement. Due to incomplete data regarding school grades, 137 persons had to be excluded from the present analyses. The final sample consisted of 171 girls and 190 boys with a mean age of 14.09 years (SD = 0.48). All students participated voluntarily with informed consent of their parents. Cognitive abilities were tested at the end of seventh grade under supervision of trained testers whereas all other measures were taken a few months later (under supervision of specially trained teachers). While school achievement, intelligence, the Big Five, self-discipline, and self-efficacy were assessed in the entire sample, the remaining personality and motivational variables were only tested within smaller subsamples. Grit was measured in a sample of 129 persons (78 girls), test anxiety in a sample of 131 persons (49 girls), and intrinsic-extrinsic motivation in a sample of 94 persons (40 girls).

Measures and Procedure To obtain a measure for school achievement, students were asked to report all grades of their last certificate (end of 7th grade). For the GPA, we computed a weighted mean score consisting of following grades: German, English, Math, Physics, Biology, Geography, and History. German, English, and Math were double-weighted as the curriculum demands twice as many credit hours for these subjects. For obtaining a measure of science, a weighted mean score consisting of Biology, Physics, and Math (with a double weight on math) was computed. For languages, the mean of German and English was computed. For intelligence, three subscales of the German test ‘‘Intelligenz-Struktur-Analyse’’ (ISA; Blum et al., 1998) were used (verbal intelligence: ‘‘Gemeinsamkeiten finden’’ – commonalities; numeric intelligence: ‘‘Zahlenreihen fortsetzen’’ – number series; visuospatial intelligence: ‘‘Figuren zusammensetzen’’ – composition of figures). The time-restricted subscales comprise 20 items each and show internal consistencies between .80 and .89. An EFA of the subscales indicated a first unrotated factor with an eigenvalue of 1.78 accounting for 41% of the variance. For obtaining a measure of general intelligence, the factor scores for the first unrotated factor were used for further analyses. To assess the Big Five, the German version of the Hierarchical Personality Inventory for Children (HiPIC; Mervielde & De Fruyt, 2002; German version by Bleidorn & Ostendorf, 2009) was used. It comprises 144 items on Journal of Individual Differences 2016; Vol. 37(1):8–15

five factors, namely Neuroticism (N), Extraversion (E), Imagination (I), Benevolence (B), and Conscientiousness (C), on a 5-point Likert-type scale (responses range from barely characteristic of me to highly characteristic of me). The factors I and B can be regarded as equivalent to the Big Five factors openness and agreeableness. The inventory shows high reliabilities (a = .83–.88) in adolescents. For measuring self-discipline and general self-efficacy, new scales were developed that should be better suited for the target sample of 13-year-olds, because the existing questionnaires seem too difficult in verbal formulations to be understood by 13-year-old students encompassing the full range of verbal abilities. To generate the new items, we extracted contents from well-established questionnaires measuring related traits (e.g., NEO-PI-R; Costa & McCrae, 1992; HEXACO-PI-R; Lee & Ashton, 2004; Self-Control Scale; Tangney, Baumeister, & Boone, 2004; general selfefficacy [Allgemeine Selbstwirksamkeitserwartung]; Jerusalem & Satow, 1999) and reformulated the contents in much simpler ways than in the classical questionnaires that are targeted at adult persons. To ensure that the new items are adequate for our adolescent sample, we conducted cognitive interviews with the students as well as pilot studies. In their final versions, the self-discipline scale was comprised of six items, the self-efficacy scale of seven items. We used a 4-point Likert-type scale ranging from not appropriate to very appropriate. The results of the confirmatory factor analyses indicated one-dimensionality for each scale (self-discipline: v2(df = 9) = 25.17, p (Bollen-Stine Bootstrap) = .040; RMSEA = .048, CFI = .97, SRMR = .029; self-efficacy: v2(df = 14) = 31.22, p (Bollen-Stine Bootstrap) = .040; RMSEA = .040, CFI = .98, SRMR = .040). The internal consistencies (for the present sample) are acceptable (self-discipline: a = .63; self-efficacy: a = .72). The validity of the new scales had been tested in pilot studies. For self-discipline, construct validity was demonstrated by high correlations with the facet scales concentration (r = .55) and perseverance (r = .60) of the Big Five factor conscientiousness (assessed using the HiPIC; Mervielde & De Fruyt, 2002), as well as by either low or moderate correlations with the other HiPIC factors (neuroticism: r = .34, extraversion: r = .12, imagination: r = .25, benevolence: r = .41) and intrinsic motivation (r = .31; measured by the Academic Self-Regulation Questionnaire SRQ-A; Ryan & Connell, 1989; German version by Müller, Hanfstingl, & Andreitz, 2007). Criterion validity could be shown by positive relationships with time spent on learning (r = .12), time of day when homework is begun (r = .27) and negative relationships with watching TV (r = .14) and playing computer games (r = .21; measured by asking for the weekly time amount spent on certain activities). For selfefficacy, we found relationships which were similar to those of the self-efficacy scale by Jerusalem and Satow (1999), for example, high negative relationships with the HiPIC factor neuroticism (r = .50) and the scale lack of confidence of the Test-Anxiety Questionnaire PAF (r = .54; ‘‘Prüfungsangstfragebogen’’; Hodapp, Rohrmann, & Ringeisen, 2011). 2016 Hogrefe Publishing

B. Dumfart & A. C. Neubauer: Prediction of School Achievement in Adolescents

Table 1. Descriptive statistics and correlations among the variables 1. Sex 2. Age 3. GPA 4. Science 5. Lang 6. IQ 7. C 8. N 9. E 10. I 11. B 12. Self-d 13. Self-e 14. Grit 15. TA 16. SDI

361 358 327 361 339 361 359 357 358 359 357 360 358 129 131 94

– 14.09 3.59 3.60 3.39 0.00 3.40 2.65 3.36 3.41 3.41 7.08 20.19 19.04 45.66 10.64

– 0.48 0.90 0.91 1.01 0.85 0.48 0.58 0.47 0.51 0.39 1.39 3.20 3.37 7.75 11.03

1 – .01 .14* .09 .15** .18** .02 .28** .21** .08 .25** .05 .08 .09 .21* .18

2 – .16** .15** .18** .14* .03 .03 .07 .02 .04 .00 .02 .05 .05 .30**

– .94** .92** .55** .34** .16** .18** .31** .19** .23** .23** .29** .14 .19

– .76** .56** .31** .14** .11* .27** .20** .25** .22** .20* .19* .22*

– .46** .28** .13* .18** .29** .14* .13* .17** .23* .08 .12

– .16** .28** .10* .28** .00 .04 .24** .15 .29** .05

– .37** .31** .54** .40** .59** .47** .67** .24** .40**

– .26** .33** .17** .34** .50** .54** .65** .32**

– .57** .16** .12* .34** .21* .05 .14

– .19** .25** .53** .39** .14 .16

– .41** .20** .39** .05 .28**

– .32** .67** .12 .43**

– .61** .28** .29**

Notes. lang = languages; self-d = self-discipline; self-e = self-efficacy; TA = test anxiety; female = 1, male = 2. *p < .05. **p < .01 (two-tailed). Variables 14–16 could not be correlated because of non-overlapping samples.

To assess grit, the Short-Grit-Scale (Duckworth & Quinn, 2009), which consists of eight items, was translated into German. We used a 4-point Likert-type scale ranging from not appropriate to very appropriate. The internal consistency in our sample was sufficient (a = .70, N = 129). Test anxiety was measured by the German test-anxiety questionnaire PAF (Hodapp et al., 2011). The questionnaire comprises of 20 items and shows high internal consistency (a = .82–.90). Intrinsic-extrinsic motivation was assessed by a German adaptation of the SRQ-A (Ryan & Connell, 1989). It consists of 17 items on four scales: intrinsic, identified, introjected, and external regulation on a 5-point Likert-type scale ranging from very appropriate to not appropriate. The items are originally formulated to refer to only one school subject, but were used here to assess general scholastic motivation (e.g., ‘‘I mainly learn or study in school because it’s fun’’). The internal consistencies for the German version are satisfying (a = .67–.90). The self-determination index (SDI) is computed as suggested by the authors using the following formula: 2 · intrinsic + identified introjected 2 · external.

Results Descriptive statistics as well as correlations among the variables can be seen in Table 1. GPA as a composite of the other two criteria is – of course – highly correlated with them; interesting is only that science and languages still shared 58% of variance. Gender was slightly associated with GPA and languages (girls performed better). Age was slightly negatively correlated with all criteria. Intelligence was highly related to school achievement, especially

2016 Hogrefe Publishing

with GPA and science, somewhat lower with languages. The correlations between the Big Five factors and school achievement were rather similar for the three criteria; conscientiousness and imagination were moderately, extraversion and benevolence slightly correlated with school achievement. Neuroticism was slightly negatively associated with school achievement. Grit and self-efficacy were slightly to moderately positively associated with the criteria. Test anxiety and intrinsic-extrinsic motivation were only slightly related to science, but not to GPA and languages. To examine the relative importance of all personality and motivational predictors, we performed hierarchical regressions for all criteria. All regressions had the same structure; to control for sex and age, these variables were introduced in the first step. Due to the hypothesis that intelligence and conscientiousness are the most important predictor variables, they were included in the second step. In the third step, additional personality or motivational predictors were included to assess incremental variance of these variables. Since grit, test anxiety, and intrinsicextrinsic motivation were not tested within the entire sample, but each of them within a subsample, not all predictors could be entered in one regression. To examine if the subsamples for grit, extrinsic-intrinsic motivation, and test anxiety are comparable to the larger sample, we performed the main regression analyses (regressions 1a–c) again within the subsamples. The results deviated only marginally with respect to the beta weights, only some of them did no longer reach significance, which could be explained by the decreasing statistical power of smaller samples. The results, summarizing total R2 after step 1 and 2 R change after step 2, as well as the regression coefficients of the final model, can be found in Tables 2–4. In none of the regression analyses, the predictor variables showed multicollinearity. The final regression models of the entire

Journal of Individual Differences 2016; Vol. 37(1):8–15

B. Dumfart & A. C. Neubauer: Prediction of School Achievement in Adolescents

Table 2. Regressions 1a-c (r1a-r1c): Predicting school achievement by age, gender (step 1), intelligence, conscientiousness (step 2), the remaining Big Five, self-discipline, and self-efficacy (step 3) GPA (r1a, N = 318) 2

R Step 1

Science (r1b, N = 351) t

.05

Step 2

.44

Step 3

.45

Sex Age IQ C N E I B Self-d Self-e

Languages (r1c, N = 330) t

.03 F(2, 315) = 8.18** .39 DF(2, 313) = 107.48** .01 DF(6, 307) = 0.80 .21 4.31** .09 2.16* .56 12.06** .18 2.85** .06 1.09 .00 0.08 .01 0.20 .04 0.88 .09 1.63 .02 0.29

.41 .43

DR2

.06 F(2, 348) = 6.07** .38 DF(2, 346) = 110.18** .02 DF(6, 340) = 2.19* .15 3.26** .08 1.85 .58 12.91** .15 2.50* .10 1.80 .05 0.89 .00 0.05 .06 1.33 .13 2.30* .04 0.66

.32 .32

F(2, 327) = 10.34** .26 DF(2, 325) = 62.05** .00 DF(6, 316) = 0.22 .22 4.07** .13 2.67** .44 8.70** .19 2.77** .01 0.24 .01 0.08 .06 0.89 .01 0.25 .01 0.21 .02 0.44

Notes. self-d = self-discipline; self-e = self-efficacy. *p < .05. **p < .01 (two-tailed).

sample (regressions 1aâ&#x20AC;&#x201C;c; Table 2) accounted for 45% of the variance in GPA and for 32% of the variance in languages. In the prediction of GPA and languages, only steps 1 and 2 (IQ, C) contributed significantly. The remaining Big Five factors, as well as self-discipline and self-efficacy (step 3), did not explain incremental variance. Intelligence was the strongest predictor, explaining 26% of unique variance in GPA and 16% in languages. Only in science, one of the step 3 variables (self-discipline) could marginally enhance the prediction, uniquely explaining 1% of the variance. Intelligence was the strongest predictor, explaining 34% of unique variance. In subsample 1 (regressions 2aâ&#x20AC;&#x201C;c; Table 3), the incremental contribution of grit over and above intelligence and conscientiousness was examined. Similar to the findings of the full sample, the amount of prediction could only be increased significantly in steps 1 and 2. Adding grit in the third step could not enhance the prediction, either for GPA, or for science or languages. Due to nonsignificant zero-order correlations between the predictors test anxiety and intrinsic-extrinsic motivation with the criteria GPA and languages, regressions for these traits were only performed for science. In subsample 2 (regression 3; Table 4), we analyzed the incremental variance of test anxiety by adding it in the third step. It could not increase the amount of prediction. In subsample 3 (regression 4; Table 4), intrinsic-extrinsic motivation was entered in the third step of the regression and could also not account for incremental variance.2

Discussion We examined the impact of specific personality and motivational predictors on the school achievement of eighthgraders comparing three criteria: GPA (average across all subjects), science subjects, and language subjects. First, we deal with the bivariate correlations between school achievement and all personality and motivational predictors: The correlation between intelligence and school achievement was high; compared to previous findings (Furnham et al., 2011) it was considerably higher with science than with languages. Although the three criterion variables were strongly intercorrelated, we found different patterns of correlates with personality and motivational traits. All Big Five traits as well as grit and self-efficacy showed the highest associations with GPA. This could easily be explained by the presumed higher reliability of this criterion compared to the others due to the higher level of aggregation. By contrast, two of the narrow traits, namely test anxiety and intrinsic-extrinsic motivation, correlated significantly only with science. Comparable results could be found in Spinath, Freudenthaler, and Neubauer (2010): neuroticism showed incremental validity over intelligence and ability self-perceptions only in Math achievement, not in languages. Steinmayr and Spinath (2007) obtained similar results and conjectured that test anxiety seems to be more important in domains where the correctness of answers can be logically determined more easily, as it is the case with Math. These findings could be interpreted in a

All regression analyses were repeated entering the narrow traits (self-discipline, self-efficacy, grit, test anxiety, and intrinsic-extrinsic motivation) in the second step and conscientiousness in the third step. The results indicated significant contributions of all narrow traits in the second step which were reduced to nonsignificance when adding conscientiousness in the following step. Conscientiousness accounted for incremental variance over and above the narrow traits (for space constraints these results are not presented here but can be obtained from the authors).

Journal of Individual Differences 2016; Vol. 37(1):8â&#x20AC;&#x201C;15

2016 Hogrefe Publishing

B. Dumfart & A. C. Neubauer: Prediction of School Achievement in Adolescents

Table 3. Regressions 2a–c (r2a–r2c): Predicting school achievement in subsample 1 by age, gender (step 1), intelligence, conscientiousness (step 2), and Grit (step 3) GPA (r2a, N = 107) R Step 1

Science (r2b, N = 126) 2

.08

Step 2

.48

Step 3

.49

Sex Age IQ C Grit

Languages (r2c, N = 114) 2

.08 F(2, 104) = 4.25** .41 DF(2, 102) = 40.23** .00 DF(1, 101) = 0.55 .21 2.80** .15 1.97 .53 6.76** .23 2.20** .08 0.74

.44 .44

DR2

.09 F(2, 123) = 5.10** .36 DF(2, 121) = 39.16** .00 DF(1, 120) = 0.02 .11 1.63 .16 2.22* .53 7.04** .21 2.25* .01 0.14

F(2, 111) = 5.75** .30 DF(2, 109) = 27.39** .00 DF(1, 108) = 0.43 .25 3.34** .17 2.16* .47 5.74** .17 1.57 .07 0.65

.40 .40

Note. *p < .05. **p < .01 (two-tailed).

way that strategies to encourage the intrinsic motivation and reduce the test anxiety of students are of particular importance in Math, maybe generally in science subjects. Most bivariate correlations between school achievement and the predictor variables correspond to previous findings (Chamorro-Premuzic & Furnham, 2005; De Raad & Schouwenburg, 1996; Poropat, 2009): moderate relations with conscientiousness, imagination (openness), self-discipline, grit, and self-efficacy as well as weak relations with benevolence (agreeableness), test anxiety, and intrinsicextrinsic motivation. Contrary to previous findings, extraversion was also slightly positively correlated with school achievement, especially with GPA and languages. Spinath et al. (2010) obtained similar results in eighth-graders (but only in girls) which they explained by the higher importance of oral performance in languages. If participating actively during class is requested, extraverted students may have advantages in these subjects. The hierarchical regressions indicated that all selected predictors together explained almost half of the variance in adolescent school achievement. Intelligence was by far

the most important predictor; although conscientiousness could uniquely contribute to the prediction, the impact of intelligence was much higher. This result is comparable to findings of studies in similar populations; in nonselective schools where intelligence is not (yet) range-restricted (Freudenthaler et al., 2008; Laidra et al., 2007). With respect to the GPA and languages, girls performed significantly better than boys while there was no difference in sciences. This is in line with previous findings (Freudenthaler et al., 2008). Age had a negative impact on school achievement. This may seem counterintuitive, but it can be explained by the fact that all students were basically from the same age cohort; most students who were older than the average have likely repeated at least one grade because of poor performance. Although most bivariate correlations were as high as expected on the basis of the literature, the third step of the hierarchical regressions showed that – in addition to conscientiousness – no other personality or motivational traits could substantially enhance the prediction. Selfdiscipline could account for only 1% of incremental

Table 4. Regressions 3 and 4 (r3, r4): Predicting Science by age, gender (step 1), intelligence, conscientiousness (step 2), and test anxiety (step 3; subsample 2), respectively, intrinsic-extrinsic motivation (step 3; subsample 3) Science (r3, N = 94) 2

R Step 1

Science (r4, N = 130) t

.05

Step 2

.49

Step 3

.49

Sex Age IQ C Test anxiety

F(2, 127) = 3.23** .44 DF(2, 125) = 53.59** .00 DF(1, 124) = 0.03 0.25 0.01 0.63 0.26 0.01

3.78** 0.08 9.04** 3.92** 0.18

R Step 1

.00

Step 2

.29

Step 3

.31

Sex Age IQ C SDI

DR2

F(2, 91) = 0.03 .28 DF(2, 89) = 17.67** .02 DF(1, 88) = 3.03 0.09 0.05 0.48 0.15 0.18

0.94 0.55 5.19** 1.49 1.74

Note. **p < .01 (two-tailed). 2016 Hogrefe Publishing

Journal of Individual Differences 2016; Vol. 37(1):8–15

B. Dumfart & A. C. Neubauer: Prediction of School Achievement in Adolescents

variance over intelligence and conscientiousness in the prediction of science; none of the narrow traits could enhance the prediction of GPA or languages. Broad traits could substantially improve the prediction over and above narrow traits, but not vice versa. That implies that in this age range tested here there seems to be little benefit of assessing narrow traits like self-discipline or grit for counseling contexts. For the target population of average-achieving adolescents it would be sufficient to examine broad personality traits, particularly conscientiousness, when supporting them in improving school achievement. The narrower traits might be of relevance in higher age ranges or more selective schools like academic high schools. It may also be possible that the broad traits outperformed the narrow traits because of the broad performance criteria used in this study. It has been shown that narrow traits are useful when predicting specific performance criterion, as for instance, counterproductive work behaviors. However, when predicting a global criteria, like overall job performance, broad traits are preferable (Ones & Viswesvaran, 1996; Dudley, Orvis, Lebiecki, & Cortina, 2006). Therefore, it would be interesting to extend our study to additional performance criteria in school, for example, class participation or oral exam performance. In summary, intelligence and conscientiousness turned out – once again – as the most powerful predictors for school achievement of adolescents. No other personality and motivational variables (the remaining Big Five, selfdiscipline, grit, self-efficacy, intrinsic-extrinsic motivation, and test anxiety) could substantially enhance the prediction, although most of them were correlated with grades. As a restriction of our study it should be mentioned that on the basis of a thorough literature research we included only the most discussed variables with respect to the prediction of school achievement. Due to economic reasons, we could not test each variable which has turned out to be associated with school achievement but focused on those which turned out to be of importance in samples of comparable age ranges and school types. For future directions, it would be interesting to explore the impact of certain other constructs over and above personality and motivation. For example, it would be interesting to also include approaches to learning. Previous research demonstrated interactions between personality traits and approaches to learning regarding academic achievement (Diseth, 2003). Based on our results, we conclude that conscientiousness is the crucial noncognitive trait in school achievement of adolescents. Some authors suggest that high conscientiousness can even compensate for low (fluid) intelligence; that is, students with lower intelligence and high conscientiousness could perform as well as their more intelligent colleagues who do not show such structured and persevering learning habits (Moutafi, Furnham, & Paltiel, 2004; Wood & Englert, 2009). So it might be useful to focus on the impact of conscientiousness in improving school achievement, for example, for school career counselors. Some conscientious behaviors – like being on time, tidying up the workplace, or keeping focused on a task – can be

Journal of Individual Differences 2016; Vol. 37(1):8–15

trained with little effort but might have considerable influence on school achievement.

References Abele, A. E., & Spurk, D. (2009). The longitudinal impact of self-efficacy and career goals on objective and subjective career success. Journal of Vocational Behavior, 74, 53–62. Bleidorn, W., & Ostendorf, F. (2009). Ein Big Five-Inventar für Kinder und Jugendliche. Die deutsche Version des Hierarchical Personality Inventory for Children (HiPIC) [A Big Five inventory for children and adolescents: The German version of the Hierarchical Personality Inventory for Children (HiPIC)]. Diagnostica, 55, 160–173. Blum, F., Didi, H. J., Fay, E., Maichle, U., Trost, G., Wahlen, J. H., & Gittler, G. (1998). Intelligenz Struktur Analyse (ISA). Ein Test zur Messung der Intelligenz [Intelligence Structure Analysis. A test for measuring intelligence]. Frankfurt, Germany: Swets & Zeitlinger B. V., Swets Test Services. Boekaerts, M. (1995). The interface between intelligence and personality as determinants of classroom learning. In D. H. Saklofske & M. Zeidner (Eds.), Handbook of Personality and Intelligence (pp. 161–183). New York, NY: Plenum. Caprara, G. V., Vecchione, M., Alessandri, M. G., & Barbaranelli, C. (2011). The contribution of personality traits and self-efficacy beliefs to academic achievement: A longitudinal study. The British Journal of Educational Psychology, 81, 78–96. Chamorro-Premuzic, T., & Furnham, A. (2005). Personality and intellectual competences. Mahwah, NJ: Erlbaum. Costa, P. T. Jr., & McCrae, R. R. (1992). Revised NEO personality inventory (NEO-PI-R) and NEO five-factor inventory (NEO-FFI): Professional manual. Odessa, FL: Psychological Assessment Resources. Deary, I. J., Strand, S., Smith, P., & Fernandes, C. (2007). Intelligence and educational achievement. Intelligence, 35, 13–21. Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and selfdetermination in human behavior. New York, NY: Plenum. Deci, E. L., Vallerand, R. J., Pelletier, L. G., & Ryan, R. M. (1991). Motivation and education: The self-determination perspective. The Educational Psychologist, 26, 325–346. De Raad, B., & Schouwenburg, H. (1996). Personality in learning and education: A review. European Journal of Personality, 10, 406–425. Diseth, Å. (2003). Personality and approaches to learning as predictors of academic achievement. European Journal of Personality, 17, 143–156. Duckworth, A. L., Peterson, C., Matthews, M. D., & Kelly, D. R. (2007). Grit: Perseverance and passion for long-term goals. Journal of Personality & Social Psychology, 92, 1087–1101. Duckworth, A. L., & Quinn, P. D. (2009). Development and validation of the Short Grit Scale (GRIT–S). Journal of Personality Assessment, 91, 166–174. Duckworth, A. L., & Seligman, M. E. P. (2005). Self-discipline outdoes IQ in predicting academic performance of adolescents. Psychological Science, 16, 939–944. Dudley, N. M., Orvis, K. A., Lebiecki, J. E., & Cortina, J. M. (2006). A meta-analytic investigation of conscientiousness in the prediction of job performance: Examining the intercorrelations and the incremental validity of narrow traits. Journal of Applied Psychology, 91, 40.

2016 Hogrefe Publishing

B. Dumfart & A. C. Neubauer: Prediction of School Achievement in Adolescents

Freudenthaler, H. H., Spinath, B., & Neubauer, A. C. (2008). Predicting school achievement in boys and girls. European Journal of Personality, 22, 231–245. Furnham, A., Rinaldelli-Tabaton, E., & Chamorro-Premuzic, T. (2011). Personality and intelligence predict arts and science school results in 16 year olds. Psychologia, 54, 39–51. Goff, M., & Ackerman, P. L. (1992). Personality-intelligence relations: Assessment of typical intellectual engagement. Journal of Educational Psychology, 84, 537–552. Gustafsson, J.-E., & Undheim, J. O. (1996). Individual differences in cognitive functions. In D. C. Berliner & R. C. Calfee (Eds.), Handbook of educational psychology (pp. 186–242). New York, NY: Prentice Hall. Hembree, R. (1988). Correlates, causes, effects, and treatment of test anxiety. Review of Educational Research, 58, 47–77. Hodapp, V., Rohrmann, S., & Ringeisen, T. (2011). PAF – Prüfungsangstfragebogen [Test anxiety questionnaire]. Göttingen, Germany: Hogrefe. Jensen, A. R. (1980). Bias in mental testing. New York, NY: The Free Press. Jerusalem, M., & Satow, L. (1999). Schulbezogene Selbstwirksamkeitserwartung [School-related self-efficacy]. In R. Schwarzer & M. Jerusalem (Eds.), Skalen zur Erfassung von Lehrer- und Schülermerkmalen (pp. 15–16). Berlin, Germany: Freie Universität Berlin. Laidra, K., Pullmann, H., & Allik, J. (2007). Personality and intelligence as predictors of academic achievement: A crosssectional study from elementary to secondary school. Personality and Individual Differences, 42, 441–451. Lee, K., & Ashton, M. C. (2004). Psychometric properties of the HEXACO personality inventory. Multivariate Behavioral Research, 39, 329–358. Lounsbury, J. W., Sundstrom, E., Loveland, J. L., & Gibson, L. W. (2003). Broad versus narrow personality traits in predicting academic performance of adolescents. Learning and Individual Differences, 14, 65–75. Mervielde, I., & De Fruyt, F. (2002). Assessing children’s traits with the Hierarchical personality inventory for children. In B. De Raad & M. Perugini (Eds.), Big Five assessment (pp. 129–142). Göttingen, Germany: Hogrefe. Moutafi, J., Furnham, A., & Paltiel, L. (2004). Why is conscientiousness negatively correlated with intelligence? Personality and Individual Differences, 37, 1013–1022. Müller, F. H., Hanfstingl, B., & Andreitz, I. (2007). Skalen zur motivationalen Regulation beim Lernen von Schülerinnen und Schülern: Adaptierte und ergänzte Version des Academic Self-Regulation Questionnaire (SRQ-A) nach Ryan & Connell [Scales to assess motivational regulation during students’ learning]. Wissenschaftliche Beiträge aus dem Institut für Unterrichts- und Schulentwicklung Nr. 1. Klagenfurt: Alpen-Adria-Universität. O’Connor, M. C., & Paunonen, S. V. (2007). Big Five personality predictors of postsecondary academic performance. Personality and Individual Differences, 43, 971–990. Ones, D. S., & Viswesvaran, C. (1996). Bandwidth-fidelity dilemma in personality measurement for personnel selection. Journal of Organizational Behavior, 17, 609–626. Paunonen, S. V., & Ashton, M. C. (2001). Big Five predictors of academic achievement. Journal of Research in Personality, 35, 78–90.

2016 Hogrefe Publishing

Poropat, A. E. (2009). A meta-analysis of the five-factor model of personality and academic performance. Psychological Bulletin, 135, 322–338. Rindermann, H., & Neubauer, A. C. (2001). The influence of personality on three aspects of cognitive performance: Processing speed, intelligence and school performance. Personality and Individual Differences, 30, 829–842. Ryan, R. M., & Connell, J. P. (1989). Perceived locus of causality and internalization: Examining reasons for acting in two domains. Journal of Personality and Social Psychology, 57, 749–761. Saks, A. M. (2006). Multiple predictors and criteria of job search success. Journal of Vocational Behavior, 68, 400–415. Spinath, B., Freudenthaler, H. H., & Neubauer, A. (2010). Domain-specific school achievement in boys and girls predicted by intelligence, personality and motivation. Personality and Individual Differences, 48, 481–486. Stankov, L., Lee, J., Luo, W., & Hogan, D. J. (2012). Confidence: A better predictor of academic achievement than selfefficacy, self-concept and anxiety? Learning and Individual Differences, 22, 747–758. Steinmayr, R., & Spinath, B. (2007). Predicting School Achievement from Motivation and Personality. Zeitschrift für Pädagogische Psychologie, 21, 207–216. Tangney, J. P., Baumeister, R. F., & Boone, A. L. (2004). High self-control predicts good adjustment, less pathology, better grades, and interpersonal success. Journal of Personality, 72, 271–323. von Stumm, S., Hell, B., & Chamorro-Premuzic, T. (2011). Curiosity predicts academic performance. The hungry mind: Intellectual curiosity as third pillar of academic performance. Perspectives on Psychological Science, 6, 574–588. Wood, P., & Englert, P. (2009). Intelligence compensation theory: A critical examination of the negative relationship between conscientiousness and fluid and crystallised intelligence. The Australian and New Zealand Journal of Organisational Psychology, 2, 19–29. Wood, R., & Bandura, A. (1989). Impact of conceptions of ability on self-regulatory mechanisms and complex decision making. Journal of Personality and Social Psychology, 56, 407–415.

Date of acceptance: March 23, 2015 Published online: February 29, 2016

Aljoscha C. Neubauer Dept. of Psychology University of Graz Maiffredygasse 12b 8010 Graz Tel. +43 316 380–5124 E-mail aljoscha.neubauer@uni-graz.at

Journal of Individual Differences 2016; Vol. 37(1):8–15

Original Article

Perception of Emotional Expressions in Adults The Role of Temperament and Mood Chit Yuen Yi,1 Matthew W. E. Murry,2 and Amy L. Gentzler1 1

Department of Psychology, West Virginia University, Morgantown, WV, USA, 2 Department of Psychology, Northeastern University, Boston, MA, USA

Abstract. Past research suggests that transient mood influences the perception of facial expressions of emotion, but relatively little is known about how trait-level emotionality (i.e., temperament) may influence emotion perception or interact with mood in this process. Consequently, we extended earlier work by examining how temperamental dimensions of negative emotionality and extraversion were associated with the perception accuracy and perceived intensity of three basic emotions and how the trait-level temperamental effect interacted with state-level selfreported mood in a sample of 88 adults (27 men, 18–51 years of age). The results indicated that higher levels of negative mood were associated with higher perception accuracy of angry and sad facial expressions, and higher levels of perceived intensity of anger. For perceived intensity of sadness, negative mood was associated with lower levels of perceived intensity, whereas negative emotionality was associated with higher levels of perceived intensity of sadness. Overall, our findings added to the limited literature on adult temperament and emotion perception. Keywords: temperament, emotion perception, event recall, mood, adults

Facial expressions of emotions are embedded in everyday interpersonal interactions. Therefore, the ability to perceive other people’s emotions from facial expressions is critical for navigating the social world because others’ emotional states enable us to make inferences about their attitudes and intentions (Ekman, 1993). Given the importance of emotion perception, a substantial body of research has focused on how people vary in their perception of others’ emotions. For instance, past research has investigated contextual or situational influences (e.g., scenarios being paired with the facial stimuli; Milanak & Berenbaum, 2014) or age difference and developmental influence on emotion perception (e.g., Isaacowitz et al., 2007; Ruffman, Henry, Livingstone, & Phillips, 2008; Trentacosta & Fine, 2010; Widen & Russell, 2003). Another key variable is the perceiver’s own emotional state, but this research has mainly focused on the effect of transient mood. The role of mood is important in understanding how affect influences emotion perception (Lee, Ng, Tang, & Chan, 2008; Schmid & Mast, 2010). Yet, enduring temperamental or personality dimensions also appear to be influential (Young & Brunet, 2011). Our study is the first (to our knowledge) to explicitly test for the influence of both state-linked mood and traitlinked temperament on the perception of emotional facial expressions in young adults. This study can therefore offer insight into the importance of both factors concurrently and if the effect of one factor is magnified in the presence of the other (e.g., if being higher in temperamental negative emotionality makes negative mood particularly impactful Journal of Individual Differences 2016; Vol. 37(1):16–23 DOI: 10.1027/1614-0001/a000183

because these individuals might see anger as more intense). We specifically focused on two aspects of emotion perception: (1) perception accuracy of emotion and (2) perceived intensity of emotional facial expressions.

Mood and Emotion Perception Mood can influence perceptual processes by facilitating the processing of mood-congruent information and hindering the processing of mood-incongruent information (Bower, 1981). In the context of perception of facial expressions, individuals may be biased toward mood-congruent emotional cues, and thus may be more accurate when identifying mood-congruent facial expressions. In extreme cases of mood disturbance, depressed, and manic individuals appear to have altered emotion perception (e.g., Cooley & Nowicki, 1989; Lembke & Ketter, 2002; Surguladze et al., 2004). Specifically, depressed individuals were less accurate in identifying happy facial expressions, compared to healthy individuals (Cavanagh & Geisler, 2006; Surguladze et al., 2004). In contrast, Bipolar I patients who were in a manic state showed impairment in identification of sadness from facial expressions (Lennox, Jacob, Calder, Lupson, & Bullmore, 2004). Similarly, research with healthy participants has found mood-specific biases in emotion perception (Bouhuys, Bloem, & Groothuis, 1995; Lee et al., 2008; Schmid & Mast, 2010). For instance, participants who were induced 2016 Hogrefe Publishing

C. Y. Yi et al.: Adult Temperament and Emotion Perception

to experience a sad mood, compared to those induced to feel happy or neutral, perceived more sadness in morphed faces (blended expressions of a positive emotion and a negative emotion; Lee et al., 2008) or in line-drawn faces (Bouhuys et al., 1995). After watching mood-evoking film clips, sad participants demonstrated higher identification accuracy for sad facial expressions than for happy ones, whereas happy participants showed a nonsignificant trend of higher accuracy of identifying happy expressions as compared to sad ones (Schmid & Mast, 2010). Together, these findings suggest that transient mood states could make mood-congruent materials more salient and accessible, which could influence people’s interpretation of other’s facial expression. However, it is worth noting these studies have mostly focused on identification of prototypical facial expressions of emotion, which can be prone to ceiling effects (Egger et al., 2011). Ceiling effects are likely because stimuli displaying prototypical facial expressions are designed to show a single emotion and be clearly representative of that emotion. Thus, people are able to often quickly and accurately identify these prototypical emotional expressions (Egger et al., 2011; Hess, Blairy, & Kleck, 1997; Surguladze et al., 2004). In the current study, in addition to the identification of expressions, we also assessed participants’ perceived intensity of emotional expressions to better capture the potential variations in emotion perception. Perceived intensity of emotion is important because even if people identify the same discrete emotion, they could infer different intensity levels of the emotion (Matsumoto, Kasri, & Kooken, 1999). Past research has shown individual variations in perception of facial expression intensity based on one’s cultural identity, current diagnosis of depression, age, and gender (Matsumoto et al., 1999; McClure, 2000; Hoffmann, Kessler, Eppel, Rukavina, & Traue, 2010; Yoon, Joormann, & Gotlib, 2009). To our knowledge, studies regarding the effect of mood on perceived intensity of facial expression of emotions have been underrepresented in healthy adult population (for exception, see Schiffenbauer, 1974). According to Schwarz’s (1990) affect-as-information model, mood can serve an informative function to help people evaluate others’ emotion. One’s mood state may increase the availability of mood-congruent information, which could bias their judgments of emotional intensity (e.g., people may perceive anger as more intense when in a negative mood state).

Temperament and Emotion Perception Despite the research establishing state-level mood effects on emotion perception (e.g., Bouhuys et al., 1995; Lee et al., 2008; Schmid & Mast, 2010), whether trait-level emotionality also impacts people’s perception of emotional facial expressions is not widely understood. Because temperament pertains to biologically-based individual differences in emotional arousal, expression, and regulation (Rothbart & Derryberry, 1981), it may influence how an individual interprets and responds to another person’s 2016 Hogrefe Publishing

emotional expressions (Buchanan, Bibas, & Adolphs, 2010). Two facets of adult temperament that should be particularly relevant to emotion perception are positive emotionality and negative emotionality, both of which relate to a person’s emotional reactivity. Positive emotionality is characterized by frequent experience of intense positive affect. However, positive emotionality is often subsumed under the broader temperament dimension of extraversion or surgency that also includes seeking pleasure from social interactions and activities involving high stimulus intensity or novelty (Evans & Rothbart, 2007). We similarly refer to the higher-order dimension as extraversion. Negative emotionality refers to the propensity to experience unpleasant affect, including sadness, frustration, and fear (Evans & Rothbart, 2007). These temperamental dimensions correspond with pervasive patterns in emotional experiences (extraversion predicting positive affect and negative emotionality predicting negative affect; e.g., Tsai, Levenson, & McCoy, 2006). Examining the role of temperament in conjunction with mood can shed light on whether affect-related variations in emotion perception mainly stem from the transient mood state or whether they are better explained by more stable emotional patterns (i.e., temperament). To our knowledge, few studies have specifically examined how the temperament dimension of extraversion and negative emotionality influence emotion perception. However, a meta-analysis has provided some evidence that extraverted individuals tend to have higher levels of interpersonal sensitivity, a broader construct of interpersonal perception that includes perception of others’ emotions (Hall, Andrzejewski, & Yopchick, 2009). In another study on temperamental characteristics and emotion perception, Young and Brunet (2011) found that higher level of sociability (a component of temperamental extraversion; Evans & Rothbart, 2007) was related to better accuracy in recognizing facial expressions of emotions, but only when there was a time constraint. As suggested by the authors, these results indicate that highly-sociable individuals may develop an advantage in quickly recognizing others’ emotions as a result of their frequent social interaction. The temperament of negative emotionality has not been explicitly examined as the predictor of emotion perception. However, previous research has shown that experiences with intense negative emotions can influence how one perceives others’ emotional expressions. In a field study with a community sample of museum visitors (Buchanan et al., 2010), participants were asked to adjust the facial features of a stimulus until it resembles a specific emotional expression. The results suggested that people who reported more intense experiences of fear and happiness had more accurate conceptions of the expressions of fear and happiness as evidenced by the close resemblance between their task responses and the prototypical expressions from the Ekman’s face database (Ekman & Friesen, 1976). Together, these studies have provided empirical support that individuals’ temperament and emotional experiences may shape the development of emotion perception skills due to familiarity and direct experiences with particular emotional states. Journal of Individual Differences 2016; Vol. 37(1):16–23

C. Y. Yi et al.: Adult Temperament and Emotion Perception

Current Study We conducted a laboratory study with young adults to assess how state-level mood and trait-level emotionality influence perception of emotional facial expressions. We used three prototypical emotional expressions from the same database (Ekman & Friesen, 1976) as Young and Brunet (2011). Based on previous work regarding moodcongruity effect on emotion perception accuracy (Lee et al., 2008; Schmid & Mast, 2010), we hypothesized that positive emotionality (i.e., extraversion) would be associated with higher scores of perception accuracy of positive emotion (i.e., happiness). Additionally, we expected that negative emotionality would predict higher scores of perception accuracy of negative emotions (i.e., anger and sadness). Because research regarding the effect of mood and temperament on perceived intensity of emotions was limited, we based our intensity hypotheses on Schwarz’s (1990) affect-as-information model and hypothesized that extraversion would be associated with higher perceived intensity of happiness, whereas negative emotionality would predict higher perceived intensity of anger and sadness. Although no prior research addressed both temperament and mood, we sought to investigate the possible interaction effect of adult temperament and mood on emotion perception. Therefore, on an exploratory basis, we examined whether the hypothesized temperamental effects would be more pronounced when the temperamental dimension and the mood are of the same valence (e.g., extraversion more strongly relates to seeing happiness as more intense when participants were in a more positive mood).

Method Participants The sample was 88 adults (69.3% women; Mage = 20.44; SDage = 4.42) who received psychology course credit for their participation in a laboratory study at a large public university. The racial composition of the sample was 88.6% Caucasian, 2.3% African Americans, 2.3% Latino, 1.1% Asian Americans, and 5.7% identified as Multiracial. The majority of the participants were college freshmen or sophomores (64.8%). To be eligible for participation, the individuals had to be at least 18 years old and not currently being treated for depression.

Procedure Laboratory Procedure After providing informed consent, participants completed mood and temperament questionnaires along with other measures online. Then, they performed several baseline

Journal of Individual Differences 2016; Vol. 37(1):16–23

cognitive tasks, along with the emotion perception task. The cognitive tasks are not the focus of the current investigation and will not be discussed further. Next, participants experienced three mood induction procedures (a failure task, negative event recall, and positive event recall). They completed variations of the emotion perception task only after the two event recalls. Thus, we do not discuss the failure task further. At the end of the study, participants were thanked and fully debriefed. Event Recall Participants were asked to think of an upsetting (for negative mood induction) and happy (for positive mood induction) event and then describe it to the experimenter. Their event descriptions were recorded for later analysis, but were not part of this investigation. Recalling emotional episodes has been used in previous studies to successfully induce mood states (e.g., Schwarz & Clore, 1983). Emotion Perception Task The participants completed versions of the perception task three times: at baseline, following the negative event recall, and after the positive event recall. The emotion perception task was delivered using PXLab software (Irtel, 2007) on a designated laptop. All 36 black-and-white face stimuli were drawn from the Ekman face database (Ekman & Friesen, 1976). The stimuli were posed by 14 posers (8 females; 6 males). We selected 9 happy, 9 angry, 9 neutral, and 9 sad facial expressions from the database. All selected stimuli (except the neutral ones) were judged to show the intended emotion by the majority (75–100%) of the raters in Ekman and Friesen (1976). We did not include other negative emotions (i.e., fear, disgust) because we wanted to have a better balance between the number of positive and negative emotions. There were 12 stimuli presented at each time point. Four additional stimuli were used as practice immediately prior to the baseline photo presentation. The face stimuli were presented one at a time at the center of the computer screen for a fixed duration of 800 ms to allow enough time for the participants to make identification and judge intensity. Each face stimulus was preceded by a fixation cross of the same duration to direct the participants’ attention to the center of the computer screen. For each stimulus, participants were instructed to respond to a force-choice question (happy, sad, anger, or neutral) regarding the type of emotion displayed by the face and to indicate their perceived intensity of the emotion on a visual analog scale (ranged from 0 to 10) anchored by not at all on one end and extremely intense on the opposite end. In the experimental sessions, the face stimuli were presented in standard order such that neither the same poser nor the same emotion would appear consecutively. Because we are interested in the effects of temperament and mood on perception of emotional expression, task results regarding neutral expression will not be discussed further.

2016 Hogrefe Publishing

C. Y. Yi et al.: Adult Temperament and Emotion Perception

Materials

Table 1. Descriptive statistics for perception task variables Accuracy

Mood Assessment

Intensity

Range

0.99 0.87 0.62

0.06 0.19 0.31

0.67–1.00 0.33–1.00 0.00–1.00

3.84 3.29 2.81

0.50 0.65 0.57

2.68–5.00 1.74–4.96 1.33–4.29

Test 1 (post-negative mood induction) Happy 0.99 0.05 0.67–1.00 3.82 Anger 0.99 0.04 0.67–1.00 3.75 Sad 0.85 0.25 0.00–1.00 2.62

0.50 0.63 0.79

2.56–4.87 1.66–4.93 0.20–4.28

Test 2 (post-positive mood induction) Happy 0.99 0.05 0.67–1.00 3.56 Anger 0.95 0.12 0.67–1.00 3.18 Sad 0.83 0.23 0.33–1.00 2.71

0.54 0.69 0.80

1.98–4.91 0.96–4.93 0.52–4.98

To measure participants’ mood, they were given the Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988) at the beginning of the study and immediately after each event recall procedure. We asked participants to rate the 20 emotions on an 11-point Likert scale (1 = very slight or not at all, and 11 = extremely). The positive affect scales (a = .91, .95, and .93 at baseline, after negative event recall, and after positive event recall, respectively) were created by averaging participants’ ratings of 10 positive emotions (e.g., interested, excited), whereas the negative affect scales (a = .72, .83, and .84 at baseline, after negative event recall, and after positive event recall, respectively) were created by averaging their ratings of 10 negative emotions (e.g., distressed, upset).

Baseline Happy Anger Sad

Temperament Measure The 77-item short form of Adult Temperament Questionnaire (ATQ; Evans & Rothbart, 2007) assessed temperament across four higher-order dimensions: effortful control, extraversion, negative emotionality, and orienting sensitivity. Participants rated each item on a 7-point Likert scale (1 = extremely untrue of you and 7 = extremely true of you). For the purpose of the current investigation, we only examined the scales that are part of the reactivity component of temperament (i.e., extraversion and negative emotionality). The 17-item extraversion factor (a = .78) includes subscales of: sociability (a = .79), high intensity pleasure (a = .62), and positive affect (a = .52). The 20item negative emotionality factor (a = .77) includes subscales of: fear (a = .60), sadness (a = .65), and frustration (a = .71).

baseline (M = 2.43, SE = .11), F(1, 87) = 18.90, p < .001, g2p = .18. Also as expected, after the positive event recall, participants reported higher levels of positive mood (M = 6.66, SE = .23) compared to baseline (M = 6.37, SE = .19), but the change from baseline was not statistically significant, F(1, 87) = 7.43, p = .75, g2p = .04. However, participants did report significantly higher levels of positive mood after positive event recall (M = 6.66, SE = .23), compared to after negative mood recall (M = 5.31, SE = .23), F(1, 87) = 75.80, p < .001, g2p = .47. In addition, participants reported significantly lower levels of negative mood (M = 1.89, SE = .10) after positive event recall, compared to baseline (M = 2.43, SE = .11), F(1, 87) = 24.68, p < .001, g2p = .22, and negative event recall (M = 3.11, SE = .16), F(1, 87) = 103.66, p < .001, g2p = .54. These results suggested that the event recall procedures were effective in manipulating the mood of the participants in the ways that we intended.

Results

Main Results

Preliminary Analyses

Multilevel modeling was used to account for clustering of data within individuals because it addresses the problems related to correlated errors that cannot be accounted for in other approaches like regression models (Garson, 2013). Data were organized in IBM SPSS 21 and analyzed in SSI’s student version of HLM 7.0. Six models with variance components covariance structure were performed to examine affect-based predictors of perception accuracy and intensity of happy, sad, and angry facial expressions at baseline (intercept) and across time (slope). The significance levels were two-tailed. In each model, the three assessment time points and self-reported mood were nested within individuals and were tested as the fixed-effect predictor at level 1. Participants’ self-reported temperament was included as the fixed-effect predictor at level-2. The model estimation method was maximum likelihood.

Participants were generally very accurate in identifying the expressions (see Table 1). Two repeated-measures ANOVAs were conducted to assess the effectiveness of positive and negative mood induction. The main effect of mood change was significant for self-reported positive mood, F(2, 174) = 40.17, p < .001, g2p = .32, and negative mood, F(2, 174) = 44.36, p < .001, g2p = .34, indicating that the mood ratings differed across the mood induction procedures. Contrast analyses revealed that participants reported significantly lower levels of positive mood after negative event recall (M = 5.31, SE = .23), compared to baseline (M = 6.37, SE = .19), F(1, 87) = 44.08, p < .001, g2p = .34. They also reported higher levels of negative mood after negative event recall (M = 3.11, SE = .16), relative to

2016 Hogrefe Publishing

Journal of Individual Differences 2016; Vol. 37(1):16–23

C. Y. Yi et al.: Adult Temperament and Emotion Perception

The mixed model equation is shown as follows (italicized variables were centered around grand mean):

Table 2. Mood and temperament predicting perception accuracy and intensity of happiness over time Happiness accuracy

Perception task outcometi ¼ b00 þ b01 Temperamenti þ b10 Timeti þ b11 Temperamenti Timeti

Fixed effects

þ b20 Mood ti þ b21 Mood i Temperamentti þ r0i þ r1i Timeti þ eti : ð1Þ

Coefficient

Intercept (b00) Extraversion (b01) Time slopes (p1) Intercept (b10) Extraversion (b11) Mood slopes (p2) Intercept (b20) Extraversion (b21)

However, the model for the perception accuracy of happiness was run without the mood and extraversion interaction terms with time, given that so few people missed identifying happy expressions at each of the three time points (i.e., Happiness accuracy = b00 + b01 · Extraversioni + b10 · Timeti + b20 · Positive Affectti + r0i + eti).

Happiness intensity SE

Coefficient

.99*** .01*

.01 .004

3.87*** 0.03

.05 .08

.002 –

.004 –

0.14*** 0.04

.03 .03

.003 –

.001 –

0.02 0.02

.02 .02

Note. *p < .05. ***p < .001 (two-tailed).

Perception of Happiness

accuracy of anger (ICC = .001, p > .05) and 41% of variance in perceived intensity of anger (ICC = .41, p < .001) were associated with interindividual differences among the participants. Partially consistent with our hypothesis, higher levels of self-reported negative mood significantly predicted higher accuracy in identifying anger. However, negative emotionality was not associated with perception accuracy for angry stimuli. The final model for anger accuracy provided a significantly better fit than the null model, 2LL = 345.45, Dv2(7) = 53.41, p < .001. With regard to anger intensity, individuals with higher levels of negative mood perceived anger as more intense, but negative emotionality did not significantly predict variations in perceived intensity of anger (see Table 3). The final model for anger intensity provided a better fit than the null model, although the reduction in deviance was only marginally significant, 2LL = 507.11, Dv2(7) = 13.86, p = .05. No interactive effect of mood and temperament was found on perception accuracy and perceived intensity of anger.

Intra-class correlations indicated that 0.4% of variance in the score of perception accuracy of happiness (ICC = .004, p > .50) and 54% of variance in perceived intensity of happiness (ICC = .54, p < .001) were associated with interindividual differences among the participants. Results from the multilevel models indicated that extraversion, but not positive mood, was significantly and positively associated with perception accuracy of happiness. On the other hand, neither positive mood nor temperament (extraversion) independently predicted perceived intensity of happiness. In addition, the effect of positive mood on perceived intensity of happiness did not vary as a function of extraversion (see Table 2). Overall, the results of model comparisons suggested that the final model for happy accuracy provided a better fit than the null model, although the reduction in deviance was nonsignificant, 2LL = 803.07, Dv2(3) = 6.69, p = .08. The final model for happy intensity reflected significantly better fit than the null model, 2LL = 307.46, Dv2(7) = 34.51, p < .001.

Perception of Sadness Perception of Anger

For sad stimuli, intra-class correlations indicated that 22% of variance in the score of perception accuracy (ICC = .22, p < .001) and 50% of variance in perceived intensity (ICC = .50, p < .001) were associated with interindividual

With regard to angry stimuli, intra-class correlations indicated that 0.10% of variance in the score of perception

Table 3. Mood and temperament predicting perception accuracy and intensity of anger and sadness over time Anger accuracy Fixed effects Intercept (b00) Negative emotionality (b01) Time slopes (p1) Intercept (b10) Negative emotionality (b11) Mood slopes (p2) Intercept (b20) Negative emotionality (b21)

Anger intensity

Sadness accuracy

Sadness intensity

Coefficient

.89*** .03

.02 .02

3.44*** 0.09

.07 .09

.66*** .05

.03 .04

2.78*** 0.17*

.06 .08

.05*** .01

.01 .02

0.03 0.02

.03 .04

.11*** .05*

.02 .02

0.07 0.09

.04 .05

.01* .01

.01 .01

0.10** 0.04

.03 .05

.03* .002

.01 .01

0.07* 0.02

.03 .04

Note. *p < .05. **p < .01. ***p < .001 (two-tailed). Journal of Individual Differences 2016; Vol. 37(1):16–23

2016 Hogrefe Publishing

C. Y. Yi et al.: Adult Temperament and Emotion Perception

differences among the participants. Our results from the multilevel models (see Table 3) provided further evidence for the mood-congruent effect, with higher levels of negative mood predicting higher scores for identifying sadness. In contrast to our hypothesis, negative emotionality was unrelated to identification accuracy of sadness. The final model for sadness accuracy reflected a significantly better fit than the null model, 2LL = 28.35, Dv2(7) = 45.78, p < .001. With regard to perceived intensity of sadness, higher levels of negative mood predicted lower levels of perceived intensity of sadness, whereas higher levels of negative emotionality predicted higher levels of perceived intensity of sadness. The final model for sadness intensity provided a significantly better fit than the null model, 2LL = 496.48, Dv2(7) = 27.15, p < .001. No interactive effects of mood and temperament were found on perception accuracy and perceived intensity of sadness.

Discussion The current study enhances the understanding of affectrelated individual differences in perception accuracy and perceived intensity of prototypical facial expressions of emotion. Partly consistent with our hypotheses, we found support for the affect-congruity effect of state-linked mood as well as that of trait-linked temperament.

Mood and Temperamental Effects on Emotion Perception Regarding happy expressions, higher levels of extraversion were significantly associated with higher accuracy of perceiving happiness. This finding suggests that extraverted people may be more attentive to happy expressions. It also coincides with research suggesting that people higher on extraversion may be more reactive to positive emotional displays, as exhibited by greater amygdala reactivity to expressions of happiness, than do people lower on extraversion (Canli, 2004). However, it is important to note that this finding should be interpreted with caution because of the ceiling effects in that it is based on only a few people who were unable to identify the happy expressions. Contrary to our expectations, positive mood did not significantly predict higher perception accuracy. In terms of perceived intensity of happiness, both extraversion and positive mood were unrelated to perceived intensity. One possible explanation for our null findings is that the presentation time (800 ms) of each face stimulus was too long, resulting in a ceiling effect. Past research indicated that healthy participants were able to correctly identify facial expressions of emotions presented for duration as brief as 200 ms (Surguladze et al., 2004). In addition, facial features indicative of happiness are generally more salient than those indicative of nonhappy expression, such as fear (Calvo & Nummenmaa, 2011). Salient facial features, such as an upturned mouth (as in a happy expression), could 2016 Hogrefe Publishing

positively influence perceived intensity of happiness (Messinger, Cassel, Acosta, Ambadar, & Cohn, 2008). The salient features of these expressions may have limited the variability in perceived intensity of happiness, reducing the likelihood of detecting the hypothesized effects. Future research could use stimuli that include less prototypical displays of emotion or that involve subtle shifts in expressions, allowing for the detection of slight individual differences. For negatively-valenced emotions, we provided further support for past literature regarding the effect of state-level mood on the identification of negative emotions (Lee et al., 2008; Schmid & Mast, 2010). Specifically, individuals who reported more negative mood were more accurate in identifying angry and sad faces. These results suggested that transient mood states may heighten individual’s sensitivity to mood-congruent facial cues, and thus, may help people correctly identify emotions from prototypical facial expressions. With regard to perceived intensity, individuals with higher levels of negative mood perceived anger as more intense. However, they perceived sadness as less intense. Our findings were consistent with a previous study with depressed patients that suggested individuals with an extreme level of negative mood viewed angry expressions as portraying greater intensity than sad expressions (Yoon et al., 2009). One explanation is that anger may be seen as more arousing than sadness (Russell & Bullock, 1985), and the arousal level of the emotional expressions may then influence people’s judgment on perceived intensity of the emotions (Yoon et al., 2009). Because we found evidence that the mood-congruity effects on perceived intensity of facial expressions may be emotion-specific, future research should use more specific mood states to predict perceived intensity of a wider range of facial expressions of emotions. Partially consistent with our hypotheses, we found some support for the affect-congruity effect of negative emotionality on perceived intensity of sadness, suggesting that people who experience more intense or frequent negative emotions in their lives also tended to see others’ sadness as more intense. This finding is consistent with research indicating that people with strong experiences of a certain emotion may think that the emotion should naturally look highly intense (Buchanan et al., 2010). It is worth noting that self-reported negative mood and temperamental negative emotionality had the opposite effect on perceived intensity of sadness in the current study in that negative emotionality predicted greater intensity, but negative mood predicted lower perceived intensity. This discrepancy in findings suggests that state-linked negative mood and trait-linked negative emotionality could influence perceived intensity of emotions through different mechanisms, at least when examined in the same study. Given the limited research attempting to tease apart state- and trait-level affect, the exact reason underlying our pattern of results is not clear. However, further investigation is warranted.

Limitations and Conclusions One major limitation of the present study is that the use of prototypical facial stimuli may reduce the variability in Journal of Individual Differences 2016; Vol. 37(1):16–23

C. Y. Yi et al.: Adult Temperament and Emotion Perception

perception accuracy and perceived intensity of the three emotions because the face stimuli were originally designed to make each expression as discernible as possible (Ekman & Friesen, 1976). In addition, there were ceiling effects in perception accuracy, which tended to be higher than what were found in other studies (e.g., Lee et al., 2008; Schmid & Mast, 2010). In addition to a longer presentation time of the face stimuli in our study, another contributing factor may be our use of a reduced set of emotions and face stimuli in the tasks. It is plausible that the participants used an elimination strategy to come to their answer during the perception accuracy task. Future research should use more sensitive stimuli, such as morphed faces that blend two basic emotions. Several other limitations are worth discussing. First, the order of our mood induction procedures was the standard protocol of a laboratory study and therefore, we could not counterbalance the order of these procedures. This ordering effect may have impacted the observed results in that the negative mood induction might limit the effectiveness of the positive mood induction, or could make the positive mood induction results less generalizable. However, in the protocol, there was a 2-minute recovery period after the negative event recall designed to allow participants’ mood to return to baseline. This resting period could therefore potentially minimize the ordering effect. Second, some of the subscales of our temperament measure and the negative affect scale of the PANAS had low reliability. Third, research indicates that people ascertain many emotional cues from other sources of information, such as their posture and context (e.g., Aviezer, Trope, & Todorov, 2012; Mondloch, 2012), however we only used emotional facial stimuli. Fourth, our sample mostly consisted of Caucasian adults, which limits the generalizability of our findings to other populations. Fifth, we did not videotape participants or use eye-tracking so we cannot offer data on how much participants actually attended to each face. Finally, other important unmeasured variables may have affected emotion perception, such as participants’ life experiences (e.g., experience of physical abuse or neglect; Pollak, Cicchetti, Hornung, & Reed, 2000; Pollak & Sinha, 2002). Despite these limitations, our study offers new evidence about how mood and the reactive components of temperament may influence an individual’s perception accuracy and perceived intensity of facial expressions of emotions. Overall, our study indicates that both state- and trait-level affect may offer unique information about individual differences in perceptual processing of emotional facial expressions. Efforts should be made to replicate and expand our findings with improved methodologies. Acknowledgments The authors would like to thank the participating students who donated their time to help with our research. The authors also wish to thank Dr. Dustin Long for statistical advice supported by the National Institute of General Medical Sciences of the National Institutes of Health under Award Number U54GM104942.

Journal of Individual Differences 2016; Vol. 37(1):16–23

References Aviezer, H., Trope, Y., & Todorov, A. (2012). Body cues, not facial expressions, discriminate between intense positive and negative emotions. Science, 338, 1225–1229. Bouhuys, A. L., Bloem, G. M., & Groothuis, T. G. (1995). Induction of depressed and elated mood by music influences the perception of facial emotional expressions in healthy subjects. Journal of Affective Disorders, 33, 215–226. Bower, G. H. (1981). Mood and memory. The American Psychologist, 36, 129–148. Buchanan, T. W., Bibas, D., & Adolphs, R. (2010). Associations between feeling and judging the emotions of happiness and fear: Findings from a large-scale field experiment. PLoS One, 5, e10640. Calvo, M. G., & Nummenmaa, L. (2011). Time course of discrimination between emotional facial expressions: The role of visual saliency. Vision Research, 51, 1751–1759. Canli, T. (2004). Functional brain mapping of extraversion and individual differences in emotion processing. Journal of Personality, 72, 1105–1132. Cavanagh, J., & Geisler, M. W. (2006). Mood effects on the ERP processing of emotional intensity in faces: A P3 investigation with depressed students. International Journal of Psychophysiology, 60, 27–33. Cooley, E. L., & Nowicki, S. (1989). Discrimination of facial expressions of emotion by depressed subjects. Genetic, Social, and General Psychology Monographs, 115, 449–465. Egger, H., Pine, D. S., Nelson, E., Leibenluft, E., Ernst, M., Towbin, K. E., & Angold, A. (2011). The NIMH Child Emotional Faces Picture Set (NIMH-ChEFS): A new set of children’s facial emotion stimuli. International Journal of Methods in Psychiatric Research, 20, 145–156. Ekman, P. (1993). Facial expression and emotion. The American Psychologist. 48, 384–392. Ekman, P., & Friesen, W. V. (1976). Pictures of facial affect. Palo Alto, CA: Consulting Psychologist Press. Evans, D. E., & Rothbart, M. K. (2007). Development of a model for adult temperament. Journal of Research in Personality, 41, 868–888. Garson, G. D. (2013). Fundamentals of hierarchical linear and multilevel modeling. In G. D. Garson (Ed.), Hierarchical Liner Modeling (pp. 3–25). Thousand Oaks, CA: Sage. Hall, J. A., Andrzejewski, S. A., & Yopchick, J. E. (2009). Psychosocial correlates of interpersonal sensitivity: A metaanalysis. Journal of Nonverbal Behavior, 33, 149–180. doi: 10.1007/s10919-009-0070-5 Hess, U., Blairy, S., & Kleck, R. E. (1997). The intensity of emotional facial expressions and decoding accuracy. Journal of Nonverbal Behavior, 21, 241–257. Hoffmann, H., Kessler, H., Eppel, T., Rukavina, S., & Traue, H. C. (2010). Expression intensity, gender and facial emotion recognition: Women recognize only subtle facial emotions better than men. Acta Psychologica, 135, 278–283. doi: 10.1016/j.actpsy.2010.07.012 Irtel, H. (2007). PXLab: The Psychological Experiments Laboratory [online] Version 2.1.11. Mannheim, Germany: University of Mannheim. Retrieved from http://www.pxlab.de Isaacowitz, D. M., Löckenhoff, C. E., Lane, R. D., Wright, R., Sechrest, L., Riedel, R., & Costa, P. T. (2007). Age differences in recognition of emotion in lexical stimuli and facial expressions. Psychology and Aging, 22, 147–159. doi: 10.1037/0882-7974.22.1.147 Lee, T. C., Ng, E. H., Tang, S. W., & Chan, C. H. (2008). Effects of sad mood on facial emotion recognition in Chinese people. Psychiatry Research, 159, 37–43.

2016 Hogrefe Publishing

C. Y. Yi et al.: Adult Temperament and Emotion Perception

Lembke, A., & Ketter, T. A. (2002). Impaired recognition of facial emotion in mania. The American Journal of Psychiatry, 159, 302–304. Lennox, B. R., Jacob, R. R., Calder, A. J., Lupson, V. V., & Bullmore, E. T. (2004). Behavioural and neurocognitive responses to sad facial affect are attenuated in patients with mania. Psychological Medicine: A Journal of Research in Psychiatry and the Allied Sciences, 34, 795–802. Matsumoto, D., Kasri, F., & Kooken, K. (1999). AmericanJapanese cultural differences in judgements of expression intensity and subjective experience. Cognition and Emotion, 13, 201–218. doi: 10.1080/026999399379339 McClure, E. B. (2000). A meta-analytic review of sex differences in facial expression processing and their development in infants, children, and adolescents. Psychological Bulletin, 126, 424–453. doi: 10.1037/0033-2909.126.3.424 Messinger, D. S., Cassel, T. D., Acosta, S. I., Ambadar, Z., & Cohn, J. F. (2008). Infant smiling dynamics and perceived positive emotion. Journal of Nonverbal Behavior, 32, 133–155. Milanak, M. E., & Berenbaum, H. (2014). The effects of context on facial affect recognition. Motivation and Emotion, 38, 560–568. doi: 10.1007/s11031-014-9401-x Mondloch, C. J. (2012). Sad or fearful? The influence of body posture on adults’ and children’s perception of facial displays of emotion. Journal of Experimental Child Psychology, 111, 180–196. Pollak, S. D., Cicchetti, D., Hornung, K., & Reed, A. (2000). Recognizing emotion in faces: Developmental effects of child abuse and neglect. Developmental Psychology, 36, 679–688. Pollak, S. D., & Sinha, P. (2002). Effects of early experience on children’s recognition of facial displays of emotion. Developmental Psychology, 38, 784–791. Rothbart, M. K., & Derryberry, D. (1981). Development of individual differences in temperament. In M. E. Lamb & A. L. Brown (Eds.), Advances in developmental psychology (Vol. 1, pp. 37–86). Hillsdale, NJ: Erlbaum. Ruffman, T., Henry, J. D., Livingstone, V., & Phillips, L. H. (2008). A meta-analytic review of emotion recognition and aging: Implications for neuropsychological models of aging. Neuroscience and Biobehavioral Reviews, 32, 863–881. Russell, J. A., & Bullock, M. (1985). Multidimensional scaling of emotional facial expressions similarity from preschoolers to adults. Journal of Personality and Social Psychology, 48, 1290–1298. Schiffenbauer, A. (1974). Effect of observer’s emotional state on judgments of the emotional state of others. Journal of Personality and Social Psychology, 30, 31–35. Schmid, P., & Mast, M. (2010). Mood effects on emotion recognition. Motivation and Emotion, 34, 288–292. Schwarz, N. (1990). Feelings as information: Informational and motivational functions of affective states. In E. Higgins, R. M. Sorrentino, E. Higgins, & R. M. Sorrentino (Eds.),

2016 Hogrefe Publishing

Handbook of motivation and cognition: Foundations of social behavior (Vol. 2, pp. 527–561). New York, NY: Guilford Press. Schwarz, N., & Clore, G. L. (1983). Mood, misattribution, and judgments of well-being: Informative and directive functions of affective states. Journal of Personality and Social Psychology, 45, 513–523. Surguladze, S. A., Young, A. W., Senior, C., Brébion, G., Travis, M. J., & Phillips, M. L. (2004). Recognition accuracy and response bias to happy and sad facial expressions in patients with major depression. Neuropsychology, 18, 212–218. Trentacosta, C. J., & Fine, S. E. (2010). Emotion knowledge, social competence, and behavior problems in childhood and adolescence: A meta-analytic review. Social Development, 19, 1–29. doi: 10.1111/j.1467-9507.2009.00543.x Tsai, J. L., Levenson, R. W., & McCoy, K. (2006). Cultural and temperamental variation in emotional response. Emotion, 6, 484–497. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54, 1063–1070. Widen, S. C., & Russell, J. A. (2003). A closer look at preschoolers’ freely produced labels for facial expressions. Developmental Psychology, 39, 114–128. doi: 10.1037/ 0012-1649.39.1.114 Yoon, K., Joormann, J., & Gotlib, I. H. (2009). Judging the intensity of facial expressions of emotion: Depressionrelated biases in the processing of positive affect. Journal of Abnormal Psychology, 118, 223–228. doi: 10.1037/ a0014658 Young, M. E., & Brunet, P. M. (2011). It’s about time: The influence of sociability – but not shyness – on the categorization of facial expressions of emotion. Journal of Individual Differences, 32, 65–73.

Date of acceptance: April 14, 2015 Published online: February 29, 2016

Chit Yuen Yi Department of Psychology West Virginia University 53 Campus Drive Morgantown, WV 26506 USA Tel. +1 734 219-8383 E-mail cyi@mix.wvu.edu

Journal of Individual Differences 2016; Vol. 37(1):16–23

Original Article

The Relationships Between the Dark Triad, the Moral Judgment Level, and the Students’ Disciplinary Choice Self-Selection, Indoctrination, or Both? Annika Krick,1 Stephanie Tresp,1 Mirijam Vatter,1 Antonia Ludwig,1 Michael Wihlenda,2 and Martin Rettenberger1,3 1

Department of Psychology, Johannes Gutenberg-University Mainz (JGU), Germany, 2Global Ethic Institute, University of Tübingen, Germany, 3Centre for Criminology, Wiesbaden, Germany

Abstract. The purpose of the present study was to examine the relationships between the personality traits of the Dark Triad, the moral judgment level, and the students’disciplinary choice. It was hypothesized that students who major in higher business and management education show higher levels of the Dark Triad and lower levels of moral judgment competence (self-selection hypothesis). According to the indoctrination hypothesis it was assumed that the differences between business and management students and other students would be higher in advanced semesters. The findings suggest that business and management students show higher levels of the Dark Triad but not of moral judgment competence. However, there was no evidence found for a higher difference in advanced students. Keywords: Dark Triad, psychopathy, narcissism, Machiavellianism, business, management, morality

The financial crisis and new dimensions of white-collar crimes and criminals like the Enron scandal and the case of Bernard Madoff have caused a growing interest in identifying relevant characteristics and personality traits that promote problematic decision-making processes of employees and problematic workplace behavior. For a long time researchers focused on the well-known construct of psychopathy but recent findings suggest that more comprehensive answers seem to be found in the concept of the Dark Triad (Paulhus & Williams, 2002). The Dark Triad consists of three socially undesirable characteristics: narcissism, Machiavellianism, and psychopathy. Narcissists are characterized by an unwarranted feeling of grandiosity, an extraordinary sense of entitlement, and a lack of empathy (Mathieu & St-Jean, 2013). They constantly try to manipulate others to their own benefit. Machiavellian behavior manifests itself in contempt of conventional morality, a lack of empathy, as well as in dishonest behavior in an effort to maintain positions of power (Jonason, Slomski, & Partyka, 2012). A lack of empathy and remorse, impulsive and egocentric behavior, irresponsibility, manipulation and deception, and antisocial behaviors define the construct of psychopathy (Hare, 1999). At the same time, researchers and clinicians have claimed that people with specific undesirable characteristics related Journal of Individual Differences 2016; Vol. 37(1):24–30 DOI: 10.1027/1614-0001/a000184

to the Dark Triad achieve higher positions more often than others (Babiak & Hare, 2006; Board & Fritzon, 2005). Consequently, there have to be some supportive job-related aspects of these traits: For example, already in 1941 the US-American psychiatrist Hervey M. Cleckley described the phenomenon of the successful psychopath in the workplace which is also supported by newer findings (Lykken, 1995). Recent studies suggest that the number of psychopaths in business and management positions is significantly higher than in other occupational fields (Babiak & Hare, 2006; Boddy, 2011). Jonason et al. (2012) examined different leadership skills used by employees with high levels of the Dark Triad and found that high levels on these traits lead to hard, aggressive, and forceful tactics of social influence. To summarize these results, it seems that employees holding such traits do not respect or consider any boundaries in order to achieve their aims and could therefore be relatively successful in some occupational areas. Some specific Dark Triad characteristics like recklessness and a lack of empathy and remorse lead to the idea that a person’s moral judgment competence should also be taken into consideration when investigating unethical and deceitful workplace behaviors. Lind (2008) introduced the term of the dual-aspect model of moral judgment competence which is based on Kohlberg’s (1964) definition that moral Ó 2016 Hogrefe Publishing

A. Krick et al.: Dark Triad and Student Disciplinary Choice

judgment competence is a person’s capability to decide and behave according to moral ideals. His theory describes that cognitive and affective mechanisms of moral judgment competence cannot be separated, although they are distinguishable from each other (Lind, 2008). Both components of the dual-aspect model of moral judgment competence refer to an individual’s attitudes, ideals, and orientations: While the affective component captures the affection for moral principles and ideals, the cognitive aspect of the model describes the ability to reason and act according to these ideals and principles (Lind, 1985). An adequate measurement of moral competence should assess both the affective and the cognitive mechanisms and should be sensitive to positive changes as a function of moral learning and intervention or to downward changes as a function of competence erosion (Lind, 2002). Up to now, a number of studies have examined the relationship between Dark Triad personality traits like psychopathy or narcissism and moral judgment capabilities (Amernic & Craig, 2010; Cleckley, 1941; Galperin, Bennett, & Aquino, 2010; Hare, 1999). However, data show inconsistent results concerning psychopaths’ moral judgment skills in moral dilemmas (e.g., Bartels & Pizarro, 2011; Cima, Tonnaer, & Hauser, 2010). Before someone is able to pursue a successful career in the economic occupational area, the person usually has to go through the business and management education system. Therefore, there has been growing criticism concerning the academic business and management education system because it has ‘‘obviously failed to instill in their students the ethical values and norms that would have helped them to conduct their professional activity with a due sense of responsibility’’ (Elegido, 2009, p. 16). Wilson and McCarthy (2011) analyzed the relationship between university students’ major and psychopathy and found that students majoring in economics had higher levels of psychopathy than students from other areas. The authors conclude that economics might attract students with higher levels of psychopathy because of the possibility of achieving powerful positions in the future (the so-called selfselection hypothesis). An alternative explanation is provided by proponents of the so-called indoctrination hypothesis which proposes that unethical, selfish, and immoral behaviors are more prevalent among business and management students because of the academic training received at the university (Elegido, 2009). For example, Davis (1987) reported that nonbusiness students experienced a significant moral growth during their years in college, while business students did not become morally more competent or even regressed. The aim of the present study was to assess the relationships between the students’ major subject, the amount of studying experience they had (i.e., novice vs. advanced student), Dark Triad traits, and their moral judgment competence. Concerning previous studies about the association between certain socially undesirable characteristics and success in higher business and management positions

Ó 2016 Hogrefe Publishing

(e.g., Babiak & Hare, 2006; Board & Fritzon, 2005; Wilson & McCarthy, 2011), it was hypothesized that business and management students show higher scores of the Dark Triad than students from other subject areas (self-selection hypothesis). Given previous investigations about the possible negative effect of business education on moral development and character formation (Davis, 1987; Elegido, 2009; Frank, Gilovich, & Regan, 1993), it was further hypothesized that the level of the Dark Triad traits will be higher and simultaneously the moral judgment competence will be lower the more advanced the students are (indoctrination hypothesis).

Method The data of the present sample were collected by means of an online survey, which was sent to various e-mail addresses of students’ faculty councils from German universities. Furthermore, the link was shared via social networks (e.g., Facebook). Two hundred seventy-six students (71.4% female) participated in this study, and the mean age was 23.59 years (SD = 4.37). The majority of the students (76.4%, n = 211) attended state universities, 19.6% (n = 54) were studying at a private university. All participants between the first and the sixth semester were categorized as beginners (57.2%, n = 158), and participants with more than six semesters at the time of survey participation were labeled as advanced students (n = 118). Of the total sample 28.3% (n = 78) studied business and management (50% female) and 71.7% (n = 198) majored in other subject areas (79.8% female). The majority of the nonbusiness and management students group indicated that their major subject area falls in the category of social sciences (54.0%, n = 149), 6.5% (n = 18) specified a subject area from the field of humanities, 6.5% (n = 18) from engineering, and 4.7% (n = 13) from the natural sciences. Of the business and management subsample 57.7% (n = 45) were categorized as beginners and 42.3% (n = 33) as advanced students. Male and female participants did not differ significantly in their moral judgment competence measured by the C-score derived from the Moral Judgment Test (see below for a detailed description of the measurement method), t(273) = .427, p = .670. However, men had generally higher scores on the Dark Triad traits measure compared to women, t(274) = 5.069, p < .001. On the first page of the online survey, participants were informed that the present survey was about the assessment of ‘‘Differences in the Personality of Students in Different Subject Areas.’’ All participants were assured of the anonymity of their data. The complete survey consisted of some single items about basic demographical characteristics of the participants and the German versions of the following standardized questionnaires: the Balanced Inventory of Desirable Responding (BIDR; Musch, Brockhaus, & Bröder, 2002; Paulhus, 1998), the Dirty Dozen (Jonason

Journal of Individual Differences 2016; Vol. 37(1):24–30

A. Krick et al.: Dark Triad and Student Disciplinary Choice

Table 1. Means (M), standard deviations (SD), and internal consistencies (a) of all scales for the total sample (N = 276) and separately for the male (n = 79) and female (n = 197) subsamples Total sample (N = 276) M (SD) Dirty Dozen total score Psychopathy Machiavellianism Narcissism MJT C-score BIDR total score BIDR self deception BIDR impression management

44.52 13.81 12.26 18.46 28.75 80.22 43.52 36.70

(15.85) (7.28) (5.93) (6.86) (18.32) (13.58) (8.35) (10.13)

a .83 .60 .81 .76 —a .67 .67 .68

Men (n = 79) M (SD) 51.84 15.35 16.56 19.92 28.00 79.14 45.73 33.41

(16.42) (6.32) (8.14) (6.62) (19.67) (13.85) (9.35) (10.62)

Women (n = 197) a .81 .52 .83 .72 —a .67 .75 .70

M (SD) 41.59 11.02 12.71 17.87 29.04 80.65 42.63 38.02

(14.66) (5.29) (6.62) (6.69) (17.80) (13.47) (7.76) (9.64)

a .82 .57 .77 .77 —a .68 .62 .65

Notes. MJT = Moral Judgment Test; BIDR = Balanced Inventory of Desirable Responding. aNo a-value was calculated because this index is statistically inadequate for the MJT.

& Webster, 2010; Küfner, Dufner, & Back, 2015), and the Moral Judgment Test (MJT; Lind, 2002, 2008). The original BIDR (Paulhus, 1998) consisted of 40 items which had to be answered using a 7-point Likert scale (from 1 = not true to 7 = very true) and which can be allocated to two factors: self-deception (sample item: ‘‘My first impressions of people usually turn out to be right.’’) and impression management (sample item: ‘‘I never cover up my mistakes.’’). The German version of the BIDR (Musch et al., 2002) uses an empirically reduced set of 20 items (same scale anchors) with 10 items on each subscale. In three independent cross-validation studies, the German version of the BIDR showed satisfactory psychometric qualities, the proposed two-factorial structure could be replicated, and good convergent and discriminant validity were found (Musch et al., 2002). The Dirty Dozen is a concise 12-item measure of the Dark Triad (Jonason & Webster, 2010), that asks for the participants agreement (from 1 = not at all to 5 = very much) with statements such as ‘‘I tend to lack remorse’’ (psychopathy), ‘‘I have used deceit or lied to get my way’’ (Machiavellianism), or ‘‘I tend to want others to pay attention to me’’ (narcissism). Both the original English (Jonason & Webster, 2010) as well as the German translation (Küfner et al., 2015) showed good psychometric properties.1 Within the MJT participants have to make comprehensive judgments about the well-known mercykilling dilemma and the worker’s dilemma (Kohlberg, 1964). The participants are asked to judge arguments which present different levels of moral reasoning (Lind, 2008). For each dilemma, participants have to judge 12 arguments, six supporting the decision that the protagonist in the story made and six arguing against his or her decision. The MJT measures the affective and cognitive aspects of moral judgment competence by using the so-called C-score, which can range from 1 to 100. The C-score measures the degree to which the participant let his or her judgment behavior be determined by moral concerns or principles 1

rather than by other psychological forces (e.g., like the human tendency to make arguments agree with one’s opinion or decision about a certain issue). In other words, the C-score reflects a person’s ability to judge arguments according to their moral quality. The higher the C-scores, the more morally competent an individual is regarded to be (Lind, 2008). All statistical analyses presented in the following result section were calculated by using IBM Statistical Package for Social Sciences (SPSS) version 21.0.0.1.

Results The internal consistency of the Dirty Dozen was Cronbach’s a = .83 which is comparable to the results of the development study (Jonason & Webster, 2010). The German version of the BDIR also yielded comparable results to previous investigations (a = .67; Musch et al., 2002). Table 1 shows means, standard deviations, and the internal consistencies of all scales for the total sample as well as separately for the male and female subsamples. For the MJT no a-value was calculated because this index is usually considered as inadequate for assessing the structural properties of the MJT (Lind, 2008). The correlation between the Dirty Dozen and the self-deception subscale of the BIDR was not significant (r = .06, p = .33), whereas the correlation between the BIDR impression management subscale and the Dirty Dozen scale was r = .57 ( p < .001). The MJT showed no significant relationship with both selfdeception (r = .00, p = .96) and impression management (r = .04, p = .54). The correlations between the BIDR and both independent variables, the students’ major subject and the amount of studying experience, were not significant (r = .08, p = .19 and r = .03, p = .60, respectively). A multivariate analysis of covariance (MANCOVA) was conducted to examine the differences in Dark Triad traits

We decided in favor of the Dirty Dozen as measure for the Dark Triad because of the fact that this 12-item instrument was internationally established (Jonason & Webster, 2010; Jonason et al., 2012) and adequately translated into German. Furthermore, recently published cross-validation data of the German version supported the use of the translated instrument (Küfner et al., 2015). Even if recently published studies sounded a note of caution regarding the use of the Dirty Dozen, the instrument was particularly regarded as a reasonable choice if one is interested in obtaining a composite score of the Dark Triad (Lee et al., 2013).

Journal of Individual Differences 2016; Vol. 37(1):24–30

Ó 2016 Hogrefe Publishing

A. Krick et al.: Dark Triad and Student Disciplinary Choice

and moral competence levels between business and management students and students majoring in other subject areas (self-selection hypothesis) and between beginners and advanced students (indoctrination hypothesis). The students’ major subject and the amount of studying experience (i.e., novices vs. advanced students) were the independent variables and the Dirty Dozen total score and the MJT C-score were the dependent variables. The BIDR scores were included as a covariate, in order to control for social desirability. Furthermore, gender was included as a second covariate because previous research indicated that women have usually lower scores on Dark Triad traits than men (e.g., Paulhus & Williams, 2002). The MANCOVA examining the differences in Dark Triad traits and the moral competence level between business and management students and students majoring in other subject areas was significant (Wilks K = .95; F(1, 271) = 7.282, p < .001; g2p = .05). The results show that business and management students (M = 53.27, SD = 16.27) exhibit higher levels of Dark Triad traits than the control group (M = 41.08, SD = 14.31; F(1, 271) = 14.090, p < .001). However, concerning their moral judgment competence business and management students (M = 28.47, SD = 18.97) and students of other subject areas (M = 28.85, SD = 18.11) did not differ in their C-scores, F(1, 270) = .229, p = .633. The indoctrination hypothesis could not be confirmed: The MANCOVA (Wilks K = 1.00; F(1, 271) = .684, p = .506; g2p = .01) revealed no significant difference between beginners (M = 45.69, SD = 16.54) and advanced students (M = 42.96, SD = 14.79) with regard to their Dark Triad scores, F(1, 271) = .545, p = .461. Similarly, there was no significant difference between beginners (M = 29.35, SD = 18.52) and advanced students (M = 27.93, SD = 18.11) in their mean C scores, F(1, 270) = .738, p = .391.2 There was no significant interaction effect, neither for the Dark Triad scores, F(1, 271) = .116, p = .733, nor for the MJT C-scores, F(1, 270) = .077, p = .782, between the major subject and the amount of studying experience.

Discussion The aim of the present study was to examine the relationships between the Dark Triad, moral judgment competence, and the students’ major subject as well as their amount of studying experience. The results show that after controlling for social desirability business and management students score higher in Dark Triad traits than students of other subject areas but both student groups did not differ in their moral judgment competence. With regard to the Dark Triad, the self-selection hypothesis could be supported because previous studies have suggested that business and management students with comparatively higher levels of Dark Triad traits may be attracted by the opportunity of

potentially powerful positions in the business world (Elegido, 2009). The findings of the present study are in line with a previous study published by Wilson and McCarthy (2011) who pointed out that students majoring in commerce have higher psychopathy scores than students of other areas. As far as narcissism is concerned, the present results are also in accordance with those of previous studies which have investigated the role of narcissism in the entrepreneurial personality (Mathieu & St-Jean, 2013). Studies about the relevance of Machiavellianism in the economic and occupational system revealed that this personality trait is associated with abusive behaviors at the workplace (Kiazad, Restubog, Zagenczyk, Kiewitz, & Tang, 2010), a focus on power and manipulativeness (Kessler et al., 2010), and a reduced commitment toward colleagues and employers (Zettler, Friedrich, & Hilbig, 2011). Therefore, these constructs seem to be highly relevant for the investigation of problematic and socially undesirable personality traits in individuals working in the economic system (Jonason et al., 2012). Taken together, one can conclude that persons with higher levels of Dark Triad personality traits are more likely to choose university courses which allow them supposedly to achieve powerful and successful positions in the employment market. The results of the previous study also supported the seminal case studies published by Cleckley (1941) who asserted that certain psychopathic traits are conductive to managerial functions. A possible theoretical explanation for the results of the present study as well as for the above-mentioned previous investigations is provided by John Holland’s theory of vocational choice (Holland, 1997) which proposes that individuals with similar personality types tend to make personal and vocational decisions in a way to increase the probability that they can be around other persons who are like them. In other words, persons with higher levels of Dark Triad traits are looking for work environments that fit their personality traits. Because various studies have recently suggested that the amount of individuals with high Dark Triad traits seems to be higher in business and management positions (Babiak & Hare, 2006; Boddy, 2011), these occupational fields could be a preferable vocational choice (Holland, 1997). Another somewhat similar explanation is provided by the person-environment-fit theory which proposes a compatibility between individual and environmental (i.e., organizational and vocational) characteristics that occurs when one entity provides what the other needs and/or both parties share similar fundamental characteristics (Kristof, 1996). Similarly to Holland’s (1997) theory of vocational choice, the person-environment-fit theory would predict a higher degree of Dark Triad traits in business and management students than in students of other areas. Contrary to the self-selection hypothesis, the present study provided no support for the indoctrination hypothesis because beginners and advanced students exhibited no differences in the level of Dark Triad traits. The indoctrination

A reanalysis with the number of semesters as independent variable instead of the dichotomized variable (beginners vs. advanced students) did not change the result pattern. We would like to thank an anonymous reviewer for this suggestion.

Ó 2016 Hogrefe Publishing

Journal of Individual Differences 2016; Vol. 37(1):24–30

A. Krick et al.: Dark Triad and Student Disciplinary Choice

hypothesis proposes that the degree of Dark Triad traits will increase during university education due to the academic training the students have received (Elegido, 2009). However, there was no evidence that beginners and advanced students of both study groups – business and management students and the control group majoring in other subjects – differed significantly. In other words, the course of studies does not function as a moderator, so one cannot conclude that course contents are associated with differences in Dark Triad characteristics in the present study. This result indicated that Board and Fritzon’s (2005) findings concerning the level of psychopathy of managers rather represent what existed before the course of studies and the occupational career than something which became boosted by the course of studies or the profession. Of course, the results of the present study might be due to cohort effects which is a possible explanation for the lack of difference between beginners and advanced students. For a methodologically sound examination of the indoctrination hypothesis, it would be necessary to use a prospective-longitudinal research design. Concerning the moral judgment competence neither the self-selection nor the indoctrination hypothesis could be confirmed. The assumptions about the relationships between psychopathy and moral judgment (e.g., Hare, 1999) and narcissism and moral judgment (e.g., Amernic & Craig, 2010) obviously cannot be simply transferred to the course of managerial economics and moral judgment. In other words, even if business and management students have higher levels of Dark Triad traits, they do not necessarily differ in their moral judgment competence. With regard to the relationship between psychopathy and decisionmaking in moral dilemmata previous research yielded somewhat contradictory results with some studies reporting that psychopathy leads to a utilitarian bias in moral judgment processes (e.g., Bartels & Pizarro, 2011), whereas other studies have shown that there is no negative influence of psychopathy on moral decision-making capacities (e.g., Cima et al., 2010). Recently, Tassy, Deruelle, Mancini, Leistedt, and Wicker (2013) provided a possible solution for these contradictions derived from previous research. Their research findings indicated that higher levels of psychopathic traits – particularly if associated with affective deficits – were related to a greater proportion of utilitarian responses only if the moral dilemma tasks included questions targeting a choice of hypothetical action (e.g., ‘‘Would you. . . in order to. . .?’’) instead of questions which focus on abstract judgments only (e.g., ‘‘Is it acceptable to. . . in order to. . .?’’). The authors concluded that high psychopathic traits influence the moral choice but not necessarily the moral judgment (Tassy et al., 2013). In relation to the results of the present study it could be hypothesized that there were no differences between the study groups because we focused on differences in moral judgment competence rather than in moral choice.

Besides the possible inappropriateness of the moral dilemma task there are further limitations which should be addressed. First, the total sample size and especially the sizes of the subsamples were relatively small.3 Second, online surveys can generally suffer from methodological problems which could limit the informative value of a scientific investigation (Blank, Fielding, & Lee, 2008). Disadvantages and potential problems which are often discussed for Internetbased surveys are, for example, sampling issues (e.g., Winters, Christoff, & Gorzalka, 2010) and mode effects because previous investigations indicated that online respondents might use scales differently compared to respondents in other data collection modes (e.g., it is discussed whether online respondents are more likely to choose midpoints in scales; Duffy, Smith, Terhanian, & Bremer, 2005). However, there is also scientific evidence available that Internet-based data collections are commonly able to produce generalizable results (e.g., Best, Krueger, Hubbard, & Smith, 2001). Third, the internal consistency of the BIDR was relatively low which could also be a potential source for biased results. Fourth, this study only used a cross-sectional research design. Therefore, cohort effects might provide one possible explanation for the lack of difference between beginners and advanced students. To test whether scores on Dark Triad traits increase and moral judgment competence actually decreases over the courses of studies as suggested by the indoctrination hypothesis, a prospective-longitudinal research design would be necessary. Fifth, there is still an ongoing debate about the benefits and costs of brief personality measures (e.g., Miller et al., 2012; Rammstedt & John, 2007; Rauthmann, 2013). Referring to the Dark Triad, Miller et al. (2012) reported that the validity of the Dirty Dozen showed flaws which were especially relevant for the psychopathy subscale. Therefore, it cannot be ruled out that the (psychometric) costs of using brief personality measures outweighed the benefits. Therefore, a replication of the present study either using other instruments for measuring the Dark Triad as a single variable or using individual measures for each dimension of the Dark Triad would be desirable. Despite these limitations, the present study contributes to the knowledge about the relationships between Dark Triad traits, moral judgment competence, and students’ disciplinary choice and provides possible starting points for future research studies. Acknowledgments We would like to thank Daniel Turner (Institute of Sex Research and Forensic Psychiatry, University Medical Center Hamburg-Eppendorf, Hamburg, Germany), Volker Lingnau and Johannes Wadle (Department of Management Accounting and Management Control Systems, Technical University Kaiserslautern, Germany) for their helpful

Furthermore, the sample of the present study consisted predominantly of female participants and previous reports indicated that women and men could differ in the degree of Dark Triad traits (Paulhus & Williams, 2002). However, we tried to countervail the potential sexrelated bias by controlling for gender in the data analyses.

Journal of Individual Differences 2016; Vol. 37(1):24–30

Ó 2016 Hogrefe Publishing

A. Krick et al.: Dark Triad and Student Disciplinary Choice

comments in the preparation of this article and their support in the data collection process. Annika Krick and Stephanie Tresp contributed equally to this work. The authors wish it to be known that, in their opinion, the first two authors should be regarded as joint first authors.

References Amernic, J. H., & Craig, R. J. (2010). Accounting as a facilitator of extreme narcissism. Journal of Business Ethics, 96, 79–93. doi: 10.1007/s10551-010-0450-0 Babiak, P., & Hare, R. D. (2006). Snakes in suits: When psychopaths go to work. New York, NY: HarperCollins. Bartels, D. M., & Pizarro, D. A. (2011). The mismeasure of morals: Antisocial personality traits predict utilitarian responses to moral dilemmas. Cognition, 121, 154–161. doi: 10.1016/j.cognition.2011.05.010 Best, S. J., Krueger, B., Hubbard, C., & Smith, A. (2001). An assessment of the generalizability of internet surveys. Social Science Computer Review, 19, 131–145. doi: 10.1177/ 089443930101900201 Blank, G., Fielding, N. G., & Lee, R. M. (2008). The SAGE handbook of online research methods. London, UK: Sage. Board, B. J., & Fritzon, K. (2005). Disordered personalities at work. Psychology, Crime, & Law, 11, 17–32. doi: 10.1080/ 10683160310001634304 Boddy, C. R. (2011). Corporate Psychopaths, bullying and unfair supervision in the workplace. Journal of Business Ethics, 100, 367–379. doi: 10.1007/s10551-010-0689-5 Cima, M., Tonnaer, F., & Hauser, M. D. (2010). Psychopaths know right from wrong but don’t care. Social Cognitive and Affective Neuroscience, 5, 59–67. doi: 10.1093/scan/nsp051 Cleckley, H. (1941). The Mask of Sanity (1st ed.). St. Louis, MO: C. V. Mosby. Davis, L. (1987). Moral judgment development of graduate management students in two cultures: Minnesota and Singapore, (Unpublished doctoral dissertation). University of Minnesota, Minneapolis. Duffy, B., Smith, K., Terhanian, G., & Bremer, J. (2005). Comparing data from online and face-to-face surveys. International Journal of Market Research, 47, 615–639. Elegido, J. (2009). Business education and erosion of character. African Journal of Business Ethics, 4, 16–24. Frank, R. H., Gilovich, T., & Regan, D. T. (1993). Does studying economics inhibit cooperation? Journal of Economic Perspectives, 7, 159–171. doi: 10.1257/jep.7.2.159 Galperin, B. L., Bennett, R. J., & Aquino, K. (2010). Status differentiation and the protean self: A social-cognitive model of unethical behavior in organizations. Journal of Business Ethics, 98, 407–424. doi: 10.1007/s10551-010-0556-4 Hare, R. D. (1999). Without conscience: The disturbing world of the psychopaths among us. New York, NY: The Guilford Press. Holland, J. L. (1997). Making vocational choices: A theory of vocational personalities and work environments. Odessa, FL: Psychological Assessment Resources. Jonason, P. K., Slomski, S., & Partyka, J. (2012). The Dark Triad at work: How toxic employees get their way. Personality and Individual Differences, 52, 449–453. doi: 10.1016/ j.paid.2011.11.008 Jonason, P. K., & Webster, G. D. (2010). The Dirty Dozen: A concise measure of the Dark Triad. Psychological Assessment, 22, 420–432. doi: 10.1037/a0019265 Kessler, S. R., Bandell, A. C., Spector, P. E., Borman, W. C., Nelson, C. E., & Penney, L. M. (2010). Re-examining Machiavelli: A three-dimensional model of Machiavellianism

Ó 2016 Hogrefe Publishing

in the workplace. Journal of Applied Social Psychology, 40, 1868–1896. doi: 10.1111/j.1559-1816.2010.00643.x Kiazad, K., Restubog, S. L., Zagenczyk, T. J., Kiewitz, C., & Tang, R. L. (2010). In pursuit of power: The role of authoritarian leadership in the relationship between supervisors’ Machiavellianism and subordinates’ perceptions of abusive supervisory behavior. Journal of Research on Personality, 44, 512–519. doi: 10.1016/j.jrp.2010.06.004 Kohlberg, L. (1964). Development of moral character and moral ideology. In M. Hoffman & L. W. Hoffman (Eds.), Review of child development research (pp. 381–432). New York, NY: Sage. Kristof, A. L. (1996). Person-organizational fit: An integrative review of its conceptualizations, measurement, and implications. Personnel Psychology, 49, 1–49. doi: 10.1111/ j.1744-6570.1996.tb01790.x Küfner, A. C. P., Dufner, M., & Back, M. (2015). Das Dreckige Dutzend und die Niederträchtigen Neun – Kurzskalen zur Erfassung von Narzissmus, Machiavellismus und Psychopathie [The dirty dozen and the naughty nine – short scales for the assessment of narcissism, machiavellianism, and psychopathy]. Diagnostica, 61, 76–91. doi: 10.1026/0012– 1924/a000124 Lee, K., Ashton, M. C., Wiltshire, J., Bourdage, J. S., Visser, B. A., & Gallucci, A. (2013). Sex, power and money: Prediction from the Dark Triad and honesty-humility. European Journal of Personality, 27, 169–184. doi: 10.1002/per.1860 Lind, G. (1985). The theory of moral-cognitive judgment: A socio-psychological assessment. In G. Lind, H. A. Hartmann, & R. Wakenhut (Eds.), Moral development and the social environment. Studies in the philosophy and psychology of moral judgment and education (pp. 21–53). Chicago, IL: Precedent. Lind, G. (2002). Ist Moral lehrbar? Ergebnisse der modernen moralpsychologischen Forschung [Can morality be taught? Research findings from modern moral psychology]. Berlin, Germany: Logos. Lind, G. (2008). The meaning and measurement of moral judgment competence. A dual-aspect model. In D. Fasko & W. Willis (Eds.), Contemporary philosophical and psychological perspectives on moral development and education (pp. 185–220). Creskill, NJ: Hampton Press. Lykken, D. T. (1995). The antisocial personalities. Mahwah, NJ: Erlbaum. Mathieu, C., & St-Jean, É. (2013). Entrepreneurial personality: The role of narcissism. Personality and Individual Differences, 55, 527–531. doi: 10.1016/j.paid.2013.04.026 Miller, J. D., Few, L. R., Seibert, L. A., Watts, A., Zeichner, A., & Lynam, D. R. (2012). An examination of the Dirty Dozen measure of psychopathy: A cautionary tale about the costs of brief measures. Psychological Assessment, 24, 1048–1053. doi: 10.1037/a0028583 Musch, J., Brockhaus, R., & Bröder, A. (2002). Ein Inventar zur Erfassung von zwei Faktoren sozialer Erwünschtheit [An inventory for the assessment of two factors of social desirability]. Diagnostica, 48, 121–129. doi: 10.1026/00121924.48.3.121 Paulhus, D. L. (1998). The Balanced Inventory of Desirable Responding. Toronto, ON: Multi-Health Systems. Paulhus, D. L., & Williams, K. M. (2002). The dark triad of personality: Narcissism, machiavellianism, and psychopathy. Journal of Research in Personality, 36, 556–563. doi: 10.1016/S0092-6566(02)00505-6 Rammstedt, B., & John, O. P. (2007). Measuring personality in one minute or less: A 10-item short version of the Big Five Inventory in English and German. Journal of Research in Personality, 41, 203–212. doi: 10.1016/j.jrp.2006.02.001

Journal of Individual Differences 2016; Vol. 37(1):24–30

A. Krick et al.: Dark Triad and Student Disciplinary Choice

Rauthmann, J. F. (2013). Investigating the MACH–IV with item response theory and proposing the trimmed MACH*. Journal of Personality Assessment, 95, 388–397. doi: 10.1080/00223891.2012.742905 Tassy, S., Deruelle, C., Mancini, J., Leistedt, S., & Wicker, B. (2013). High levels of psychopathic traits alters moral choice but not moral judgment. Frontiers in Human Neuroscience, 7, 229. doi: 10.3389/fnhum.2013.00229 Wilson, M. S., & McCarthy, K. (2011). Greed is good? Student disciplinary choice and self-reported psychopathy. Personality and Individual Differences, 51, 873–876. doi: 10.1016/ j.paid.2011.07.028 Winters, J., Christoff, K., & Gorzalka, B. B. (2010). Dysregulated sexuality and high sexual desire: Distinct constructs? Archives of Sexual Behavior, 39, 1029–1043. doi: 10.1007/ s10508-009-9591-6 Zettler, I., Friedrich, N., & Hilbig, B. E. (2011). Dissecting work commitment: The role of Machiavellianism. Career Development International, 16, 20–35. doi: 10.1108/ 13620431111107793

Journal of Individual Differences 2016; Vol. 37(1):24–30

Date of acceptance: April 14, 2015 Published online: February 29, 2016

Martin Rettenberger Centre for Criminology Viktoriastraße 35 65189 Wiesbaden Germany Tel. +49 611 15758-0 Fax +49 611 15758-10 E-mail m.rettenberger@krimz.de

Ó 2016 Hogrefe Publishing

Original Article

Social Support, Emotional Intelligence, and Posttraumatic Stress Disorder Symptoms A Mediation Analysis Nicole L. Hofman, Austin M. Hahn, Christine K. Tirabassi, and Raluca M. Gaher The University of South Dakota, Vermillion, SD, USA Abstract. Exposure to traumatic events and the associated risk of developing Posttraumatic stress disorder (PTSD) symptoms is a significant and overlooked concern in the college population. It is important for current research to identify potential protective factors associated with the development and maintenance of PTSD symptoms unique to this population. Emotional intelligence and perceived social support are two identified protective factors that influence the association between exposure to traumatic events and PTSD symptomology. The current study examined the mediating role of social support in the relationship between emotional intelligence and PTSD symptoms. Participants included 443 trauma-exposed university students who completed online questionnaires. The results of this study indicated that social support mediates the relationship between emotional intelligence and reported PTSD symptoms. Thus, emotional intelligence is significantly associated with PTSD symptoms and social support may play an integral role in the relationship between emotional intelligence and PTSD. The current study is the first to investigate the role of social support in the relationship between emotional intelligence and PTSD symptoms. These findings have important treatment and prevention implications with regard to PTSD. Keywords: social support, trauma, PTSD, emotional intelligence

Exposure to traumatic events is a common, yet historically overlooked, experience among college students. Research has consistently demonstrated high prevalence of traumatic events in samples of college students, with rates ranging from 56% to 84% (Marx & Sloan, 2003; Read, Ouimette, White, Colder, & Farrow, 2011; Smyth, Hockemeyer, Heron, Wonderlich, & Pennebaker, 2008). College students are at a particularly high risk to experience motor-vehicle accidents, life-threatening medical illness, unexpected death of a significant other, natural disasters, physical and sexual assault, and exposure to combat (Marx & Sloan, 2003; Read et al., 2011). A potential consequence of exposure to traumatic events is posttraumatic stress disorder (PTSD). The lifetime prevalence of PTSD for an adult in the United States is 8.7% and the 12-month prevalence rate for the same population is 3.5% (Kessler, Chiu, Demler, & Walters, 2005). The current prevalence of PTSD is estimated to range from 9% to 16% among college student populations (Read et al., 2011; Smyth et al., 2008; Twamley, Hami, & Stein, 2004). Given the high rates of exposure to traumatic events among college students, it is clear that the associated risk of developing PTSD symptomology is a significant concern. Thus, identifying potential risk and protective factors

Ă&#x201C; 2016 Hogrefe Publishing

associated with the development and maintenance of PTSD symptoms, unique to the college population, is important. Identification of these factors can contribute to developing appropriate prevention and intervention strategies that may be particularly relevant for the college student population. As such, the current study investigates the role of emotional intelligence (EI) and perceived social support (SS) on severity of PTSD symptoms within a college sample. Emotional intelligence (EI) is a construct that encompasses perception of emotional experiences, understanding and managing emotions in self and others, and using emotions to facilitate problem solving (Salovey & Mayer, 1990). Within the literature, there are two differing models concerning the conceptualization and measurement of EI. Mayer, Salovey, and Caruso (2004) suggest EI is best defined as ability, much like, cognitive intelligence, while others describe and measure EI as a trait (Neubauer & Freudenthaler, 2005; Petrides & Furnham, 2003). Although research indicates ability and trait EI are mutually exclusive conceptualizations, Schutte and Malouff (2008) suggest ability EI may ultimately facilitate and encourage the development of trait EI. Thus, ability and trait models are corresponding, and individuals with high levels of ability EI may, in turn, demonstrate more EI traits and behaviors.

Journal of Individual Differences 2016; Vol. 37(1):31â&#x20AC;&#x201C;39 DOI: 10.1027/1614-0001/a000185

N. L. Hofman et al.: Social Support, Emotional Intelligence, PTSD

The current study investigates trait EI because based on a meta-analysis demonstrates stronger associations with mental distress compared to ability EI (Martins, Ramalho, & Morin, 2010). Results from various studies suggest trait EI is a protective factor against the development of psychological symptoms following a traumatic event (Hunt & Evans, 2004; Kwako, Szanton, Saligan, & Gill, 2011). For example, Armstrong, Galligan, and Critchley (2011) found higher trait EI predicted fewer reported anxiety and depression symptoms in response to negative life events among adults. In addition, Hunt and Evans (2004) concluded participants with higher levels of trait EI reported fewer PTSD symptoms after experiencing a traumatic event. Furthermore, Kwako et al. (2011) found that deficits in trait EI were linked to the onset of depression symptoms among individuals who recently experienced a traumatic event. Thus, research suggests that trait EI is associated with fewer negative consequences, such as symptoms of PTSD, following a traumatic event. Trait EI is an important facet of emotion regulation, thus providing a potential explanation for the observed association between trait EI and PTSD symptoms (Hunt & Evans, 2004; Mayer & Salovey, 1995; Salovey & Mayer, 1990). High levels of trait EI are associated with more effective management of emotions in distressing situations. Individuals with high levels of trait EI may be better able to identify and describe their emotional experience during and in response to the event (Hunt & Evans, 2004) and subsequently regulate their emotions more effectively. In addition, Mayer and Salovey (1995) suggested deficits in emotion regulation could lead to poorer mental health, as an individual is unable to process important emotional information in relation to an event. Thus, as trait EI is associated with effective emotion regulation, deficits in trait EI may lead to difficulties processing emotional information associated with a traumatic experience (Mayer & Salovey, 1995). Although research indicates that EI is associated with PTSD, the mechanisms responsible for this relationship are not fully understood. Perceived social support (SS) may be an important variable underlying the relationship between EI and PTSD. In general, SS is a broad construct that encompasses many subtypes of social interactions including perceived and received SS (Hobfoll, 1988). Perceived SS is related to perceived availability of satisfying relationships that can provide the individual with care and help as needed (Haber, Cohen, Lucas, & Baltes, 2007). Conversely, received SS is the actual assistance provided by others (Hobfoll, 1988). Perceived SS depends on idiosyncratic evaluation of available resources rather than objective measures of existing relationships, and has been found to be a better predictor of mental health outcomes (DiMatteo, 2004; Norris & Kaniasty, 1996; Uchino, 2009; Smerglia, Miller, & Kort-Butler, 1999). In addition, perceived SS is a factor related to the development and maintenance of PTSD following a traumatic experience in

Journal of Individual Differences 2016; Vol. 37(1):31–39

both civilian and military samples (Brewin, Andrews, & Valentine, 2000; Ozer, Best, Lipsey, & Weiss, 2008; Smith, Benight, & Cieslak, 2013). Given the associations with perceived SS and mental health outcomes, the current study focuses on this particular subtype of SS. Research examining the relationship between SS and PTSD suggests that SS is a protective factor with regard to the development and maintenance of symptoms following a traumatic event (Cohen & Wills, 1985) as it increases an individual’s ability to cope and recover (Burgess & Holmstrom, 1978). In addition, perceived SS may increase the probability that an individual shares emotional concerns and information related to their distress with those in their support network (Chaudoir & Fisher, 2010; Fenlason & Beehr, 1994). The opportunity to share personal experiences and feel empathy from others (Mohay & Forbes, 2009) is associated with fewer PTSD symptoms (Cohen et al., 2000; Lepore, Silver, Wortman, & Wayment, 1996). Perceived SS has also been shown to increase perceived safety, which leads to decreased PTSD symptoms (Cai, Ding, Tang, Wu, & Yang, 2014). Finally, perceived SS may decrease PTSD symptoms because it can lead to appraisals of stress as manageable (Mohay & Forbes, 2009). An individual’s ability to recognize adequate SS in their environment may be influenced by individual processes, such as EI. Research indicates a functional association between EI and perceived SS, such that individuals with higher EI perceive greater availability of SS (Gallagher & Vella-Brodrick, 2008; Montes-Berges & Augusto, 2007). In addition, high levels of trait EI are positively related with social network size and quality (Austin, Saklofske, & Egan, 2005). Furthermore, individuals who score high on measures of trait EI are more likely to report higher levels of relationship satisfaction (Schröder-Abé & Schütz, 2011). Researchers theorize that individuals with higher EI have better SS because greater ability to manage emotions makes it easier to have positive interactions and establish relationships with others (Salovey, Bedell, Detweiler, & Mayer, 2000; Furnham & Petrides, 2003; Schutte et al., 2002). In addition, EI may be associated with social capability, allowing for adequate development of relationships and enhancement of the availability of SS resources (Salovey et al., 2000). For instance, individuals with better emotion regulation abilities have more positive interactions (Lopes et al., 2004, 2011) and fewer negative interactions with others (Lopes, Salovey, & Straus, 2003). Thus, trait EI appears to influence the quality and availability of SS, resulting in greater perceived SS. In summary, people with higher EI tend to perceive more SS (Gallagher & Vella-Brodrick, 2008). Perceived SS may, consequently, contribute to the maintenance of or reduction of PTSD symptoms following a traumatic experience, as people with lower perceived SS report more PTSD symptoms than those with higher perceived SS (Brewin et al., 2000; Ozer et al., 2008; Smith et al., 2013). Therefore, it likely that EI is associated with fewer PTSD symptoms due to the experience of perceived SS. This notion is

Ó 2016 Hogrefe Publishing

N. L. Hofman et al.: Social Support, Emotional Intelligence, PTSD

supported by findings that perceived SS mediates the relationship between EI and mental distress (Kong, Zhao, & You, 2012).

from one to 14 (M = 3.05, SD = 2.16). The most common traumatic events reported included serious accidents (18%), natural disasters (18%), fears of being killed (13%), seeing someone injured or killed (16%), seeing dead bodies (21%), sexual assault (11%), and physical abuse (10%).

Current Study The current study examined the relationships between EI, perceived SS, and PTSD symptoms in a sample of trauma-exposed college students. Research indicates that a positive relationship exists between perceived SS and trait EI. More specifically, higher levels of trait EI predict more perceived SS (Gallagher & Vella-Brodrick, 2008; MontesBerges & Augusto, 2007). Moreover, amount and quality of perceived SS is associated with PTSD symptoms and may account for the maintenance of PTSD symptomology. Although independent associations have been found between trait EI and PTSD symptoms, as well as perceived SS and PTSD symptoms, the combined mechanisms that may account for these relations have not been identified. Thus, the purpose of the current study is to examine the role of SS on relationship between trait EI and PTSD symptoms. It is hypothesized perceived SS will mediate the relationship between trait EI and PTSD symptoms. In addition, the current study controlled for negative affectivity and gender because negative affectivity is associated with SS (Measelle, Stice, & Springer, 2006; Scholz, Kliegel, Luszczynska, & Knoll, 2012), EI (Wang, 2002), and PTSD symptoms (Fetzner, Collimore, Carleton, & Asmundson, 2012), and gender differences exist across all variables, especially with regard to PTSD (Tolin & Foa, 2008). In addition, an individual’s negative affectivity could impact his or her perception of social support networks and increase an individual’s self-reported posttraumatic stress symptomology. The current study was interested in understanding the relationships between social support, emotional intelligence, and posttraumatic stress symptoms over and above the effects of negative affectivity. We also tested if negative affectivity and gender had an association with PTSD symptoms. Negative affectivity was significantly associated with PTSD symptoms, but gender was not.

Method Participants Complete data was collected from 573 undergraduate students from a mid-size Midwestern United States university. However, analyses within the current study utilized data from 443 participants who endorsed experiencing at least one traumatic event on the Trauma History Questionnaire, which assesses exposure to 24 different traumatic events. Within this subsample, the age of participants ranged from 18 to 25 years (M = 19.70, SD = 1.53). In addition, a majority of the sample was female (69%). With regard to race, 95.9% identified as white, 1.8% as African American, 0.5% as Asian, 1.0% as Latino or Hispanic, and 1.0% as Multiracial. Number of reported traumatic events ranged Ó 2016 Hogrefe Publishing

Measures Emotional Intelligence Scale (EIS; Schutte et al., 1998) The EIS is a 33-item self-report inventory designed to measure an individual’s appraisal of emotional experiences, emotion regulation, and how one utilizes emotions. Total scores can range from 33 to 165 and higher scores are indicative of a higher level of EI. The EIS is one of the most widely used assessments of trait EI and the validity and reliability of this measure are widely supported with many diverse samples (Petrides & Furnham, 2000; Schutte et al., 1998, 2001; Schutte, Schuettpelz, & Malouff, 2000). In samples of college students, reported alphas are relatively high (i.e., Cronbach’s a = .84, Austin, Saklofske, Huang, & McKenney, 2004; Cronbach’s a = .90, Gardner & Qualter, 2010). In the current sample, the reliability coefficient for the total score was Cronbach’s a = .92. Posttraumatic Stress Disorder Checklist-Specific (PCL-S; Weathers, Litz, Herman, Huska, & Keane, 1993) PTSD symptoms were assessed utilizing the PCL-Specific version in which participants are asked to focus on a specific traumatic event. The PCL-S is a 17-item self-report measure that examines each of the diagnostic symptoms outlined in the Diagnostic and Statistical Manual of Mental Disorders (4th ed.; American Psychiatric Association, 1994) for PTSD. Total scores can range from 17 to 85, with higher scores being indicative of a greater degree and severity of PTSD symptoms. The PCL is the most widely used instrument among professionals (Elhai, Gray, Kashdan, & Franklin, 2005) and the psychometric properties of the measure have been examined extensively (Blanchard, JonesAlexander, Buckley, & Forneris, 1996; Ruggiero, Del Ben, Scotti, & Rabalais, 2003; Weathers et al., 1993). Previous reliability estimates of the PCL in college samples are high (i.e., Cronbach’s a = .90; Adkins, Weathers, McDevittMurphy, & Daniels, 2008). The alpha in this sample for the total score was Cronbach’s a = .85. The Multidimensional Scale of Perceived Social Support (MSPSS; Zimet, Dahlem, Zimet, & Farley, 1988) The MSPSS is a 12-item self-report inventory designed to measure perceived SS from family, friends, and a significant other. A total score and three subscale scores can be obtained with higher scores being indicative of higher levels of perceived SS. Research has demonstrated Journal of Individual Differences 2016; Vol. 37(1):31–39

N. L. Hofman et al.: Social Support, Emotional Intelligence, PTSD

that the MSPSS has good psychometric properties and the measure has previously been utilized with college student populations (Canty-Mitchell & Zimet, 2000; Eker & Arkar, 1995; Zimet et al., 1988). The Cronbach alpha estimate for the total score in this sample was a = .91. Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988) We used the trait version of the PANAS. The PANAS lists 10 positive and 10 negative emotions and scores for negative affect range from 10 to 50, with higher scores representing higher levels of negative affect. Watson, Clark, and Tellegen (1988) report alphas of .84 to .97 for the Negative Affect (NA) scale. The alpha in this sample was Cronbach’s a = .88.

support mediated the relationship between emotional intelligence and PTSD symptoms. In the first regression, PTSD symptoms were regressed on emotional intelligence. Second, social support was regressed on emotional intelligence. Last, PTSD symptoms were regressed on emotional intelligence and social support. In each of the regression analyses, negative affect and gender were included as covariates because each has been associated with reported differences in PTSD symptoms (Frazier et al., 2011; Tolin & Foa, 2008). Based on recommendations from Holmbeck (2002), the direct effect included in the full model was then calculated by conducting a Sobel test (Sobel, 1982). The significance of this effect was examined by calculating bias-corrected bootstrapped confidence intervals for the coefficients using Stata 12 (StataCorp, 2011).

Results

Procedure Recruitment and Data Collection Participants were recruited via an online psychology recruitment program and college students between the ages of 18 and 25 years were eligible to participate. Participants received extra credit in specified courses as compensation for volunteering. All questionnaires were completed online and each participant’s responses were anonymous. In the current study, researchers adhered to the ethical guidelines as outlined by the American Psychological Association’s Ethical Principles of Psychologists and Code of Conduct (American Psychological Association, 2002) and the study was approved by the university’s Institutional Review Board (IRB).

Data Analyses All statistical analyses were calculated utilizing Stata statistical software version 12 (StataCorp, 2011). Two-tailed significances were reported. Following the recommendations of Baron and Kenny (1986), a series of standard multiple regression analyses were completed to examine whether social

Correlations among variables can be found in Table 1. Negative affect was significantly negatively correlated with SS and EI and positively related with PTSD symptoms as measured by the PCL. In addition, scores on the measure of perceived SS were significantly positively associated with higher levels of EI and negatively associated with PTSD symptoms. EI was also significantly related to PTSD symptoms. With regard to the distribution of scores on the various measures, scores for EI ranged from 50 to 163 (M = 122.64, SD = 14.46). The mean score for PTSD symptoms was 34.75 (SD = 14.22) and the scores ranged from 17 to 85. The National Center for PTSD (2014) suggests a cut-point score of 30 for general population (e.g., nonclinical, civilian, etc.). Seventy percent of the traumaexposed sample (N = 310) scored on or above the cut point, indicating clinical significance. With regard to various types of symptoms, scores on the Reexperiencing subscale ranged from 5 to 25 (M = 10.42, SD = 4.86), Avoidance subscale ranged from 7 to 33 (M = 13.47, SD = 6.11), and the scale of Hyperarousal symptoms ranged from 5 to 24 (M = 10.32, SD = 4.49). Scores on the SS measure ranged from 12 to 84 (M = 67.03, SD = 14.21). Negative affect scores ranged from 1 to 4.5 (M = 2.20, SD = 0.71).

Table 1. Correlations between social support, EI, and PTSD symptoms 1. 2. 3. 4. 5. 6. 7. 8.

Gender NA MSPSS EIS PCL PCL-B PCL-C PCL-D

–

.08 –

3 .17** .16** –

4 .09 .35** .41** –

5 .07 .59** .30** .39** –

6 .14** .54*** .20*** .29*** .89*** –

7 .01 .52*** .33*** .42*** .93*** .75*** –

8 .04 .55*** .26*** .32*** .88*** .68*** .74*** –

Notes. **Correlation is significant at the .01 level. ***Correlation is significant at the < .001 level. Gender = Women (0), M = Men (1). NA = Negative Affect Total Score. MSPSS = Social Support Total Score. EIS = Emotional Intelligence Scale Total Score. PCL = PTSD Symptoms Total Score. PCL-B = PTSD Cluster B: PCL Reexperiencing Subscale Score. PCL-C = PTSD Cluster C: PCL Avoidance Subscale Score. PCL-D = PTSD Cluster D: PCL Hyperarousal Subscale Score. Journal of Individual Differences 2016; Vol. 37(1):31–39

Ó 2016 Hogrefe Publishing

N. L. Hofman et al.: Social Support, Emotional Intelligence, PTSD

Table 2. Tests of social support as a mediator in the relationship between EI and PTSD symptoms b

gp2

Model with dependent variable (PTSD) regressed on independent variable (EI) Gender 1.41 1.21 Negative affect 10.23 12.82 Emotional intelligence 0.21 5.35

.228 < .001 < .001

.05 .51 .21

< .01 .23 .04

Model with mediator (social support) regressed on independent variable (EI) Gender 4.31 3.20 Negative affect 0.75 0.81 Emotional intelligence 0.38 8.28

.001 .416 < .001

.14 .04 .38

.01 < .01 .13

Model with (PTSD) regressed on mediator (social support) and independent variable (EI) Gender 2.12 1.83 .068 Negative affect 10.11 12.87 < .001 Emotional intelligence 0.15 3.57 < .001 Social support 0.17 4.07 < .001

.07 .51 .15 .17

< .01 .22 .02 .02

Note. Gender was coded women (0), men (1).

Mediation Analyses Three multiple regression analyses were completed to test the conditions required to test for a mediation effect. Covariates in each analysis included gender and negative affect. The full model included PTSD total symptom score as the criterion variable, the hypothesized mediator, SS total score, the predictor, EI total score, and the covariates. A summary of all mediation analyses, including effect sizes, can be found in Table 2. In the first model, results indicated that EI was a significant predictor of PTSD symptoms, F(3, 439) = 93.43, p < .0001, R2 = 0.39. Results of the second model indicate that perceived SS was significantly related to EI, F(3, 439) = 33.55, p < .0001, R2 = 0.19. With regard to the full model, results indicated that the model was significant, F(4, 438) = 76.71, p < .0001, R2 = 0.41. SS was inversely related to PTSD symptoms (b = .166, p < .0001). In this model, negative affect also emerged as a significant predictor (b = 10.11, p < .0001); however, gender was not a significant predictor of PTSD symptoms (b = 2.12, p = .07). The direct effect between perceived SS and PTSD symptoms was calculated and the significance of the direct effect was determined by utilizing a Sobel Mediation Test and bias-corrected bootstrapped confidence intervals. Results from these analyses indicate that SS was a significant mediator (ab = .062 95% CI ( .101 to .024). In addition, SS mediated 30% of the total effect. EI remained significant in the full model, which indicates that perceived SS mediated the association between EI and PTSD symptoms. A summary of the direct and total effects can be found in Figure 1.

Discussion The current study examined the mediating role of SS in the relationship between EI and PTSD symptoms within a trauma-exposed sample of college students. As hypothesized, perceived SS was a significant mediator in the Ó 2016 Hogrefe Publishing

Social Support -.166**

.376**

-.211** Emotional Intelligence

-.148**

PTSD Symptoms

Total Effect Direct Effect

Figure 1. Total and direct effects in the mediation model examining the role of social support in the relationship between EI and PTSD symptoms. relationship between EI and reported PTSD symptoms. Thus, EI is associated with increased levels of perceived SS, which is then related to a decrease in reported PTSD symptoms. This is the first study to demonstrate this relationship among EI, perceived SS, and PTSD symptoms. Results contribute to the literature in several ways. First, these findings support prior research, which found that SS mediated the relationship between EI and mental distress (Kong et al., 2012) and life satisfaction (Koydemir, Simsek, Schütz, & Tipandjan, 2011). Findings also indicate that EI is associated with better interpersonal functioning, which is also congruent with previous research findings (Schutte et al., 2001). Higher levels of EI may be associated with better social functioning as it may allow an individual to more effectively engage in social interactions leading to the establishment of more relationships and a better quality SS network. Finally, our results suggest that perceived amount and quality of SS may be associated with fewer Journal of Individual Differences 2016; Vol. 37(1):31–39

N. L. Hofman et al.: Social Support, Emotional Intelligence, PTSD

PTSD symptoms following a traumatic experience. These findings are similar to results from previous studies, which identified perceived SS as a protective factor in the development and maintenance of PTSD symptoms (Brewin et al., 2000; Ozer et al., 2008; Trickey, Siddaway, MeiserStedman, Serpell, & Field, 2012). Therefore, SS may serve to facilitate a person’s ability to effectively regulate the emotional experience associated with the traumatic event. Due to many possible risk and protective factors, both emotional intelligence and social support produced small, but significant effects. Despite the small effect sizes, the findings of the current study have important implications in terms of treatment and prevention interventions with regard to PTSD symptoms. A significant amount of research has examined protective factors, such as EI, that may influence negative consequences associated with the experience of a traumatic event as well as identifying specific components that may be useful in terms of reducing PTSD symptoms (Irving, Telfer, & Blake, 1997; Keller, Zoellner, & Feeny, 2010). For example, some research has indicated that by enhancing EI, psychological effects of stress may be reduced. The results of the current study identified a potential mechanism as for why this relationship exists. These results suggest enhancement of EI may influence the likelihood of seeking out SS after experiencing a traumatic event, thus leading to a reduction in PTSD symptoms. Similarly, perceived SS plays a vital protective role with regard to preventing and reducing PTSD symptoms (Brewin et al., 2000). More specifically, by increasing a person’s perceived SS network, PTSD symptoms may be reduced (Irving et al., 1997). However, the results of the current study suggest that within treatment, it may be more effective to focus on enhancing EI rather than simply focusing on enhancing a person’s network and quality of SS. There were some important limitations of this study. First, the design of the study was cross-sectional in nature. Therefore, although results support EI as a predictor of current PTSD symptoms via perceived SS, causal implications cannot be made. Future research using longitudinal designs could provide evidence for the temporal order of these variables. Similarly, because data was collected at a single time point, it is unknown how the relationships among the variables may vary across time. Results indicated that perceived SS only partially accounted for the relationship between EI and PTSD symptoms, therefore other mechanisms may influence the relationship between EI, perceived SS, and PTSD symptoms that were not identified in the current study. For example, unacknowledged social interactions can reduce mental distress to a greater extent than acknowledged SS (Bolger, Zuckerman, & Kessler). Therefore, aspects of SS that were not measured may be impacting the relationship between EI and PTSD. Additionally, given the specificity of the sample with regard to ethnicity, age, and other demographic characteristics the achieved results cannot be generalized to other populations. Finally, the current study included self-report inventories, and thus may have provided a limited picture of PTSD symptoms, perceived SS, and EI.

Journal of Individual Differences 2016; Vol. 37(1):31–39

There are various avenues for future research. First, these results should be replicated utilizing more diverse samples. In addition, given that perceived SS was only identified as a partial mediator, other possible mediators could be identified. Furthermore, studies utilizing a longitudinal design would be useful in terms of identifying how the relationship among EI, SS, and PTSD symptoms may change over time. Also, future research could focus on other mediation relationships among risk and protective factors in the development and maintenance of PTSD symptoms. It is imperative that future research target these limitations due to the high rate of traumatic experiences among college students (Marx & Sloan, 2003; Read et al., 2011; Smyth et al., 2008). The current study provides novel evidence that EI and SS are important variables associated with PTSD symptoms, and these are resources that can be readily targeted for prevention and intervention among college populations.

References Adkins, J. W., Weathers, F. W., McDevitt-Murphy, M., & Daniels, J. B. (2008). Psychometric properties of seven self-report measures of posttraumatic stress disorder in college students with mixed civilian trauma exposure. Journal of Anxiety Disorders, 22, 1393–1402. doi: 10.1016/j.janxdis.2008.02.002 American Psychiatric Association. (1994). Diagnostic and statistical manual of mental health disorders (4th ed.). Washington DC: Author. doi: 10.1176/appi.books. 9780890425596.991543 American Psychological Association. (2002). Ethical principles of psychologists and code of conduct. The American Psychologist, 57, 1060–1073. doi: 10.1037/e305322003-001 Armstrong, A. R., Galligan, R. F., & Critchley, C. R. (2011). Emotional intelligence and psychological resilience to negative life events. Personality and Individual Differences, 51, 331–336. doi: 10.1016/j.paid.2011.03.025 Austin, E. J., Saklofske, D. H., & Egan, V. (2005). Personality, well-being and health correlates of trait emotional intelligence. Personality and Individual Differences, 38, 547–558. doi: 10.1016/j.paid.204.05.009 Austin, E. J., Saklofske, D. H., Huang, S. H. S., & McKenney, D. (2004). Measurement of trait emotional intelligence: Testing and cross-validating a modified version of Schutte et al.’s (1998) measure. Personality and Individual Differences, 36, 555–562. doi: 10.1016/S0191-8869(03)00114-4 Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. doi: 10.1037/0022-3514.51.6.1173 Blanchard, E. B., Jones-Alexander, J., Buckley, T. C., & Forneris, C. A. (1996). Psychometric properties of the PTSD Checklist (PCL). Behaviour Research and Therapy, 34, 669–673. doi: 10.1016/0005-7967(96)00033-2 Bolger, N., Zuckerman, A., & Kessler, R. C. (2000). Invisible support and adjustment to stress. Journal of Personality and Social Psychology, 79, 953–961. doi: 10.1037/0022-3514. 79.6.953 Brewin, C. R., Andrews, B., & Valentine, J. D. (2000). Metaanalysis of risk factors for posttraumatic stress disorder in trauma-exposed adults. Journal of Consulting and Clinical Psychology, 68, 748–766. doi: 10.1037/0022-006X.68.5.748

Ó 2016 Hogrefe Publishing

N. L. Hofman et al.: Social Support, Emotional Intelligence, PTSD

Burgess, A. W., & Holmstrom, L. L. (1978). Recovery from rape and prior life stress. Research in Nursing and Health, 1, 165–174. doi: 10.1002/nur.4770010404 Cai, W., Ding, C., Tang, Y., Wu, S., & Yang, D. (2014). Effects of social supports on posttraumatic stress disorder symptoms: Moderating role of perceived safety. Psychological Trauma: Theory, Research, Practice, and Policy, 6, 724–730. doi: 10.1037/a0036342 Canty-Mitchell, J., & Zimet, G. D. (2000). Psychometric properties of the Multidimensional Scale of Perceived Social Support in urban adolescents. American Journal of Community Psychology, 28, 391–400. doi: 10.1023/ A:1005109522457 Chaudoir, S. R., & Fisher, J. D. (2010). The disclosure processes model: Understanding disclosure decision-making and post disclosure outcomes among people living with a concealable stigmatized identity. Psychological Bulletin, 136, 236–256. doi: 10.1037/a0018193 Cohen, H., Benjamin, J., Geva, A. B., Matar, M. A., Kaplan, Z., & Kotler, M. (2000). Autonomic dysregulation in panic disorder and in post-traumatic stress disorder: Application of power spectrum analysis of heart rate variability at rest and in response to recollection of trauma or panic attacks. Psychiatry Research, 96, 1–13. doi: 10.1016/S01651781(00)00195-5 Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310–357. doi: 10.1037/0033-2909.98.2.310 DiMatteo, M. (2004). Social support and patient adherence to medical treatment: A meta-analysis. Health Psychology, 23, 207–218. doi: 10.1037/0278-6133.23.2.207 Eker, D. D., & Arkar, H. H. (1995). Perceived social support: Psychometric properties of the MSPSS in normal and pathological groups in a developing country. Social Psychiatry and Psychiatric Epidemiology, 30, 121–126. doi: 10.1007/BF00802040 Elhai, J. D., Gray, M. J., Kashdan, T. B., & Franklin, C. (2005). Which instruments are most commonly used to assess traumatic event exposure and posttraumatic effects? A survey of traumatic stress professionals. Journal of Traumatic Stress, 18, 541–545. doi: 10.1002/jts.20062 Fenlason, K. J., & Beehr, T. A. (1994). Social support and occupational stress: Effects of talking to others. Journal of Organizational Behavior, 15, 157–175. doi: 10.1002/ job.4030150205 Fetzner, M. G., Collimore, K. C., Carleton, R. N., & Asmundson, G. G. (2012). Clarifying the relationship between AS dimensions and PTSD symptom clusters: Are negative and positive affectivity theoretically relevant constructs? Cognitive Behaviour Therapy, 41, 15–25. doi: 10.1080/16506073.2011.621971 Frazier, P. A., Gavian, M., Hirai, R., Park, C., Tennen, H., Tomich, P., & Tashiro, T. (2011). Prospective predictors of posttraumatic stress disorder symptoms: Direct and mediated relations. Psychological Trauma: Theory, Research, Practice, and Policy, 3, 27–36. doi: 10.1037/a0019894 Furnham, A., & Petrides, K. V. (2003). Trait emotional intelligence and happiness. Social Behavior and Personality, 31, 815–824. doi: 10.2224/sbp.2003.31.8.815 Gallagher, E. N., & Vella-Brodrick, D. A. (2008). Social support and emotional intelligence as predictors of subjective well-being. Personality and Individual Differences, 44, 1551–1561. doi: 10.1016/j.paid.2008.01.011 Gardner, K., & Qualter, P. (2010). Concurrent and incremental validity of three trait emotional intelligence measures. Australian Journal of Psychology, 62, 5–13. doi: 10.1080/ 00049530903312857 Haber, M. G., Cohen, J. L., Lucas, T., & Baltes, B. B. (2007). The relationship between self-reported received and Ó 2016 Hogrefe Publishing

perceived social support: A meta-analytic review. American Journal of Community Psychology, 39, 133–144. doi: 10.1007/s10464-007-9100-9 Hobfoll, S. E. (1988). The ecology of stress. Washington, DC: Hemisphere Publishing. Holmbeck, G. N. (2002). Post-hoc probing of significant moderational and mediational effects in studies of pediatric populations. Journal of Pediatric Psychology, 27, 87–96. doi: 10.1093/jpepsy/27.1.87 Hunt, N., & Evans, D. (2004). Predicting traumatic stress using emotional intelligence. Behaviour Research and Therapy, 42, 791–798. doi: 10.1016/j.brat.2003.07.009 Irving, L. M., Telfer, L., & Blake, D. D. (1997). Hope, coping, and social support in combat-related posttraumatic stress disorder. Journal of Traumatic Stress, 10, 465–479. doi: 10.1023/A:1024897406135 Keller, S. M., Zoellner, L. A., & Feeny, N. C. (2010). Understanding factors associated with early therapeutic alliance in PTSD treatment: Adherence, childhood sexual abuse history, and social support. Journal of Consulting and Clinical Psychology, 78, 974–979. doi: 10.1037/a0020758 Kessler, R. C., Chiu, W. T., Demler, O., & Walters, E. E. (2005). Prevalence, severity, and comorbidity of twelve-month DSM-IV disorders in the National Comorbidity Survey Replication (NCS-R). Archives of General Psychiatry, 62, 617–627. doi: 10.1001/archpsyc.62.6.617 Kong, F., Zhao, J., & You, X. (2012). Social support mediates the impact of emotional intelligence on mental distress and life satisfaction in Chinese young adults. Personality and Individual Differences, 53, 513–517. doi: 10.1016/j.paid. 2012.04.021 Koydemir, S., Simsek, O. F., Schütz, A., & Tipandjan, A. (2011). Emotional intelligence predicts components of subjective well-being beyond personality: A two-country study using self and informant reports. Journal of Positive Psychology, 7, 107–118. doi: 10.1080/17439760.2011.647050 Kwako, L. E., Szanton, S. J., Saligan, L. N., & Gill, J. M. (2011). Major depressive disorder in persons exposed to trauma: Relationship between emotional intelligence and social support. Journal of the American Psychiatric Nurses Association, 17, 237–245. doi: 10.1177/1078390311402498 Lepore, S. J., Silver, R., Wortman, C. B., & Wayment, H. A. (1996). Social constraints, intrusive thoughts, and depressive symptoms among bereaved mothers. Journal of Personality and Social Psychology, 70, 271–282. doi: 10.1037/00223514.70.2.271 Lopes, P. N., Brackett, M. A., Nezlek, J. B., Schütz, A., Sellin, I. & Salovey, P. (2004). Emotional intelligence and social interaction. Personality and Social Psychology Bulletin, 30, 1018–1034. doi: 10.1177/0146167204264762 Lopes, P. N., Nezlek, J. B., Extremera, N., Hertel, J., FernándezBerrocal, P., Schütz, A., & Salovey, P. (2011). Emotion regulation and the quality of social interaction: Does the ability to evaluate emotional situations and identify effective responses matter? Journal of Personality, 79, 429–467. doi: 10.1111/j.1467-6494.2010.00689.x Lopes, P. N., Salovey, P., & Straus, R. (2003). Emotional intelligence, personality, and the perceived quality of social relationships. Personality and Individual Differences, 35, 641–658. doi: 10.1016/S0191-8869(02)00242-8 Martins, A., Ramalho, N., & Morin, E. (2010). A comprehensive meta-analysis of the relationship between emotional intelligence and health. Personality and Individual Differences, 49, 554–564. doi: 10.1016/j.paid.2010.05.029 Marx, B. P., & Sloan, D. M. (2003). The effects of trauma history, gender, and race on alcohol use and posttraumatic stress symptoms in a college student sample. Addictive Behaviors, 28, 1631–1647. doi: 10.1016/j.addbeh.2003. 08.039 Journal of Individual Differences 2016; Vol. 37(1):31–39

N. L. Hofman et al.: Social Support, Emotional Intelligence, PTSD

Mayer, J. D., & Salovey, P. (1995). Emotional intelligence and the construction and regulation of feelings. Applied and Preventive Psychology, 4, 197–208. doi: 10.1016/s09621849(05)80058-7 Mayer, J. D., Salovey, P., & Caruso, D. (2004). Emotional intelligence: Theory, findings, and implications. Psychological Inquiry, 15, 197–215. doi: 10.1207/ s15327965pli1503_02 Measelle, J. R., Stice, E., & Springer, D. W. (2006). A prospective test of the negative affect model of substance abuse: Moderating effects of social support. Psychology of Addictive Behaviors, 20, 225–233. doi: 10.1037/0893-164X.20. 3.225 Mohay, H., & Forbes, N. (2009). Reducing the risk of posttraumatic stress disorder in children following natural disasters. Australian Journal of Guidance and Counselling, 19, 179–195. doi: 10.1375/ajgc.19.2.179 Montes-Berges, B. B., & Augusto, J. M. (2007). Exploring the relationship between perceived emotional intelligence, coping, social support and mental health in nursing students. Journal of Psychiatric and Mental Health Nursing, 14, 163–171. doi: 10.1111/j.1365-2850.2007.01059.x National Center for PTSD. (2014). Using the PTSD checklist for DSM-IV. Retrieved from http://www.ptsd.va.gov/professional/ pages/assessments/assessment-pdf/PCL-handout.pdf Neubauer, A. C., & Freudenthaler, H. H. (2005). Models of emotional intelligence. In R. Schultz & R. D. Roberts (Eds.), Emotional intelligence: An international handbook (pp. 31– 50). Cambridge, MA: Hogrefe. Norris, F. H., & Kaniasty, K. (1996). Received and perceived social support in times of stress: A test of the social support deterioration deterrence model. Journal of Personality and Social Psychology, 71, 498–511. Ozer, E. J., Best, S. R., Lipsey, T. L., & Weiss, D. S. (2008). Predictors of posttraumatic stress disorder and symptoms in adults: A meta-analysis. Psychological Trauma: Theory, Research, Practice, and Policy, 5, 3–36. doi: 10.1037/19429681.S.1.3 Petrides, K. V., & Furnham, A. (2000). Gender differences in measured and self-estimated trait emotional intelligence. Sex Roles, 42, 449–461. Petrides, K. V., & Furnham, A. (2003). Trait emotional intelligence: Behavioural validation in two studies of emotion recognition and reactivity to mood induction. European Journal of Personality, 17, 39–57. doi: 10.1002/ per.466 Read, J. P., Ouimette, P., White, J., Colder, C., & Farrow, S. (2011). Rates of DSM-IV-TR trauma exposure and posttraumatic stress disorder among newly matriculated college students. Psychological Trauma: Theory, Research, Practice, and Policy, 3, 148–156. doi: 10.1037/a0021260 Ruggiero, K. J., Del Ben, K., Scotti, J. R., & Rabalais, A. E. (2003). Psychometric properties of the PTSD ChecklistCivilian Version. Journal of Traumatic Stress, 16, 495–502. doi: 10.1023/a:1025714729117 Salovey, P., Bedell, B. T., Detweiler, J. B., & Mayer, J. D. (2000). Current directions in emotional intelligence research. In M. Lewis & J. M. Haviland (Eds.), Handbook of Emotions (2nd ed., pp. 504–520). New York, NY: Guilford Press. Salovey, P., & Mayer, J. D. (1990). Emotional intelligence. Imagination, Cognition and Personality, 9, 185–211. doi: 10.2190/dugg-p24e-52wk-6cdg Scholz, U., Kliegel, M., Luszczynska, A., & Knoll, N. (2012). Associated between received social support and positive and negative affect: Evidence for age differences from a dailydiary study. European Journal of Ageing, 9, 361–371. doi: 10.1007/s10433-012-0236-6

Journal of Individual Differences 2016; Vol. 37(1):31–39

Schröder-Abé, M., & Schütz, A. (2011). Walking in each other’s shoes: Perspective taking mediates effects of emotional intelligence on relationship quality. European Journal of Personality, 25, 155–169. doi: 10.1002/per.818 Schutte, N. S., & Malouff, J. M. (2008, July). A dimensional model of adaptive emotional functioning. Presented at the XXIX International Congress of Psychology, Berlin. Schutte, N. S., Malouff, J. M., Bobik, C., Coston, T. D., Greeson, C., Jedlicka, C., . . . Wendorf, G. (2001). Emotional intelligence and interpersonal relations. The Journal of Social Psychology, 141, 523–536. doi: 10.1080/ 00224540109600569 Schutte, N. S., Malouff, J. M., Hall, L. E., Haggerty, D. J., Cooper, J. T., Golden, C. J., & Dornheim, L. (1998). Development and validation of a measure of emotional intelligence. Personality and Individual Differences, 25, 167–177. doi: 10.1016/S0191-8869(98)00001-4 Schutte, N. S., Malouff, J. M., Simunek, M., Hollander, S., & McKenley, J. (2002). Characteristic emotional intelligence and emotional well-being. Cognition and Emotion, 16, 769–786. doi: 10.1080/02699930143000482 Schutte, N. S., Schuettpelz, E., & Malouff, J. M. (2000). Emotional intelligence and task performance. Imagination, Cognition and Personality, 20, 347–354. doi: 10.2190/J0X6BHTG-KPV6-2UX Smerglia, V. L., Miller, N. B., & Kort-Butler, L. (1999). The impact of social support on women’s adjustment to divorce: A literature review and analysis. Journal of Divorce & Remarriage, 32, 63–89. doi: 10.1300/J087v32n01_05 Smith, A. J., Benight, C. C., & Cieslak, R. (2013). Social support and postdeployment coping self-efficacy as predictors of distress among combat veterans. Military Psychology, 25, 452–461. doi: 10.1037/mil0000013 Smyth, J. M., Hockemeyer, J. R., Heron, K. E., Wonderlich, S. A., & Pennebaker, J. W. (2008). Prevalence, type, disclosure, and severity of adverse life events in college students. Journal of American College Health, 57, 69–76. doi: 10.3200/JACH.57.1.69-76 Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equation models. In S. Leinhardt (Ed.), Sociological methodology 1982 (pp. 290–312). Washington, DC: American Sociological Association. doi: 10.2307/270723 StataCorp. (2011). Stata Statistical Software: Release 12. College Station, TX: StataCorp LP. Tolin, D. F., & Foa, E. B. (2008). Sex differences in trauma and posttraumatic stress disorder: A quantitative review of 25 years of research. Psychological Trauma: Theory, Research, Practice, and Policy, 5, 37–85. doi: 10.1037/0033-2909. 132.6.959 Trickey, D., Siddaway, A. P., Meiser-Stedman, R., Serpell, L., & Field, A. P. (2012). A meta-analysis of risk factors for posttraumatic stress disorder in children and adolescents. Clinical Psychology Review, 32, 122–138. doi: 10.1016/ j.cpr.2011.12.001 Twamley, E. W., Hami, S., & Stein, M. B. (2004). Neuropsychological function in college students with and without posttraumatic stress disorder. Psychiatry Research, 126, 265–274. doi: 10.1016/j.psychres.2004.01.008 Uchino, B. N. (2009). Understanding the links between social support and physical health: A life-span perspective with emphasis on the separability of perceived and received support. Perspectives on Psychological Science, 4, 236–255. doi: 10.1111/j.1745-6924.2009.01122.x Wang, C. (2002). The relationship between emotional intelligence and anxiety, depression, and mood in a sample of college students. Chinese Journal of Clinical Psychology, 10, 298–299.

Ó 2016 Hogrefe Publishing

N. L. Hofman et al.: Social Support, Emotional Intelligence, PTSD

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54, 1063–1070. doi: 10.1037/0022-3514.54. 6.1063 Weathers, F. W., Litz, B. T., Herman, D. S., Huska, J. A., & Keane, T. M. (1993, October). The PTSD Checklist (PCL): Reliability, validity, and diagnostic utility. Paper presented at the annual meeting of the International Society for Traumatic Stress Studies, San Antonio, TX. Zimet, G. D., Dahlem, N. W., Zimet, S. G., & Farley, G. K. (1988). The Multidimensional Scale of Perceived Social Support. Journal of Personality Assessment, 52, 30–41. doi: 10.1207/s15327752jpa5201_2

Ó 2016 Hogrefe Publishing

Date of acceptance: May 12, 2015 Published online: February 29, 2016 Austin M. Hahn Clinical Psychology Training Program Department of Psychology The University of South Dakota 414 East Clark Street – SDU 107 Vermillion, South Dakota 57069-2390 USA Tel. +1 605 677-5353 Fax +1 605 677-3195 E-mail austin.hahn@usd.edu

Journal of Individual Differences 2016; Vol. 37(1):31–39

An essential reference and tool-kit for treating trauma survivors “Still the best book on trauma therapy in print... This book is a must for your trauma bookshelf.” Dr. Michael Dubi, President of the International Association of Trauma Professionals

Anna B. Baranowsky / J. Eric Gentry

Trauma Practice

Tools for Stabilization and Recovery 3rd ed. 2015, xii + 204 pp., large format US $59.00 / € 41.95 ISBN 978-0-88937-471-3 Also available as an eBook Trauma Practice, now in its 3rd edition, is back by popular demand! Filled with new resources, this book based on the tri-phasic trauma treatment model is a guide for both seasoned trauma therapists and newer mental health professionals seeking practical approaches that work. Clearly written and detailed, Trauma Practice provides the reader with an array of techniques, protocols and interventions for effectively helping trauma survivors. This book will help you address the (cognitive, behavioral, body-oriented, and emotional / relational) aftermath of trauma using impactful care approaches. In addition to presenting the foun-

www.hogrefe.com

dations of CBT trauma treatment, the authors also provide step-by-step explanations of many popular and effective CBT techniques developed through the lens of phased trauma therapy. Interventions include Trigger List Development, 3-6 Breath Training, Layering, Systematic Desensitization, Exposure Therapy, Story-Telling Approaches, as well as new approaches inspired by recent research on neuroplasticity such as Picture Positive, Corrective Messages from Old Storylines, and Thematic Map. Completely new sections are devoted to forward-facing trauma therapy, and clinician self-care. This is a manual that you will find useful everyday in your trauma practice.

Journal of Personnel Psychology nline free o issue le samp

Editor-in-Chief Bernd Marcus University of Hagen, Germany Managing Editor Petra Gelléri University of Hagen, Germany

ISSN-Print 1866-5888 ISSN-Online 2190-5150 ISSN-L 1866-5888 4 issues per annum (= 1 volume)

Subscription rates (2016) Libraries / Institutions US $324.00 / € 238.00 Individuals US $159.00 / € 114.00 Postage / Handling US $16.00 / € 12.00

www.hogrefe.com

About the Journal The Journal of Personnel Psychology welcomes excellent empirical and theoretical contributions to basic and applied research in personnel psychology and related methodology. Reviews are also welcome. Articles deal with all fields in personnel psychology, including selection, performance measurement, motivation, leadership, organizational commitment, personnel development and training, new test developments, and job analysis. As many topics in personnel psychology are closely related to issues in other branches of psychology or, more generally, the social sciences and human resource management, the journal is open to contributions of an interdisciplinary nature. There are five categories of submission: original articles, research notes, review articles, registered reports and hybrid registered reports. Full details are available on the journal‘s website at www.hogrefe.com/ journals/jpp.

Associate Editors Ian Gellatly, Alberta, Canada Barbara Griffin, Sydney, Australia Laurenz Meier, Fribourg, Switzerland Sandra Ohly, Kassel, Germany Despoina Xanthopoulou, Thessaloniki, Greece Xin-An Zhang, Shanghai, China

Manuscript Submissions All manuscripts should be submitted online at www.editorialmanager.com/jppsy, where full instructions to authors are also available. Electronic Full Text The full text of the journal – current and past issues (from 2002 onward) – is available online at http://econtent.hogrefe.com/ loi/pps (included in subscription price). A free sample issue is also available here. Abstracting Services The Journal of Personnel Psychology is abstracted/indexed in Current Contents/Social and Behavioral Sciences (CC/S&BS), Social Sciences Citation Index (SSCI), PsyJOURNALS, PsycINFO, PSYNDEX, IBZ, and IBR.

Original Article

Rank-Order Consistency and Profile Stability of Self- and InformantReports of Personal Values in Comparison to Personality Traits Henrik Dobewall1 and Toivo Aavik2 1

Universitat Pompeu Fabra, Barcelona, Spain, 2University of Tartu, Estonia

Abstract. This study compares the three-year rank-order consistency of informant-reports of values with stability estimates of self-rated values as meta-analytically reviewed and within the same sample. Whether the hierarchy of values attributed to an individual is as stable as in target’s self-reports was assessed with profile correlations. Self- and informant-reports of personality traits were available for direct comparison. Results indicated that informant-reports of values were not less stable across time than self-rated values or than other-ratings of traits. This was true for the relative position of a person within a sample as well as the relative ordering of these measures within the same individuals. The observed longitudinal stability of informant-reports of values implies that they can serve as a reliable source of information. Moreover, the temporal stability of value/trait profiles was found to predict subjective well-being. Keywords: Schwartz’s values and Big Five personality, self- and informant-reports, rank-order consistency, longitudinal profile stability, adjustment

Stability and change in individuals can be distinguished by at least four orthogonal forms (Roberts & DelVecchio, 2000). Most commonly reported for personal values (Schwartz, 1992) is rank-order consistency (Bardi, Buchanan, Goodwin, Slabu, & Robinson, 2014; Bardi, Lee, Hofmann-Towfigh, & Soutar, 2009; Lönnqvist, Jasinskaja-Lahti, & Verkasalo, 2011; Schwartz, 2005), a stability estimate assessed by test-retest correlations, which refers to the relative placement of individuals within a sample over time. Change in terms of mean-level (or normative) development refers to increases or decreases in value priorities of most people as they grow older (e.g., Dobewall, Tormos, & Vauclair, in press). Researchers have also studied intra-individual differences in value change (e.g., Lönnqvist, Jasinskaja-Lahti, & Verkasalo, 2013) and the structure of value change (Bardi et al., 2009). A final form, called ipsative consistency, has been applied in other fields (Klimstra, Luyckx, Hale, Goossens, & Meeus, 2010; Lönnqvist, Mäkinen, Paunonen, Henriksson, & Verkasalo, 2008), but not yet in values’ research. Often assessed by profile correlations (Furr, 2008), it addresses the question of whether the ranking of a set of variables is consistent between two measurement time points. This last observation is surprising, as personal values develop their meaning by their ‘‘relative’’ importance within a person (Gollan & Witte, 2014; Schwartz, 1992). Journal of Individual Differences 2016; Vol. 37(1):40–48 DOI: 10.1027/1614-0001/a000186

Further, there are problems associated with self-report methodology, namely, that peoples’ self-ratings are potentially biased. Participants may respond in a socially desirable (Schwartz, Verkasalo, Antonovsky, & Sagiv, 1997) or self-serving (Christopher & Schlenker, 2004) way and/or have a tendency to apply diverse response styles (He, Van de Vijver, Dominguez-Espinosa, & Mui, 2014). An alternative is to assess a person’s social representation as held by close and knowledgeable others (Vazire, 2006), such as, good friends, family members, or partners. Other-ratings of personal values (see Dobewall, Aavik, Konstabel, Schwartz, & Realo, 2014) are especially useful when self-reports cannot be obtained (as in the case of very young children) or are simply not available anymore (e.g., if the target has passed away). It is obvious, nevertheless, that informants, even those who know the target well, miss information that is less visible (McAdams, 1995) or that cannot be revealed (Pronin, Fleming, & Steffel, 2008). At the same time, observers sometimes know a person better than the target knows her/himself (Vazire & Carlson, 2011). Thus, informant-reports add unique insights about a person and may have predictive validity for daily behavior incremental to those obtained from a self-report methodology (Vazire & Mehl, 2008). Dobewall, Aavik, and Realo (in press), for example, showed that other-rated values can be used to substitute and complement self-reports of values Ó 2016 Hogrefe Publishing

H. Dobewall & T. Aavik: Longitudinal Stability of Self- and Other-Reports of Values and Traits

in explaining individual differences in academic achievement. Finally, Dobewall et al. (2014) found that informants can, indeed, make quite accurate ratings of their target’s values. While studies in support of the stability of self-rated values are becoming increasingly available, not a single study on personal values has assessed the temporal consistency of a person’s social representation as held by close others. This study, therefore, goes beyond earlier work by focusing on the longitudinal stability of other-ratings of values as compared to self-reports. We take personality as a point of reference because the question of longitudinal consistency (i.e., within a sample as well as within the same person) has been repeatedly asked in relation to personality traits. Finally, personality theory (e.g., Roberts, Caspi, & Moffitt, 2001) has proposed that ‘‘maturity is related to changes toward a desirable endpoint and that the likelihood of personality change diminishes as individuals come closer to that endpoint’’ (Klimstra et al., 2010, p. 1180). And in support of this claim, the temporal stability of people’s personality trait profiles has repeatedly been found to predict psychological adjustment (e.g., Lönnqvist et al., 2008). Still, research on informant-reports of traits (cf. Klimstra et al., 2010), as well as on facets, is warranted (cf. Costa & McCrae, 2006). Sheldon (2005) showed that change in values (rank-order) is also able to account for positive change in psychological well-being. However, it is yet an unanswered question if the stability of an individual’s hierarchy of values (self- or other-rated) has any predictive validity for psychological adjustment.

The Temporary Stability of Values and Traits Schwartz’s (1992) theory distinguishes between ten specific value types, which – as guiding principles in a person’s life – vary in importance. According to the Big Five Factor theory of personality (FFT; McCrae & Costa, 1999, 2008), traits are tendencies to behave, think, and feel in consistent ways. Traits can be considered to be positive or negative and vary in terms of how much and the intensity with which individuals exhibit them (Bilsky & Schwartz, 1994). Even though values and traits can be referred to in very similar terms, it is often the case that those individuals who highly endorse a value type (an exemplar Achievement item: ‘‘Being very successful is important to him/her. He/she hopes people will recognize his/her achievements’’) do not exhibit the corresponding trait facet (an exemplar Achievement striving item (C4; reversed): ‘‘I work just as much as I have to and I am not very ambitious; I do not set very high goals for myself’’). Most importantly, no consensus has been reached on the issue of whether the one psychological construct is more stable than the other. Peoples’ values are seen as ‘‘relatively stable’’ (Rokeach, 1973, p. 11), even though they can change under certain conditions (Bardi & Goodwin, 2011). Changes in values can happen due to maturational factors (Dobewall, Tormos, & Vauclair, in press; Schwartz, 2005) and contextual variation (Bardi et al., 2014; Tormos, Ó 2016 Hogrefe Publishing

Vauclair, & Dobewall, in press), or a combination of the two. Nevertheless, the relative ordering of people’s value priorities in a sample, as well as the ranking of these measures within the same individuals, should largely remain stable. The FFT, however, classifies values as characteristic adaptations and, as such, as more malleable than a person’s traits, due to interactions with the environment. Personality traits are also viewed as ‘‘relatively stable’’ (McCrae & Costa, 2008, p. 167). What is more, the FFT postulates that there is little, if any, transfer from the environment (e.g., life experiences) to personality traits (McCrae & Costa, 2008). Unfortunately, there is no meta-analysis available on the rank-order consistency of the Schwartz (1992) ten specific value types that could give more certainty about whether personal values show test-retest correlations at the same high level as personality traits. Also, moderators of the rank-order consistency of values have yet to be identified. Roberts and DelVecchio (2000) meta-analytically reviewed self- and other-rated traits and found a rank-order consistency of r > .30 in early childhood. The consistency estimates increased stepwise until leveling, at r > .70, at age 50 and older. Moreover, for personality traits, it was found that the longitudinal stability of other-ratings is at a comparable level to self-reported data (Costa & McCrae, 1988; Laidra, Allik, Harro, Merenäkk, & Harro, 2006). So, while value theory does not lead us to expect any large differences between traits and values in their rankorder consistency, FFT views traits as more stable than values. Research on other-rated traits allows us to be optimistic regarding the temporal stability of informant-reports of values.

Aim of the Current Study The aim of the current study is to examine if informantreports of values are more (or less) stable across time than self-rated values and/or other-rated traits. We are interested in the extent to which the relative position of a person within a sample changes (rank-order consistency), as well as in the stability of trait/value hierarchies within the same individual (ipsative consistency). Only if other-rated values display a considerably high level of longitudinal stability can they be considered as a reliable source of information about a person’s personality. A secondary aim of this study is to test the predictive validity of the temporal stability of peoples’ value/trait profiles (self- and other-rated) for psychological adjustment.

Analytical Strategy After conducting a meta-analysis of the extant literature on the rank-order consistency of the ten self-rated value types, we assess the longitudinal correlations between informantreports and self-reported data as well as self- and otherrated trait facets. These analyses can indicate, in this context, if person A is seen by his/her close others at both T1 and T2 as more universalistic than person B, but not Journal of Individual Differences 2016; Vol. 37(1):40–48

H. Dobewall & T. Aavik: Longitudinal Stability of Self- and Other-Reports of Values and Traits

if Universalism has changed in its importance within person A relative to the other nine value types. We therefore conduct profile correlations (Furr, 2008) in order to test if an individual’s hierarchy of values/traits has changed over time (self- and other-rated). The overall profile consistency, however, includes a component common to the average profile within a population. Several authors (e.g., Klimstra et al., 2010), furthermore, point to the importance of ‘‘distinctive or unique aspects of one’s personality at Time t and the distinctive aspects of the personality of that same person at Time t + 1’’ (p. 1166). The distinctiveness of a pair of a person’s self- and other-rated profiles can be analyzed after z-standardizing these ratings for both groups of raters separately (M = 0; SD = 1) before correlating the profiles. These distinctive consistency estimates will be used to assess the stability of the unique aspects of an individual’s self-image and of that part of a person’s personality that makes them special in the eye of close others. Longitudinal profile correlations assign a stability score to each individual, which makes it possible to correlate these scores with a dependent variable. Finally, we examine if the stability (overall and distinct) of peoples’ value and trait profiles explains individual differences in psychological adjustment. This is done using the example of subjective well-being (SWB; Diener, Suh, Lucas, & Smith, 1999).

Method Participants The initial survey (T1) – conducted in 2011 as part of a Self-Presentation and Impression Management course – consisted of 96 informant- and 101 self-reports of students at the University of Tartu (see Dobewall et al., 2014, for details). In 2014, we invited the (now former) undergraduate students via email and/or http://www. facebook.com to do a follow-up survey. Fifty-three (former) students (7 male, 46 female) accepted the invitation. At T2, targets’ mean age was 29.3 (SD = 8.7) years. On the last page of the questionnaire they were given we asked for the email address of two informants who knew them well. We emphasized that it would be an advantage if the informants were the same as at T1. At T2, 71.2% of the informants were new. However, Laidra and colleagues (2006) showed that, in their study, test-retest correlations were only marginally affected by using the same versus new informants. In 14 out of the total of 41 cases, two other-ratings were available. We used the average of the two-informantreports, making them a rather abstract but reliable measure of the social representation of a person’s values and traits (Dobewall et al., 2014). The majority of informants (T2) were friends (55.9%), but many were spouses (17.6%), children (13.2%), or girl- or boyfriends (8.8%) of the targets; the rest were either other family members or workmates. The other-ratings (average age 33.1, SD = 12.6) were slightly gender-biased (only 30.9% were male). The average length of Journal of Individual Differences 2016; Vol. 37(1):40–48

acquaintance was 13.4 years (SD = 10.6), with the majority of the informants seeing the target at least once a week. For their participation, all participants were eligible to win a prize of one of ten ten-euro shopping vouchers. We also promised to give feedback about their individual value and trait profiles, about how their close others saw them, and on how their self-image and social representations had changed over the years.

Measures Schwartz’s Values The original Schwartz Values Survey (SVS; 1992) uses 56 ratings of importance to assess people’s guiding principles in life (e.g., ‘‘mature love’’). The Portrait Values Questionnaire (PVQ; Schwartz et al., 2001), developed later, measures values in a less abstract way: indirectly through a comparison task with a fictive person’s goals, aspirations, and wishes. The 21-item version of the PVQ, which is included in the core questionnaire of the European Social Survey (http://www.europeansocialsurvey.org/), was used. Respondents had to decide how much the person described was like them (self-rated form) or like the target (informant form). Individuals are known to differ in how they respond to a values questionnaire, however, and it has been found that the average rating an individual gives to all value items independent of their content (coined MRAT) is governed by several response styles, such as, acquiescence or social desirability bias (Schwartz et al., 1997). In order to correct for differences in peoples’ tendency to respond in different ways, for the PQV, we subtracted MRAT from each raw item before aggregating them into the ten specific value types of Schwartz’s theory (1992). Paired sample t-tests were conducted. We found at least marginally significant differences in means across time in the following four value pairs: Hedonism-self at T1 (M = 0.2, SD = 0.84) and T2 (M = 0.4, SD = 0.83); t(52) = 1.73, p = .09, Conformity-self at T1 (M = 0.6, SD = 1.06) and T2 (M = 0.9, SD = 1.05); t(52) = 2.20, p = .03, Self-Direction-other at T1 (M = 0.6, SD = 0.67) and T2 (M = 1.0, SD = 0.55); t(40) = 3.36, p = .01, and Security-other at T1 (M = 0.1, SD = 0.70) and T2 (M = 0.22, SD = 0.87); t(40) = 2.23, p = .03. These mean-level developments could be interpreted as normative value change. Big Five Personality Traits Self-report and informant-report forms of the Short Five (S5; Konstabel, Lönnqvist, Walkowitz, Konstabel, & Verkasalo, 2012) were administered. The NEO personality inventory (NEO – PI; Costa & McCrae, 1992) has 30 positively-coded and 30 reversed items. Each balanced pair assesses one of six personality facets, which then can be aggregated into their broader domains (the Big Five). This study was conducted at the level of trait facets in order to allow for a fair comparison with Schwartz’s (1992) ten Ó 2016 Hogrefe Publishing

H. Dobewall & T. Aavik: Longitudinal Stability of Self- and Other-Reports of Values and Traits

specific value types (cf. Dobewall et al., 2014). Mean-level development took place at the level of trait facets in five pairs. There were significant differences in the scores for N2-self at T1 (M = 0.2, SD = 3.41) and T2 (M = 0.5, SD = 3.09); t(50) = 1.76, p = .09, O6: Openness to values-self at T1 (M = 1.4, SD = 2.26) and T2 (M = 2.7, SD = 2.53); t(50) = 3.69, p = .01, A6: Tender-mindedness-self at T1 (M = 0.7, SD = 2.21) and T2 (M = 4.0, SD = 1.94); t(50) = 10.97, p = .00, O1: Openness to fantasy-other at T1 (M = 2.5, SD = 2.32) and T2 (M = 1.6, SD = 2.67); t(38) = 1.95, p = .06, and O3: Openness to feelings-other at T1 (M = 4.1, SD = 1.60) and T2 (M = 3.3, SD = 1.95); t(38) = 2.53, p = .02.

value types for each study. The 11 studies were coded for the age group of the sample (adolescents 1,5; students 1; representative 0; adults 1), sample size as an indicator of study quality, time between T1 and T2 (months), and if measured with the SVS (1) or the PVQ (0), which differ in their reliabilities (see Schwartz, 2005). After standardizing all variables, a multiple regression analysis in which we regressed the effect sizes on the four potential moderators was conducted (R2 = 0.81; F(4, 6) = 6.389, p = .024). Regression results indicated that the time that had elapsed between the fist and the last time of measurement was the only significant moderator of a study’s average rank-order consistency (t(6) = 4.642, p < .01). Effect sizes ¼ 0:997 time þ 0:277 quality

Subjective Well-Being

þ 0:192 age group þ 0:095

The concept of SWB is ‘‘down-to-earth, innocent, and nonideological, but promising and in everyone’s vocabulary rut,’’ and, as it has a motivational component, happiness, life satisfaction ‘‘triggers behavioral responses’’ (Brockmann & Delhey, 2010, p. 1). Diener, Emmons, Larsen, and Griffin’s (1985) 5-item Satisfaction with Life Scale (SWLS) was administered at T2. The SWLS was developed to access overall cognitive evaluations of a person’s life and used in the current study as a measure of psychological adjustment. For each respondent, a sum score was computed based on the answers provided on the 5-point agree-disagree scales.

SVS

Results Rank-Order Consistency of Self- and OtherRated Values and Personality Traits Meta-Analysis We meta-analytically reviewed the test-retest correlations of published studies of Schwartz’s (1992) ten value types, using either the SVS or the PVQ (Bardi et al., 2009, 2014; Lönnqvist et al., 2011; Schwartz, 2005).1 Table 1 presents the results of this meta-analysis (self-reports only). Across the 11 samples, the weighted mean effect size for the ten value types was r = .61, ranging from .55 for Benevolence to .66 for Tradition. The table also shows the standard error of the mean and the 95% confidence intervals. This moderate level of stability in values is already remarkable, as the studies vary in life contexts – with some respondents going through a major a life transition – and were drawn from several different countries. In order to test for potential interaction effects (Wolf, 1986), we averaged the rank-order consistency of the ten 1 2

ð1Þ

Comparing the Rank-Order Stability of Selfand Other-Rated Values/Traits Within the Same Sample Second, we assessed how longitudinally stable self- and other-reports of the ten specific value types and the Big Five trait facets are in our dataset in terms of the relative position of a person within a sample. Three years had elapsed between the first and the second measurement times. The participants were in the middle of their studies at T1 and most of them had already graduated at T2. After this life transition, the rank-order consistencies for values (Table 2) in informant-reports were r = .37. The highest longitudinal stability was observed for Stimulation (r = .59) and the lowest for Hedonism (r = .15 ns). Selfrated values correlated at r = .50, ranging from Tradition (r = .23 ns) to Conformity (r = .65). Therefore, the average rank-order consistency was lower in informant-reports of values, but not significantly so when compared to self-reports of values (Z = 0.75, p = .45; Lowry, 2014).2 We also calculated test-retest correlations for the 30 narrow trait facets (Table 3). Informant-report longitudinal estimates of personality traits ranged from r = .15 (ns) for N5: Impulsiveness to .60 for E1: Warmth, and had an average test-retest correlation of r = .38. Thus, differences in test-retest correlations of other-rated specific value types and other-rated trait facets were also not statistically significant (Z = 0.05, p = .96). For the self-ratings of trait faces, the correlations averaged r = .63, peaking for A1: Trust (r = .80) and at their minimum for A6: Tender-mindedness (r = .46). It should be further noted that self-rated values were neither significantly less stable across time than in our meta-analysis (Z = 1.03, p = .30) nor the self-rated trait facets in the same sample (Z = 0.95, p = .34).

The meta-analysis and the profile correlations were conducted in Microsoft Excel 2010 (see also Electronic Supplementary Material 1). For all other analyses IBM SPSS Statistics 22.0 was used. We also calculated the self-other agreement in the ten value types at T1 and T2. The analysis (N = 41) indicated that the average self-other agreement did to change over time, with rs of .35 (SD = 0.16) and .37 (SD = 0.12), respectively. This positive self-other correlation after three years is another stronger sign for the claim that informant-reports of values are valid information about a person’s personality.

Ó 2016 Hogrefe Publishing

Journal of Individual Differences 2016; Vol. 37(1):40–48

0.62 0.61 0.63 0.63 0.57 0.65 0.55 0.66 0.63 0.58 0.52 0.44 0.58 0.40 0.55 0.60 0.44 0.63 0.49 0.53 0.53 0.57 0.67 0.60 0.70 0.76 0.59 0.82 0.66 0.12 0.52 0.63 0.53 0.41 0.49 0.53 0.37 0.63 0.54 0.46 0.64 0.50 0.53 0.68 0.54 0.70 0.65 0.53 0.50 0.53 0.66 0.61 0.64 0.65 0.58 0.62 0.60 0.66 0.68 0.58 0.70 0.62 0.60 0.67 0.64 0.76 0.48 0.66 0.70 0.57 0.58 0.63 0.65 0.62 0.53 0.64 0.50 0.66 0.59 0.59 0.75 0.74 0.70 0.76 0.71 0.70 0.76 0.76 0.77 0.82 0.77 0.82 0.65 0.76 0.70 0.75 0.62 0.80 0.72 0.70 0.94 0.93 0.67 0.87 0.79 0.91 0.84 0.90 0.77 0.81 Power Achievement Hedonism Stimulation Self-direction Universalism Benevolence Tradition Conformity Security

Journal of Individual Differences 2016; Vol. 37(1):40–48

Notes. PVQ = Portrait Values Questionnaire; SVS = Schwartz Values Questionnaire; SE = Standard error of the mean; CI 95% = 95% confidence intervals. We could not correct the results for attenuation because some of the included papers did not report Cronbach’s alphas for each time point/study separately.

.70 .71 .67 .71 .64 .71 .64 .74 .69 .69 0.54 0.52 0.59 0.54 0.51 0.58 0.47 0.59 0.56 0.46

14 261

0.50 0.34 0.49 0.50 0.44 0.58 0.48 0.50 0.45 0.39

Weighted mean

PVQ Adults 19 145

Bardi et al. (2014) Bardi et al. (2009) Schwartz (2005) Data from

Values measure used PVQ PVQ SVS PVQ SVS PVQ SVS Age of sample Adults Students Students Representative Students Adolescents Students Time between T1 and T2 (months) 1 1.5 1.5 24 3 9 12 N= 26 157 205 870 119 807 129

Table 1. Meta-analysis of the rank-order consistency of ten the value types (self-rated)

Lönnqvist et al. (2011)

SVS PVQ SVS Adults Adults Students 9 18 24 63 151 196

0.04 0.05 0.02 0.04 0.03 0.03 0.04 0.04 0.03 0.06

( ) (+)

H. Dobewall & T. Aavik: Longitudinal Stability of Self- and Other-Reports of Values and Traits

CI 95%

Table 2. Three-year rank-order consistency estimates for the ten specific value types (self- and otherrated) Self-reports N= Benevolence Universalism Self-direction Stimulation Hedonism Achievement Power Security Tradition Conformity

53 .57*** .38*** .55*** .58*** .36** .59*** .51*** .55*** .23* .65***

Informant-reports 41 .15 .34** .43*** .59*** .22 .47*** .33** .45*** .30* .43***

Notes. Marked test-retest correlations are significant at *p < .10, **p < .05, or at ***p < .01, two-tailed.

Analyzing the Ipsative Consistency of Self- and Other-Rated Values/Traits Third, we computed the overall and distinct profile stability scores for each individual. Table 4 shows that, on average, the overall consistency of other-rated value types, r = .59, was no less stable than for self-reports of value types, r = .67 (Z = 0.62, p = .54) or for informant-reports of trait facets, r = .63 (Z = 0.27, p = .78). There was also not any significant difference between the average distinct profile consistency for other-rated traits, r = .39, when compared to informant-reports of values, r = .32 (Z = 0.34, p = .73). The unique aspects of a target’s selfrated value types (r = .52) were clearly more stable but not significantly more so than the unique aspects of informantreports of the same measure (Z = 1.14, p = .25). When assessed within the same individuals, self-rated value and trait stabilities also did not differ for the overall (r = .67/.71, respectively) or for the distinct (r = .52/.58) profiles (Z = 0.38, p = .70/Z = 0.43, p = .66, respectively). Testing the Predictive Validity of Peoples’ Profile Stability Fourth, we correlated the obtained ipsative consistency scores with our adjustment variable, namely, respondents’ self-reports of their overall SWB. In self-rated data, both the overall and distinct profile consistencies of personal values (r = .28/.30) and personality traits (r = .31/.42, respectively) correlated significantly with the SWLS. In other-rated data, SWB showed associations only with the stability of the unique aspects of the target’s value hierarchy, r = .31.

Discussion Dobewall and colleagues (2014), Dobewall, Aavik, et al. (in press) were pioneers in systematically studying Ó 2016 Hogrefe Publishing

Ó 2016 Hogrefe Publishing

Notes. For self-ratings, all test-retest correlations are significant at p < .01. For other-ratings, test-retest correlations r > .26 are significant at p < .10; r > .32 are significant at p < .05; r > .40 are significant at p < .01; all two-tailed.

.70 .66 .49 .62 .46 .61 .74 .66 .63 .72 .58 .80 .69 .73 .58 .54 .67 .54 .80 .59 .63 .61 .60 .46 .53 .79 .63 .59 .65 .67 .29 .57 .46 .52 .15 .40 .60 .32 .42 .57 .38 .49 .39 .21 .48 .49 .50 .48 .26 .44 .27 .60 .30 .38 .39 .24 .17 .16 .26 .23 51 39 Self-reports Informant-reports

E4 E5 E6 O1 O2 O3 O4 O5 O6 A1 A2 A3 A4 A5 A6 C1 C2 C3 C4 C5 C6 N = N1 N2 N3 N4 N5 N6 E1 E2 E3

Table 3. Three-year rank-order consistency estimates for the 30 personality trait facets (self- and other-rated)

H. Dobewall & T. Aavik: Longitudinal Stability of Self- and Other-Reports of Values and Traits

informant-reports of Schwartz’s (1992) personal values in terms of their agreement with self-reports and their predictive validity. This study contributes to the literature by examining the longitudinal stability of these constructs. Stability and change in individuals can take diverse forms (cf. Roberts & DelVecchio, 2000). We observed only minor mean-level (or normative) development in the ten value types and the 30 personality facets (nine out of 80 longitudinal pairs showed significant differences). Furthermore, the observed pattern of normative change in values – with increases in Conformity and Security and decreases in Hedonism and Self-Direction – was actually in exactly the opposite direction as found in earlier cross-sectional data (e.g., Schwartz, 2005). This finding calls for further investigation, as change in value priorities can happen due to both maturational factors (e.g., Dobewall, Tormos, & Vauclair, in press) and interactions with the environment (e.g., Bardi et al., 2014). Our main hypothesis found support in that the perceptions of knowledgeable informants about another person’s values showed three-year rank-order consistencies at a comparable level to self-reports (i.e., in the same data and as meta-analytically reviewed), and that these were no less stable than other-rated personality trait facets. Also, selfreports of values and traits did not show significant differences in their test-retest correlations. Thus, it is not only the self-image of a person that is rather resistant to change. The presented results imply that how others see you (i.e., a person’s social representation) might also change less with time than one would expect. The presented meta-analysis of rank-order consistencies in self-reported values, however, might suffer from publication bias. The analysis, nevertheless, found that the length of the time had a negative effect on the observed test-retest correlations. The other tested moderators did not have a significant effect. The intra-individual (ipsative) stability of a person’s hierarchy of values and traits was decomposed into two components: overall profile consistency and distinct profile consistency. We found that the temporal consistency of the ranking of the ten value types and the 30 trait facets within the same individuals are also at comparable levels. This result is of special interest for value researchers as personal values are proposed to develop their meaning by their relative importance within a person (Gollan & Witte, 2013; Schwartz, 1992). Contrary to findings in some of the literature (e.g., Pronin et al., 2008), the distinct ipsative stability of other-rated values further suggests that people can reveal to others what makes them unique. Our findings also contribute to personality theory, as we found ipsative stability (overall and distinct) for other-rated personality at the level of trait facets, which usually receives less attention than self-reports of personality domains (cf. Costa & McCrae, 2006; Klimstra et al., 2010). Therefore, when looking at the reported three-year estimates of rank-order consistency and ipsative consistency, we arrive at the same solution, namely, that (even though being smaller) there exists no significant difference in the stability of informant-reports of another person’s values as compared to his/her self-reports. There was also not Journal of Individual Differences 2016; Vol. 37(1):40–48

H. Dobewall & T. Aavik: Longitudinal Stability of Self- and Other-Reports of Values and Traits

Table 4. Three-year ipsative consistency estimates of self- and other-rated value and trait profiles and their correlation with the Satisfaction with Life Scale (SWLS) Self-reports Personal values

N= Overall stability Distinctive stability

Informant-reports

Personality traits

Personal values

Personality traits

M (SD)

SWLS

M (SD)

SWLS

M (SD)

SWLS

M (SD)

SWLS

53 0.67 (0.28) 0.52 (0.34)

46 .28* .30*

51 0.71 (0.58) 0.58 (0.22)

46 .31** .42***

41 0.59 (0.31) 0.32 (0.38)

39 .06 .31**

39 0.63 (0.23) 0.39 (0.31)

35 .18 .13

Notes. Marked correlation coefficients are significant at *p < .10, **p < .05, or at ***p < .01, two-tailed.

any significant difference in informant-ratings of the person’s personality traits. This implies that other-ratings of personal values can serve as a reliable measure of a person’s personality, just like informant-reports of traits (Roberts & DelVecchio, 2000). Close others (friends, spouses or partners, parents, etc.) should be used more frequently in value research, not only when self-reports are not available, but also in order to add unique information about the target. The observed moderate longitudinal stability of otherreports of value types/trait facets over 3 years is also remarkable, especially regarding the distinct aspects of the target profiles, due to the fact that many of the informants were not the same at T2 as at T1. In line with earlier research (e.g., Laidra et al., 2006), however, we do not have any indication if the longitudinal stability estimates of other-ratings of personal values/personality traits were strongly affected by the use of both new and old informants. We also tested if it is beneficial for an individual to be consistent in his or her hierarchy of values/traits. For personality traits, a diverse set of adjustment variables was previously found to correlate with their ipsative stability (e.g., Lönnqvist et al., 2008). The causality appears to flow in both directions. On the one hand, psychological adjustment fosters changes toward more desirable personality traits (Roberts et al., 2001). On the other hand, when individuals have reached a temporally stable profile, they tend to report more positive and less negative psychological outcomes. Based on the presented results, and keeping the small N of the current study in mind, it seems that the overall as well as the distinct stability of traits indeed have validity for predicting SWB (our adjustment variable). This is in some conflict with Klimstra and colleagues’ (2010) findings and replication studies are therefore needed. Surveying samples from non-Western countries might also be useful for testing if the predictive validity of temporal profile consistency also applies cross-culturally (see Slabu, Lenton, Sedikides, & Bruder, 2014). As values are commonly viewed as relatively stable (e.g., Rokeach, 1973), most research in values has not addressed the question of whether intra-individual value change is of any relevance for psychological adjustment (see Sheldon, 2005, for an exception). Even though Gollan and Witte (2014) did not find different value profiles and SWB to be associated, we found both the overall (in selfreports only) and the distinct (self- and other-rated) profile stability of values to predict individual differences on the SWLS in our data. Taking Gollan and Witte’s (2014) and Journal of Individual Differences 2016; Vol. 37(1):40–48

our findings together indicates that the endorsement of any combination of values can make people happy and satisfied. Yet, that people report high SWB levels appears to be more often the case if they have maintained a stable hierarchy of values over time. It is especially interesting that, if close others think that another person consistently holds a unique set of values, then these targets also scored higher in SWB. These other-perceptions of personality might even be a source of happiness and life satisfaction for the targets themselves (Dobewall, Realo, Allik, Esko, & Metspalu, 2013). Other-ratings of SWB, in return, eventually also contribute to increases (or decreases) in an individual’s profile consistency. The current paper notably shows that whether or not values change or remain stable within the same individual can have meaningful implications for theory development. Personality (e.g., Klimstra et al., 2010; Roberts et al., 2001) and happiness researchers (e.g., Brockmann & Delhey, 2010) have independently proposed that subjective life evaluations motivate behavioral responses. Future work is yet needed to test if the causality also goes from positive life evaluations to value stability, as has been shown for personality traits. Furthermore, even with a time gap of merely three years, the average rank-order consistency (r = .50) and the average ipsative stability (r = .67, overall/.52, distinct) of self-rated values did not exceed moderate levels. One could expect these correlations to drop when more years elapse. At the same time, personal values appear to be less malleable than most other constructs characterized as ‘‘characteristic adaptations.’’

Acknowledgments This research was supported by a postdoctoral research fellowship of the Spanish Ministry of Economy and Competitiveness and by a grant from the Estonian Research Council (PUT78). We thank Anu Realo and Delaney Michael Skerrett for their comments on a previous version of this paper. Electronic Supplementary Material The electronic supplementary material is available with the online version of the article at http://dx.doi.org/10.1027/ 1614-0001/a000186 Ó 2016 Hogrefe Publishing

H. Dobewall & T. Aavik: Longitudinal Stability of Self- and Other-Reports of Values and Traits

ESM 1. Matrix (Excel list). Complete results of the study.

References *Marked articles are included in the meta-analysis. Bardi, A., Buchanan, K. E., Goodwin, R., Slabu, L., & Robinson, M. (2014). Value stability and change during self-chosen life transitions: Self-selection versus socialization effects. Journal of Personality and Social Psychology, 106, 131–147. Bardi, A., & Goodwin, R. (2011). The dual route to value change: Individual processes and cultural moderators. Journal of Cross-Cultural Psychology, 42, 271–287. *Bardi, A., Lee, J. A., Hofmann-Towfigh, N., & Soutar, G. (2009). The structure of intraindividual value change. Journal of Personality and Social Psychology, 97, 913–929. Bilsky, W., & Schwartz, S. H. (1994). Values and personality. European Journal of Personality, 8, 163–181. Brockmann, H., & Delhey, J. (2010). The dynamics of happiness and the dynamics of happiness research. Social Indicator Research, 97, 1–5. Costa, P. T., & McCrae, R. R. (1988). Personality in adulthood: A six-year longitudinal study of self-reports and spouse ratings on the NEO Personality Inventory. Journal of Personality and Social Psychology, 54, 853–863. Costa, P. T., & McCrae, R. R. (1992). Revised NEO personality inventory (NEO PI-R) and NEO five-factor inventory (NEOFFI) professional manual. Odessa, FL: Psychological Assessment Resources. Costa, P. T., & McCrae, R. R. (2006). Age changes in personality and their origins: Comment on Roberts, Walton, and Viechtbauer. Psychological Bulletin, 132, 26–28. Christopher, A. N., & Schlenker, B. R. (2004). Materialism and affect: The role of self-presentation concerns. Journal of Social and Clinical Psychology, 23, 260–272. Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction with life scale. Journal of Personality Assessment, 49, 71–75. Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being: Three decades of progress. Psychological Bulletin, 125, 276–302. Dobewall, H., Aavik, T., Konstabel, K., Schwartz, S. H., & Realo, A. (2014). A comparison of self-other agreement in personal values versus the Big Five personality traits. Journal of Research in Personality, 50, 1–10. Dobewall, H., Aavik, T., & Realo, A. (in press). Informantreports of basic values: A valid predictor of academic performance? Learning and Individual Differences. Submitted for publication. Dobewall, H., Realo, A., Allik, J., Esko, T., & Metspalu, A. (2013). Self-other agreement in happiness and life-satisfaction: The role of personality traits. Social Indicators Research, 114, 479–492. Dobewall, H., Tormos, R., & Vauclair, M.-C. (in press). Normative value change across the human life cycle: An examination of cultural differences in age and cohort effects. International Journal of Psychology. Submitted for publication. Furr, R. M. (2008). A framework for profile similarity: Integrating similarity, normativeness, and distinctiveness. Journal of Personality, 76, 1267–1316. Gollan, T., & Witte, E. H. (2014). From the interindividual to the intraindividual level: Is the circumplex model of values applicable to intraindividual value profiles? Journal of Cross-Cultural Psychology, 45, 452–467. He, J., Van de Vijver, F. J. R., Dominguez-Espinosa, A., & Mui, P. H. C. (2014). Toward a unification of acquiescent, Ó 2016 Hogrefe Publishing

extreme, and midpoint response styles: A multilevel study. International Journal of Cross Cultural Management, 14, 306–322. doi: 10.1177/1470595814541424 Klimstra, T. A., Luyckx, K., Hale, W. W. III, Goossens, L., & Meeus, W. H. J. (2010). Longitudinal associations between personality profile stability and adjustment in college students: Distinguishing among overall stability, distinctive stability, and within-time normativeness. Journal of Personality, 78, 1163–1184. Konstabel, K., Lönnqvist, J.-E., Walkowitz, G., Konstabel, K., & Verkasalo, M. (2012). The ‘‘Short Five’’ (S5): Measuring personality traits using comprehensive single items. European Journal of Personality, 26, 13–29. Laidra, K., Allik, J., Harro, M., Merenäkk, L., & Harro, J. (2006). Agreement among adolescents, parents, and teachers on adolescent personality. Assessment, 13, 187–196. *Lönnqvist, J.-E., Jasinskaja-Lahti, I., & Verkasalo, M. (2011). Personal values before and after migration: A longitudinal case study on value change in Ingrian-Finnish migrants. Social Psychological and Personality Science, 2, 584–591. Lönnqvist, J.-E., Jasinskaja-Lahti, I., & Verkasalo, M. (2013). Rebound effect in personal values: Ingrian-Finnish migrants’ values two years after migration. Journal of Cross-Cultural Psychology, 44, 1122–1126. Lönnqvist, J.-E., Mäkinen, S., Paunonen, S. V., Henriksson, M., & Verkasalo, M. (2008). Psychosocial functioning in young men predicts their personality stability over 15 years. Journal of Research in Personality, 42, 599–621. Lowry, R. (2014). Difference between correlations. Retrieved from http://vassarstats.net/rdiff.html McAdams, D. P. (1995). What do we know when we know a person? Journal of Personality, 63, 365–396. McCrae, R. R., & Costa, P. T. (1999). A five-factory theory of personality. In L. A. Pervin & O. P. John (Eds.), Handbook of personality: Theory and research (2nd ed., pp. 139–153). New York, NY/London, UK: The Guilford Press. McCrae, R. R., & Costa, P. T. (2008). The five-factor theory of personality. In O. P. John, R. W. Robins, & L. A. Pervin (Eds.), Handbook of personality: Theory and research (3rd ed., pp. 159–181). New York, NY/London, UK: The Guilford Press. Pronin, E., Fleming, J. J., & Steffel, M. (2008). Value revelations: Disclosure is in the eye of the beholder. Journal of Personality and Social Psychology, 95, 795–809. Roberts, B. W., Caspi, A., & Moffitt, T. E. (2001). The kids are alright: Growth and stability in personality development from adolescence to adulthood. Journal of Personality and Social Psychology, 81, 670–683. Roberts, B. W., & DelVecchio, W. F. (2000). The rank-order consistency of personality from childhood to old age: A quantitative review of longitudinal studies. Psychological Bulletin, 126, 3–25. Rokeach, M. (1973). The nature of human values. New York, NY: Free Press. Schwartz, S. H. (1992). Universals in the content and structure of values: Theory and empirical tests in 20 countries. In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 25, pp. 1–65). New York, NY: Academic Press. *Schwartz, S. H. (2005). Robustness and fruitfulness of a theory of universals in individual human values. In A. Tamayo & J. Porto (Eds.), Valores e trabalho [Values and work]. Brasilia, Brazil: Editora Universidade de Brasilia. Schwartz, S. H., Melech, G., Lehmann, A., Burgess, S., Harris, M., & Owens, V. (2001). Extending the cross-cultural validity of the theory of basic human values with a different method of measurement. Journal of Cross-Cultural Psychology, 32, 519–542. Schwartz, S. H., Verkasalo, M., Antonovsky, A., & Sagiv, L. (1997). Value priorities and social desirability: Much Journal of Individual Differences 2016; Vol. 37(1):40–48

H. Dobewall & T. Aavik: Longitudinal Stability of Self- and Other-Reports of Values and Traits

substance, some style. The British Journal of Social Psychology, 36, 3–18. Sheldon, K. M. (2005). Positive change during college: Normative trends and individual differences. Journal of Research in Personality, 39, 209–223. Slabu, L., Lenton, A. P., Sedikides, C., & Bruder, M. (2014). Trait and state authenticity across cultures. Journal of CrossCultural Psychology, 45, 1347–1373. Tormos, R., Vauclair, M.-C., & Dobewall, H. (in press). What contextual factors contribute to individuals’ value change? A dynamic comparative multilevel analysis across 32 European countries. Journal of Cross-Cultural Psychology. Manuscript under review. Vazire, S. (2006). Informant reports: A cheap, fast, and easy method for personality assessment. Journal of Research in Personality, 40, 472–481. Vazire, S., & Carlson, E. N. (2011). Others sometimes know us better than we know ourselves. Current Directions in Psychological Science, 20, 104–108. Vazire, S., & Mehl, M. R. (2008). Knowing me, knowing you: The accuracy and unique predictive validity of self-ratings and other-ratings of daily behavior. Journal of Personality and Social Psychology, 95, 1202–1216.

Journal of Individual Differences 2016; Vol. 37(1):40–48

Wolf, F. M. (1986). Meta-analysis: Quantitative methods for research synthesis. Newbury Park, CA: Sage.

Date of acceptance: May 12, 2015 Published online: February 29, 2016

Henrik Dobewall Research and Expertise Centre for Survey Methodology (RECSM) Universitat Pompeu Fabra Ramón Trías Fargas 25-27 08005 Barcelona Spain Tel. +34 9 3542-1162 Fax +34 50 7191-2174 E-mail henrik.dobewall@gmx.net

Ó 2016 Hogrefe Publishing

Original Article

Personality of Clown Doctors An Exploratory Study Alberto Dionigi Department of Education, Cultural Heritage and Tourism, University of Macerata, Italy Abstract. In recent years, both professional and volunteer clowns have become familiar in health settings. The clown represents a peculiar humorist’s character, strictly associated with the performer’s own personality. In this study, the Big Five personality traits (BFI) of 155 Italian clown doctors (130 volunteers and 25 professionals) were compared to published data for the normal population. This study highlighted specific differences between clown doctors and the general population: Clown doctors showed higher agreeableness, conscientiousness, openness, and extraversion, as well as lower neuroticism compared to other people. Moreover, specific differences emerged comparing volunteers and professionals: Professional clowns showed significantly lower in agreeableness compared to their unpaid colleagues. The results are also discussed with reference to previous studies conducted on groups of humorists. Clowns’ personalities showed some peculiarities that can help to explain the facility for their performances in the health setting and that are different than those of other groups of humorists. Keywords: clown doctors, personality profile, Big Five, volunteers, professionals

Humor is an important part of life. Numerous studies have been conducted in order to understand why people laugh, and theories have been proposed to explain the mechanisms of humor (for a review, see Ruch, 2008). Research has mainly dealt with individual differences in the appreciation of humor by most people while the psychological characteristics of comic performers, such as cartoonists, humor writers, stand-up comedians, and clowns, have only marginally been investigated. In particular, studies have been conducted on humorists aimed at investigating whether they have particular personality traits that differ from those of the others and whether there are differences with amateur or less effective humor producers. Generally, a distinction in personality traits depending upon being a creator (e.g., writers) or an interpreter (e.g., actors) was highlighted (Kogan, 2002). The psychological research in this field has mainly focused on comedians’ personality, which revealed a specific pattern. In an early study, following a psychoanalytical approach based both on projective tests and analyses of dreams, comedians were found to be sad, depressive, despondent, and angry (Janus, 1975). When investigating gender differences, male comedians were more introverted than females, who were found to be vivacious, frenetic, and hypomanic (Janus, Bess, & Janus, 1978). When compared to the general population, both professional and amateur comedians showed significantly lower conscientiousness, extraversion, agreeableness, and higher openness compared to college students (Greengross & Miller, 2009). Interestingly, comedians and actors were also found to score higher in Psychoticism with respect to the rest of society’s norms (Ando, Claridge, & Clark, 2014). These findings are Ó 2016 Hogrefe Publishing

consistent with research conducted on male professional British cartoonists (Pearson, 1983), who were found to be high on Psychoticism suggesting a link between artistic creativity and Psychoticism. The psychological profile of professional comedians was also compared to clowns’ profiles in a Rorschach inkblot test (Fisher & Fisher, 1981). Comedians showed more references to themes of good and evil and a lower perception of self-worthiness and made a higher number of negative remarks about themselves. According to these results, an unusual personality structure may help to explain the facility for comedians’ performances and be helpful in making people laugh. As Psychoticism is closely related to creativity (Eysenck, 1995), the high P-scorer is more capable of coming up with unusual, incongruous, and wittier punch lines (Köhler & Ruch, 1996). Research on the personality of humorists has mainly focused on comedians, although clowns – as pranksters, jesters, jokers, harlequins, and mythologized tricksters – have been around for quite a long time and have attracted much popular interest. The clown is, by definition, a curious figure that projects itself on a different wavelength than that of societies’ status quo (Dionigi, Ruch, & Platt, 2014). The clown is a performer who acts foolishly and childishly in order to elicit positive emotions, and it is strictly associated with the performer’s own personality, physical body, and subjectivity (Lecoq, 2011). This personality is often also reflected in a specific costume, as adopting a costume frequently helps to construct and define a new role (Miller, Jasper, & Hill, 1991). Differently from an actor, a clown does not play a character and develops this based on the performer’s physical and psychological characteristics Journal of Individual Differences 2016; Vol. 37(1):49–55 DOI: 10.1027/1614-0001/a000187

A. Dionigi: Clown Doctor’s Personality

(Peacock, 2009). Clowns have appeared in a large variety of environments, and, in the last three decades, many more have been integrated into health settings. This practice was first started by Michael Christensen in 1986, a professional clown who set up the first Clown Care Unit (Dionigi, Flangini, & Gremigni, 2012). The art of clinic clowning is rooted both in the figures of fools and jesters and in shamans and folk healers. In many cultures, people believe that fools, clowns, and tricksters possess magical powers, enabling healing (Grinberg, Pendzik, Kowalsky, & Goshen, 2012). Although the figures of the circus clown and of the clown doctors are often linked, fundamental differences exist between them. These differences lie in the goal and methodology of clowning utilized in a health setting. The main aim of circus and street clowns is to entertain and amuse audiences, while the purpose of clown doctors is to assist patients in the healing process, distracting them from painful procedures, reducing anxiety triggered by hospitalization, and enhancing their moods (Dionigi, Sangiorgi, & Flangini, 2014; Grinberg et al., 2012; Pendzik & Raviv, 2011). To accomplish this goal, clown doctors must be bright performers and talented humorists, yet be empathetic and attentive to patients’ needs. Moreover, they must have high emotional intelligence as their purpose is to modify the moods of patients and change their emotional states to positive ones (Dionigi et al., 2012; Warren & Spitzer, 2013). Research has been carried out on a variety of other humorists and social groups, as yet no empirical studies have been done on the personality traits of clown doctors. Therefore, to further understand the effectiveness of these clowns in their dual roles as healers and humorists, it is of interest to establish what the typical psychological characteristics of clown doctors are. The purpose of this study is to investigate the personalities of a sample of clown doctors, using the five-factor model of personality (FFM) (McCrae & Costa, 2003). The hypothesis was that the personality profiles of clown doctors are significantly different from the general non-clown population. As performing artists, clowns show similarities in personality when compared to actors and comedians as clown doctors perform in front of crowds to elicit positive emotions. High extroversion is associated with the desire for being the center of social attention (Nettle, 2006), so clown doctors were expected to score high on extroversion. Clown doctors, similar to comedians, tune their acts to the crowd’s reactions (Greengross & Miller, 2009). They must also put the patients’ needs first and cooperate with colleagues, healthcare staff, and patients’ relatives. Therefore, clown doctors were predicted to score high on agreeableness. Previous studies have shown that creative people are low on conscientiousness and high on openness (Greengross & Miller, 2009; Nettle, 2006; Nowakowska, Strong, Santosa, Wang, & Ketter, 2005). Creativity is required in clown doctors as well; however, since they perform in peculiar settings, such as hospitals, they must manage their emotions and be self-disciplined. Clown doctors were thus expected to score high on both conscientiousness and openness. Journal of Individual Differences 2016; Vol. 37(1):49–55

Neurotic individuals are more prone to experience anxiety, worry about life events, and evaluate themselves more critically (Costa & McCrae, 1992). Furthermore, anxious individuals are more susceptible to focusing on potential threats in their environment (Eysenck & Derakshan, 2011). High neuroticism in performers leads to stronger stage fright (Steptoe et al., 1995). As clown doctors are required to stay in tune with patients and manage their and others’ emotions in order to perform well and accomplish their task, they were predicted to score low on neuroticism as compared to the general non-clown population. Finally, as an increasing number of people practice as clown doctors, they include a wide variety of professional backgrounds (Koller & Gryski, 2008). Currently, it is common to find both professional and volunteer clown doctors in hospitals (Dionigi, Ruch, & Platt, 2014), so differences in personality traits between volunteers and professionals were also investigated.

Aim of the Study At the time of this writing, there is no reported research about personalities of clowns working in health settings. The purpose of this study was to investigate the psychological characteristics of a sample of Italian clown doctors, both professionals and volunteers, and to compare findings with published normative data. Due to the characteristics required to work as a clown in health settings, it was expected that clown doctors would score higher in agreeableness, extraversion, conscientiousness, and openness and lower in neuroticism when compared to the normal population. Moreover, as friendliness and sociability go along with the need to help it was expected that volunteers score higher in agreeableness compared to professional clowns.

Method Participants The total sample consisted of 155 clown doctors (40 males, 115 females) varying in age between 21 and 66 years (M = 37.99, SD = 10.25). Participants were well-educated adults (0.6% primary school, 5.8% low secondary school, 41.9% upper secondary school, 52.3% university). With reference to marital status, 86 were not married, 51 were married or cohabiting, 15 were divorced, and three were widowed. Participants had different levels of experience in the art of clowning in health settings (M = 5.51 years, SD = 3.07, range = 0–13 years).

Instruments Participants completed a short demographic questionnaire and the Big Five Inventory (BFI; John & Srivastava, 1999) Ó 2016 Hogrefe Publishing

A. Dionigi: Clown Doctor’s Personality

to assess the Big Five dimensions of personality. The BFI consists of 44 short-phrase items, rated on a 5-point scale (1 = disagree strongly to 5 = agree strongly). The items assess the core traits that define each Big Five domain. The BFI items are assigned to five scales measuring Extraversion (E; 8 items), Agreeableness (A; 9 items), Conscientiousness (C; 9 items), Neuroticism (N; 8 items), and Openness to experience (O; 10 items). Prior researchers have presented adequate evidence of the reliability and validity of the BFI scales (John & Srivastava, 1999). In this study, the Italian version translated and validated by Ubbiali, Chiorri, Hampton, and Donati (2013) was used.

Procedures Data was collected in an online survey, and the study was conducted among a broad sample of Italian clown doctors. Participants were recruited via e-mails sent to the coordinators of Italian clown care units asking them to forward to clown doctors belonging to each unit. The e-mail contained a link to the survey created on Survey Monkey and an explanation of the aim of the study. All participants were guaranteed anonymity. The sample that fully completed the questionnaire was composed of 155 clown doctors. All statistical procedures were performed using the software package SPSS (21.0, IBM Inc., New York, NY, USA).

Results Comparison Between the Clown Doctor Sample and the General Population Comparison data for the FFM personality traits were taken from a large Italian general sample composed of 1,023

participants (female = 56.9%, M = 34.80, SD = 14.53, range 18–80 years old) recruited to study the psychometric properties of the Italian adaptation of the BFI (Ubbiali et al., 2013). The clown doctor scores on the scales showed good to acceptable reliabilities (Cronbach’s a: Openness to experience: .80; Conscientiousness: .79; Extraversion: .77; Agreeableness: .70; Neuroticism: .75). Comparisons between means, standard deviations of BFI traits, as well as correlations with age, and education are reported in Table 1. Each dimension of the FFM was compared to the values of the Italian population (Ubbiali et al., 2013) using a onesample t-test (Table 1). A significant difference in each concept of the Big Five was found (all p < .001), with a large effect size in O (d = 0.86) and medium effect size in A (d = 0.67), N (d = 0.66), E (d = 0.46), C (d = 0.42). Thus, the sample of clown doctors investigated can generally be described as more extraverted, agreeable, conscientious, open, and low in neuroticism compared to the general population. In the clown doctors’sample, a positive correlation (r = .19, p < .05) emerged between age and agreeableness while older clown doctors scored lower in neuroticism (r = .16, p < .05). No significant correlations were found between the five personality traits and education. An independent-samples t-test was also conducted to compare the scores on the Big Five dimensions of male and female clown doctors (Table 2). Female clowns scored higher in neuroticism compared to males (Mmales = 2.37, SD = 0.66; Mfemales = 2.69, SD = 0.64); t(153) = 2.76, p = .06, with a moderate effect size (d = 0.49), while no other significant gender-related differences emerged. In their study, Ubbiali et al. (2013) computed a point-biserial correlation, showing a significant relationship between genders of Italian population and specific dimensions of the Big Five: Females were found to be significantly higher in Agreeableness and Neuroticism compared to males.

Table 1. BFI scale means, standard deviations, associations with age and education of clown doctors and italian population Clown doctors

General population

Min Max M SD

2.25 4.88 3.63 0.61

2.44 5.00 4.09 0.51

1.89 5.00 3.90 0.60

1.00 4.50 2.61 0.66

2.50 5.00 4.18 0.51

t d

6.88*** 0.46

8.94*** 0.67

5.78*** 0.42

t-test values (df = 154) 9.17*** 12.64*** 0.66 0.86

Age Education

.06 .05

.19* .13

.11 .01

E 1.00 5.00 3.32 0.74

Associations with age and education .16* .08 .00 .05 .07 .01

1.89 5.00 3.72 0.59

1.33 5.00 3.62 0.74

1.00 5.00 3.09 0.78

1.00 5.00 3.67 0.67

.09** .03

.26** .10**

.18** .14**

.06* .08**

Notes. E = Extraversion; A = Agreeableness; C = Conscientiousness; N = Neuroticism; O = Openness; Min/Max = minimum/ maximum; M = Mean; SD = Standard Deviation; df = degree of freedom; d = Effect size; Number of Clown Doctors = 155 (%female = 74.20, M =37.99, SD = 10.25 range 21–66 years old). Source of general population data: Ubbiali et al. (2013), n = 1023 (%female = 56.9, M = 34.80, SD = 14.53, range 18–80 years old). *p < .05. **p < .01. ***p < .001. All tests were two-tailed. Ó 2016 Hogrefe Publishing

Journal of Individual Differences 2016; Vol. 37(1):49–55

A. Dionigi: Clown Doctor’s Personality

Table 2. Independent t-test for gender differences in five factor model personality traits in clown doctors Females

Extraversion Agreeableness Conscientiousness Neuroticism Openness

Males

3.64 4.09 3.91 2.69 4.19

0.62 0.49 0.58 0.64 0.51

3.70 4.08 3.88 2.37 4.16

0.58 0.57 0.66 0.66 0.49

Effect size

0.54 0.18 0.26 2.76** 0.29

.10 .02 .05 .49 .06

Notes. Number of female clown doctors = 115; number of male clown doctors = 40; Effect size = d. **p < 0.01 (two-tailed).

Differences Between Volunteers and Professionals In order to assess differences between clowns, two subgroups were established: Volunteer and professional clown doctors. The volunteer group was composed of participants whose clown activity does not represent their main job and who are not paid for it. The professional group includes participants who reported to do this activity as a job and who receive money for it. The volunteer group was composed of 130 clown doctors (31 males and 99 females; M = 38.15, SD = 10.61, range = 21–66 years old) while the professional group included 25 people (9 males and 16 females; M = 37.16; SD = 8.28; range = 27–62 years old). In the next step, the Big Five dimensions of volunteer and professional clowns were compared. Due to the small sample of professional clowns, the assumption of normality of all five dimensions of the Big Five was examined using normal probability plots. No apparent deviations from normality emerged. Following, Levene’s homogeneity of variance test was conducted: The variances of the Big Five dimensions of the two groups were not different from each other (Extraversion: F(1, 153) = 0.32, p = .57; Agreeableness: F(1, 153) = 0.26, p = .61; Conscientiousness: F(1, 153) = 1.61, p = .21; Neuroticism: F(1, 153) = 3.19, p = .77; Openness to experience: F(1, 153) = 0.95, p = .33). Therefore, it was safe to continue with the ANOVA, controlling for both age and gender. Results are presented in Table 3. Table 3 shows the comparisons among volunteer and professional clown doctors groups based on the Big Five personality. An ANCOVA [between-subjects factor: professionalism (volunteer, professional); covariates: age

and gender] revealed a significant group difference for Agreeableness as volunteer clowns scored higher than professionals, F(1, 151) = 6.29, p < .01, and a moderate effect size ( f = .57). ANCOVA revealed no main effects of gender, F(1, 151) = .02, p = .88, and a main effect of age, F(1, 151) = 5.67, p < .05.

Discussion This is the first study aimed at investigating personality characteristics of clown doctors; the data for this study shows that this sample of clown doctors showed peculiar characteristics that distinguish them from the general Italian non-clown population. Clown doctors score higher in Extraversion, Agreeableness, Conscientiousness, and Openness to experience, and lower in Neuroticism compared to published normative data of the populace. These results confirm what was suggested by theoretical manuscripts written in this area (Dionigi et al., 2012; Warren & Spitzer, 2013). The clown doctor is a comic character required to be creative in order to improvise according to what is found in the healthcare setting. Extraversion is related to the fluency component of creativity (Eysenck, 1995) and humor, creativity, and extraversion have been found to be positively correlated (Koppel & Sechrest, 1970). Extraverts are characterized by traits such as being sociable, lively, active, assertive, tending to enjoy human interactions, and being enthusiastic, talkative, assertive, and gregarious. In the field of humor, extraverts (as compared to introverts) are more cheerful, less serious, and able to produce a higher quantity

Table 3. Comparisons between volunteer and professional clown doctors Volunteer clowns

Extraversion Agreeableness Conscientiousness Neuroticism Openness

Professional clowns

Effect size

3.67 4.13 3.94 2.60 4.18

0.61 0.51 0.61 0.68 0.51

3.60 3.85 3.71 2.62 4.21

0.58 0.48 0.51 0.50 0.49

0.53 6.29** 2.83 0.10 0.16

.12 .57 .41 .03 .06

Notes. Number of volunteers = 130; number of professionals = 25; Effect size = f. **p < 0.01 (two-tailed). Journal of Individual Differences 2016; Vol. 37(1):49–55

Ó 2016 Hogrefe Publishing

A. Dionigi: Clown Doctor’s Personality

of humor content (Köhler & Ruch, 1996). As clowns expose themselves with the intention of being the laughing stock for others, probably extraversion may provide responsiveness toward the interpersonal reward of being the center of an audience’s attention (Nettle, 2006). Clown doctors show high Agreeableness compared to other people. This dimension is related to the ability to be sensitive toward others’ needs, compassionate, and cooperative. Moreover, other public figures who want to be loved and appreciated, such as actors and politicians, are high on this dimension (Caprara, Barbaranelli, Consiglio, Picconi, & Zimbardo, 2003; Nettle, 2006). Agreeableness represents the core dispositional trait contributing to prosocial behaviors as agreeable individuals are altruistic, straightforward, trusting, softhearted, modest, and compliant (McCrae & Costa, 2003). Researchers found significant positive relations between agreeableness and volunteering, as people scoring high on Agreeableness are more likely to comply with requests from others (Carlo, Okun, Knight, & de Guzman, 2005). In the present study, volunteers scored higher than professionals in agreeableness. One possible explanation can be that volunteers are more altruistic, modest, and compliant compared to their paid colleagues (Liao-Troth, 2005). The sample of clown doctors was also high on Conscientiousness compared to normative data. Clowns are required to attend rigorous training before entering the health setting and must be aware of their role as well as of the risks related to a wrong approach to patients or to a lack of competence. As conscientious individuals are able to manage emotions, be self-disciplined, and focus on goal achievement, it is not surprising that clown doctors score high on this trait. As predicted, clown doctors have been found to be more open to experience than the general population. This finding is consistent with those of previous studies that showed that other creative groups are also high on Openness to experience (Nettle, 2006). Clown doctors are required to be curious and possess a high imagination as they must look at the world in a different, unconventional, and imaginative way where (almost) anything is possible and permitted (Dionigi, Ruch, & Platt, 2014). Furthermore, Openness to experience has consistently been positively related to creativity in general, and clown doctors need to continually adjust their way to approach patients in order to make their work evolve. Finally, clown doctors scored low in Neuroticism. This result is not surprising, as neuroticism is made up of traits like anxious, depressed, guilt feelings, low self-esteem, irrational, and shyness; that is not representative of the clown’s profile. Moreover, neuroticism was found to be negatively correlated with cheerfulness (Köhler & Ruch, 1996). Studies on other creative people like poets and writers show that they are high in this dimension. However, differently from clown doctors, they do not have to perform their creation on stage (Nowakowska et al., 2005). Research conducted on comedians showed no differences in neuroticism compared with the rest of society (Greengross & Miller, 2009). This finding is consistent with what has been recently found about the ability to remain in the role during Ó 2016 Hogrefe Publishing

the performance. Clown doctors who were better able to differentiate between their persona and their clown persona were less interfered by external cues and internal thoughts; they experienced less anxiety before and during the activity (Dionigi, Ruch, & Platt, 2014). The results of the present study also shed light on how groups of humorists may differ. A cheerful temperament is involved in the sense of humor and it goes along with high extraversion, emotional stability, agreeableness, and openness, and low conscientiousness (Ruch & Köhler, 1998). Compared to this, clowns are relatively similar with the exception that they are higher (and not lower) in conscientiousness. In particular, volunteers are high on both agreeableness and conscientiousness. Clowns thus seem to be very different though from the witty professionals and witty amateurs. Cartoonists (Pearson, 1983) and those able to provide funnier punch lines are high in P (Köhler & Ruch, 1996) and stand-up comedians are higher in schizotypy (Ando et al., 2014). Likewise, stand-up comedians in Greengross and Miller’s (2009) study were significantly lower in both Conscientiousness and Agreeableness. High P is an indicator of low Agreeableness and low Conscientiousness, and here we can assume that witty people and clowns are on the opposite poles of the Psychoticism dimensions. However, clowns (in this study), comedians (Greengross, Martin, & Miller, 2012; Greengross & Miller, 2009), and witty people share higher openness to experience. In their study, Ruch and Köhler (1998) showed that the ability to produce witty punch lines goes along with high openness and low seriousness. The clowns in the present study were higher in Extraversion and lower in Neuroticism, which is again different from the witty. Comedians were lower in Extraversion (Ando et al., 2014; Greengross & Miller, 2009; Greengross et al., 2012). The results of this study showed that the sample of clown doctors examined possess specific personality characteristics in comparison to Italian non-clown population. Significant differences were found between volunteers and professionals. Although these findings shed light on a topic that previous authors have only addressed theoretically, some limitations to extending the findings to other clown doctors need to be acknowledged. The study was conducted on a sample of Italian clown doctors, and further research is needed to confirm these results for other cultures and nationalities. The relatively small sample of professional clown doctors further limits the findings, and replications on larger samples are required. A comparison was not possible with published normative data where study samples had been separated into subgroups by age and gender, which prevented verification of personality differences using these criteria. Future studies will include a control group, instead of comparing results with published data, in order to address this concern, as well as to investigate possible differences in personalities of clinic versus circus clowns. Finally, as personality is a complex phenomenon, further research should explore specific personality traits that can play a role in clinic clowning, such as narcissism and emotional intelligence. Narcissists support their self-image through feedback and admiration received from others (Atlas & Them, 2008), which may lead to feelings of Journal of Individual Differences 2016; Vol. 37(1):49–55

A. Dionigi: Clown Doctor’s Personality

anxiety after perceived negative feedback. These individuals may fail to develop appropriate strategies for dealing with intense emotions. On the other hand, emotional intelligence has been indicated as important for clown doctors, since individuals scoring high on emotional intelligence are better able to use, manage, understand, and pay attention to emotions (Mayer, Salovey, & Caruso, 2004). Acknowledgments The study was conducted during the research stay of the author at the University of Zürich: the author is grateful to Willibald Ruch, Jennifer Hofmann and Tracey Platt for the precious help and assistance received. Many thanks to Carlo Chiorri, Chiara Manfredi and Giulia Casu for the helpful hints and comments.

References Ando, V., Claridge, G., & Clark, K. (2014). Psychotic traits in comedians. The British Journal of Psychiatry, 204, 341–345. doi: bjp-bp.113.134569 Atlas, G. D., & Them, M. A. (2008). Narcissism and sensitivity to criticism: A preliminary investigation. Current Psychology, 27, 62–76. doi: 10.1007/s12144-008-9023-0 Caprara, G. V., Barbaranelli, C., Consiglio, C., Picconi, L., & Zimbardo, P. G. (2003). Personalities of politicians and voters: Unique and synergistic relationships. Journal of Personality and Social Psychology, 84, 849. doi: 10.1037/ 0022-3514.84.4.849 Carlo, G., Okun, M. A., Knight, G. P., & de Guzman, M. R. T. (2005). The interplay of traits and motives on volunteering: Agreeableness, extraversion and prosocial value motivation. Personality and Individual Differences, 38, 1293–1305. doi: 10.1016/j.paid.2004.08.012 Costa, P. T. Jr., & McCrae, R. R. (1992). Revised NEO personality inventory (NEO-PI-R and NEO five-factor inventory (NEO-FFI) professional manual. Odessa, FL: Psychological Assessment Resources. Dionigi, A., Flangini, R., & Gremigni, P. (2012). Clowns in hospitals. In P. Gremigni (Ed.), Humor and health promotion (pp. 213–227). New York, NY: Nova Science. Dionigi, A., Ruch, W., & Platt, T. (2014). Components and determinants of the shift between own persona and the clown persona: A hierarchical analysis. European Journal of Humour Research, 1, 58–80. Dionigi, A., Sangiorgi, D., & Flangini, R. (2014). Clown intervention to reduce preoperative anxiety in children and parents: A randomized controlled trial. Journal of Health Psychology, 19, 369–380. doi: 10.1177/1359105312471567 Eysenck, H. J. (1995). Creativity as a product of intelligence and personality. In D. H. Saklofske & M. Zeidner (Eds.), International handbook of personality and intelligence (pp. 231–247). New York, NY: Plenum Press. Eysenck, M. W., & Derakshan, N. (2011). New perspectives in attentional control theory. Personality and Individual Differences, 50, 955–960. doi: 10.1016/j.paid.2010.08.019 Fisher, S., & Fisher, R. L. (1981). Pretend the world is funny and forever: A psychological analysis of comedians, clowns, and actors. Hillsdale, NJ: Erlbaum. Greengross, G., Martin, R. A., & Miller, G. F. (2012). Personality traits, intelligence, humor styles, and humor production

Journal of Individual Differences 2016; Vol. 37(1):49–55

ability of professional stand-up comedians compared to college students. Psychology of Aesthetics, Creativity and the Arts, 6, 74–82. doi: 10.1037/a0025774 Greengross, G., & Miller, G. F. (2009). The Big Five personality traits of professional comedians compared to amateur comedians, comedy writers, and college students. Personality and Individual Differences, 47, 79–83. doi: 10.1016/ j.paid.2009.01.045 Grinberg, Z., Pendzik, S., Kowalsky, R., & Goshen, Y. (2012). Drama therapy role theory as a context for understanding medical clowning. The Arts in Psychotherapy, 39, 42–51. doi: 10.1016/j.aip.2011.08.005 Janus, S. S. (1975). The great comedians: Personality and other factors. American Journal of Psychoanalysis, 35, 169–174. doi: 10.1007/BF01358189 Janus, S. S., Bess, B. E., & Janus, B. R. (1978). The great comediennes: Personality and other factors. The American Journal of Psychoanalysis, 38, 367–372. doi: 10.1007/ BF01253595 John, O. P., & Srivastava, S. (1999). The big five trait taxonomy: History, measurement, and theoretical perspectives. In L. A. Pervin & O. P. John (Eds.), Handbook of personality: Theory and research (pp. 102–138). New York, NY: Guilford Press. Kogan, N. (2002). Careers in the performing arts: A psychological perspective. Creativity Research Journal, 14, 1–16. doi: 10.1207/S15326934CRJ1401_1 Köhler, G., & Ruch, W. (1996). Sources of variance in current sense of humor inventories: How much substance, how much method variance? Humor, 9, 363–398. doi: 10.1515/ humr.1996.9.3-4.363 Koller, D., & Gryski, C. (2008). The life threatened child and the life enhancing clown: Towards a model of therapeutic clowning. Evidence-Based Complementary and Alternative Medicine, 5, 17–25. doi: 10.1093/ecam/nem033 Koppel, M. A., & Sechrest, L. (1970). A multitrait-multimethod matrix analysis of sense of humor. Educational and Psychological Measurement, 30, 77–85. doi: 10.3102/ 00346543046003407 Lecoq, J. (2011). The moving body (Le corps poetique): Teaching creative theatre. London, UK: Methuen. Liao-Troth, M. A. (2005). Are they here for the long haul? The effects of functional motives and personality factors on the psychological contracts of volunteers. Nonprofit and Voluntary Sector Quarterly, 34, 510–530. doi: 10.1177/ 0899764005279513 Mayer, J. D., Salovey, P., & Caruso, D. R. (2004). TARGET ARTICLES: ‘‘Emotional intelligence: Theory, findings, and implications’’. Psychological Inquiry, 15, 197–215. doi: 10.1207/s15327965pli1503_02 McCrae, R. R., & Costa, P. T. Jr. (2003). Personality in adulthood: A five-factor theory perspective. New York, NY: Guilford Press. Miller, K. A., Jasper, C. R., & Hill, D. R. (1991). Costume and the perception of identity and role. Perceptual and Motor Skills, 72, 807–813. Nettle, D. (2006). Psychological profiles of professional actors. Personality and Individual Differences, 40, 375–383. doi: 10.1016/j.paid.2005.07.008 Nowakowska, C., Strong, C. M., Santosa, C. M., Wang, P. W., & Ketter, T. A. (2005). Temperamental commonalities and differences in euthymic mood disorder patients, creative controls, and healthy controls. Journal of Affective Disorders, 85, 207–215. doi: 10.1016/j.jad.2003.11.012 Peacock, L. (2009). Serious play: Modern clown performance. Bristol, UK: Intellect Books. Pearson, P. (1983). Personality characteristics of cartoonists. Personality and Individual Differences, 4, 227–228. doi: 10.1016/0191-8869(83)90030-2

Ó 2016 Hogrefe Publishing

A. Dionigi: Clown Doctor’s Personality

Pendzik, S., & Raviv, A. (2011). Therapeutic clowning and drama therapy: A family resemblance. The Arts in Psychotherapy, 38, 267–275. doi: 10.1016/j.aip.2011.08.005 Ruch, W. (2008). Psychology of humor. In V. Raskin (Ed.), The primer of humor research (pp. 17–100). Berlin, Germany: Mouton de Gruyter. Ruch, W., & Köhler, G. (1998). A temperament approach to humor. In W. Ruch (Ed.), The sense of humor: Explorations of a personality characteristic (pp. 203–230). Berlin, Germany: Mouton de Gruyter. Steptoe, A., Malik, F., Pay, C., Pearson, P., Price, C., & Win, Z. (1995). The impact of stage fright on student actors. British Journal of Psychology, 86, 27–39. doi: 10.1111/j.20448295.1995.tb02544.x Ubbiali, A., Chiorri, C., Hampton, P., & Donati, D. (2013). Psychometric properties of the Italian adaptation of the Big Five Inventory (BFI). Bollettino di Psicologia Applicata, 266, 37–48. Warren, B., & Spitzer, P. (2013). Smiles are everywhere: Integrating clown-play into healthcare practice. London, UK: Routledge.

Ó 2016 Hogrefe Publishing

Date of acceptance: May 12, 2015 Published online: February 29, 2016

Alberto Dionigi Department of Education Cultural Heritage and Tourism University of Macerata Piazzale Luigi Bertelli (Contrada Vallebona) 62100 Macerata Italy Tel. +39 05 4169-7122 Fax +39 05 4160-0386 E-mail alberto.dionigi@unimc.it

Journal of Individual Differences 2016; Vol. 37(1):49–55

Original Article

Validation and Revision of a German Version of the Balanced Measure of Psychological Needs Scale Andreas B. Neubauer and Andreas Voss Institute of Psychology, University of Heidelberg, Germany Abstract. In this study we tested the psychometric properties of a German version of the Balanced Measure of Psychological Needs scale (BMNP; Sheldon & Hilpert, 2012), a questionnaire to assess the degree of fulfillment of the three basic psychological needs for autonomy, competence, and relatedness. In Study 1, 251 participants completed this questionnaire, as well as measures of life satisfaction, self-esteem, depression, loneliness, and personality traits via online assessment. Results indicate that a six-dimensional structure fit the data adequately well. Furthermore, all three needs independently predicted life satisfaction and depression over and above personality traits. Self-esteem was only predicted by relatedness satisfaction and competence dissatisfaction, and loneliness was only predicted by relatedness. In Study 2, we revised the BMPN by replacing one item and largely replicated the results obtained in Study 1. Study 3 showed that the subscales of the BMPN are only moderately stable over 1 week supporting the assumption of the BMPN being a state measure. Together, these results suggest that the revised German version of the BMPN is a reliable and valid measure to assess satisfaction and dissatisfaction of the psychological needs for autonomy, competence, and relatedness. Keywords: self-determination, well-being, need fulfillment, confirmatory factor analysis

Self-Determination Theory (SDT; Deci & Ryan, 1985, 2000) has become one of the most influential theoretical frameworks for studying motivation and social psychological processes at large. It has also been widely applied in the area of research on subjective well-being (SWB; Ryan & Deci, 2001). The core premise of SDT is that there are three fundamental human needs: the need for autonomy, for competence, and for relatedness. For optimal psychological functioning (including SWB), all three needs must be fulfilled. While there is ample evidence for this proposition, there is some disagreement concerning the operationalization of need fulfillment (Johnston & Finney, 2010; Sheldon & Hilpert, 2012). The aim of the present study was to test the psychometric properties of the German version of one operationalization of need fulfillment, the Balanced Measure of Psychological Needs scale (BMPN; Sheldon & Hilpert, 2012).

Self-Determination Theory as a Framework for Subjective Well-Being SDT proposes that there are three innate and basic psychological needs which must be fulfilled for optimal psychological functioning: (1) The need for autonomy refers to the feeling of volition and freedom in one’s actions. (2) Competence relates to the feeling of being effective Journal of Individual Differences 2016; Vol. 37(1):56–72 DOI: 10.1027/1614-0001/a000188

in one’s actions and in mastering one’s environment. (3) Relatedness refers to feeling close to and cared for by other human beings. Prior research has shown that fulfillment of all three needs is associated with increased wellbeing and reduced ill-being such as depression and loneliness (e.g., Vansteenkiste, Lens, Soenens, & Luyckx, 2006; Wei, Shaffer, Young, & Zakalik, 2005). Despite these positive findings, the diversity of operationalizations used to measure need fulfillment complicates their integration. For example, in two daily-diary studies (Reis, Sheldon, Gable, Roscoe, & Ryan, 2000; Sheldon, Ryan, & Reis, 1996) participants were instructed to recall the three activities they spent the most time doing at this day and answer several questions about these activities. In this context, relatedness and competence were assessed using one item each and autonomy using four items. Although daily competence and relatedness were averaged across the ratings for all three activities the participants reported, it is unclear whether one item (‘‘How effective did you feel in doing this activity?’’/‘‘How close and connected did you feel with the people you were with?’’) can fully capture the fulfillments of the needs for competence and relatedness. Moreover, using this one-item approach comes at the expense of unknown – but probably low – reliability of the measurement. Gagné (2003) developed a measure called the Basic Psychological Needs Scale (BPNS) assessing the degree to which the three basic psychological needs are fulfilled using Ó 2016 Hogrefe Publishing

A. B. Neubauer & A. Voss: Validation of the German BMPN

several items for each need. Although this instrument expands the rather narrow approach of single item measures, this particular scale comes with another limitation: It uses a different number of items to assess the three needs (seven items assessing autonomy, six items assessing competence, eight items assessing relatedness), which might give some needs more weight than others: By assessing the three dimensions with a different number of items, the dimension with more items (relatedness, eight items) might be assessed with higher reliability than, for example, competence (six items). Therefore, when competence and relatedness are competing for predictive validity (as in multiple regression), relatedness might have an advantage due to its higher reliability. Additionally, there is no clear indication whether the three needs should be interpreted separately or whether they should (and can) be combined in one overall need satisfaction score. Moreover, this scale showed an unsatisfactory factorial structure in confirmatory factor analyses (Johnston & Finney, 2010; Sheldon & Hilpert, 2012). In response to these limitations Sheldon and Hilpert (2012) developed the Balanced Measure of Psychological Needs (BMPN) scale, which assesses the three needs with six items each and has proven both good internal consistency and factorial structure. In the BMPN, three of the six items assessing each need are worded positively, indicating need satisfaction, while the other items are worded negatively, thus indicating need dissatisfaction. This is important since prior research showed that in some instances the effects of need satisfaction and dissatisfaction, respectively, cannot be considered as mere opposites of one another: For example, Sheldon, Abad, and Hinsch (2011) reported positive correlations of Facebook use with both relatedness satisfaction and relatedness dissatisfaction. This pattern of results would be a paradox if need satisfaction and need dissatisfaction were psychometric opposites. The authors showed longitudinally that relatedness dissatisfaction (but not lack of relatedness satisfaction) promotes Facebook use, and Facebook use in turn increases relatedness satisfaction (but does not decrease relatedness dissatisfaction). Thus, they assume that satisfaction and dissatisfaction operate at different time points. Hence, a scale that assesses need satisfaction and need dissatisfaction for all three needs separately would be a great improvement over previous measures. The one-item measures used in early work (Reis et al., 2000; Sheldon et al., 1996) cannot accomplish this. The BMPN, on the other hand, can be used to assess either overall fulfillment of the three needs, or need satisfaction and dissatisfaction separately (Sheldon & Hilpert, 2012) which makes it a very flexible tool to measure fulfillment of the basic psychological needs postulated in SDT. Sheldon and Hilpert (2012) recommend using the BPMN to assess the three needs separately instead of combining them into one overall need score. From their results, one could, however, also argue that the six subscales (autonomy satisfaction, autonomy dissatisfaction, competence satisfaction, competence dissatisfaction,

relatedness satisfaction, relatedness dissatisfaction) should not be combined into three need scores (autonomy, competence, relatedness). The argument in favor of the threedimensional structure is based on the finding that the fit of the measurement model improved when two latent ‘‘method’’ factors (a dissatisfaction factor and a satisfaction factor) were included. However, inspection of the factor loadings of these method factors revealed that the satisfaction factor was almost exclusively built by the three competence satisfaction items, while the dissatisfaction factor was more strongly related to the autonomy dissatisfaction items. In other words: the ‘‘method’’ effects were stronger for competence and autonomy than they were for relatedness. This hinders interpretability of the three subscales. The subscales can only be interpreted properly if the method artifacts are contained in all three scales to an equal degree (i.e., if the factor loadings for the method factors are constrained to equality). Such a model was, however, not tested by these authors. Furthermore, Sheldon and Hilpert (2012) do not report data on a six-factor model which would be an alternative model in light of the findings on a dissociation of the effects of need satisfaction and need dissatisfaction on behavior (Sheldon et al., 2011; Sheldon & Gunz, 2009).

The Present Research The aims of the present studies were as follows: First, we wanted to test whether we could replicate the factor structure obtained by Sheldon and Hilpert (2012) for a German version of the BMPN. It is expected that the best fitting model reported by these authors (three correlated needs and two uncorrelated methods) will adequately fit the data of our sample. We are not aware of any previous attempts to validate a German scale assessing need fulfillment and are aiming at filling this gap in the literature. Second, we aimed at testing alternative measurement models to ease the interpretation of the subscale scores. We will reduce the correlated trait uncorrelated method model reported by Sheldon and Hilpert (2012) to a correlated trait correlated (method-1) model.1 This model has been recommended for structurally different methods in the multitrait multimethod framework (Eid et al., 2008). Furthermore, we will test an alternative model with six correlated factors. Third, we will validate the BMPN by investigating the construct validity of the scale by inspecting relations to indicators of well-being (life satisfaction and self-esteem) and ill-being (loneliness and depression). Life satisfaction, the cognitive component of subjective well-being (SWB; Diener, 1984; Ryan & Deci, 2001), is a subjective evaluation of one’s life circumstances. Self-esteem, a positive evaluation of one’s self, has been discussed as indicator for good psychological adjustment (DeWall et al., 2011) and should therefore be predicted by need fulfillment. Although other theoretical accounts (Greenberg, Pyszczynski, & Solomon, 1986; Williams, 2001) consider self-esteem to be a psychological need

Please note that a (method-1) model leaves only one method factor in our example. Hence, there were no ‘‘correlated’’ methods and we will further use the term correlated trait (method-1) model.

Ó 2016 Hogrefe Publishing

Journal of Individual Differences 2016; Vol. 37(1):56–72

A. B. Neubauer & A. Voss: Validation of the German BMPN

itself, SDT considers it an outcome of need fulfillment and therefore secondary to fulfillment of the basic needs for autonomy, competence, and relatedness (Ryan & Brown, 2003). Therefore, life satisfaction and self-esteem are both used as outcomes of fulfillment of the basic needs postulated by SDT. We expect that all three needs independently predict levels of life satisfaction and self-esteem. While need fulfillment is positively associated with psychological well-being, it should be negatively associated with psychological malfunctioning. Depression is often used as a marker for psychological malfunctioning (e.g., DeWall et al., 2011). It is therefore expected that depression is predicted independently by all three needs. Loneliness, on the other hand, should be tied more specifically to lack of relatedness. It is therefore expected that only relatedness, but not competence or autonomy uniquely predicts loneliness. Such a pattern could be taken as evidence for discriminant validity of the three needs. Fourth and finally, we will examine the stability of the BPMN over the course of 1 week. Since the BMPN is supposed to measure state need satisfaction and dissatisfaction, test-retest correlations are expected to be modest only, representing only little stability of the measurements.

Study 1 Method Sample and Procedure

(Sheldon & Hilpert, 2012). This questionnaire consists of 18 items which measure satisfaction and dissatisfaction of the three needs autonomy, competence, and relatedness. Participants were instructed to indicate on a 7-point Likert scale to what degree each statement applies to them with respect to their last month (ranging from ‘‘not at all’’ to ‘‘completely’’). The items were translated to German by the first author of this work and back translated by a German native speaker fluent in English. Inconsistencies were resolved in a final discussion of the translation. The order of items was alternated from the three subscales. German wording for all items is presented in the Appendix. Sheldon and Hilpert (2012) argue that need fulfillment can be computed by either building three subscales (autonomy, competence, and relatedness) consisting of six items each, or by building six subscales (autonomy satisfaction, autonomy dissatisfaction, competence satisfaction, competence dissatisfaction, relatedness satisfaction, and relatedness dissatisfaction) consisting of three items each. Internal consistencies (Cronbach’s a) were therefore computed for both the three subscales version and the six subscales version. Combined across satisfaction and dissatisfaction (three subscales solution), internal consistencies were satisfactory for all three needs, a = .75 (autonomy), a = .77 (competence), and a = .78 (relatedness). For the six subscales solution, internal consistency was higher for the satisfaction subscales, a = .72 (autonomy satisfaction), a = .85 (competence satisfaction), and a = .85 (relatedness satisfaction), than for the dissatisfaction subscales, a = .66 (autonomy dissatisfaction), a = .75 (competence dissatisfaction), and a = .67 (relatedness dissatisfaction). These reliability estimates are similar to the ones reported for the original scale (Sheldon & Hilpert, 2012).

Data was collected using an online questionnaire. The link to this questionnaire was posted on the homepages of ‘‘Psychologie heute’’ and ‘‘Forschung erleben.’’2 These two websites provide information on current psychological research for interested laypersons. Additionally, the link was sent to approximately 740 members of a mailing list on ‘‘Forschung erleben’’: Visitors of ‘‘Forschung erleben’’ could subscribe to this mailing list, if they were interested in taking part in any sort of online surveys related to social psychological research. Finally, the link was also distributed via word-of-mouth recommendation. All in all, 323 people clicked on the link for the questionnaire, and 251 participants (Mage = 26.2 years, SD = 7.3, range = 14–59; 78% female) filled in the questionnaire (all completed at least 95% of the questions). All basic analyses were performed using the statistical software R (R Core Team, 2015). Confirmatory factor analyses were computed using Mplus Version 7 (Muthén & Muthén, 2015).

A short form of the Big Five Inventory (Rammstedt & John, 2005) was administered. This scale measures the ‘‘Big Five’’ (Extraversion, Neuroticism, Agreeableness, Conscientiousness, Openness) with 4–5 items per dimension. Participants had to indicate on a 5-point Likert scale to what extent they agree with each statement presented (ranging from ‘‘completely disagree’’ to ‘‘completely agree’’). Internal consistencies for the five subscales were a = .84 (Extraversion), a = .79 (Neuroticism), a = .66 (Agreeableness), a = .74 (Conscientiousness), and a = .79 (Openness).

Measurements

Life Satisfaction

Need Fulfillment

The German version (Glaesmer, Grande, Braehler, & Roth, 2011) of the Satisfaction with Life Scale (SWLS; Diener, Emmons, Larsen, & Griffin, 1985) was used to assess life satisfaction. This scale consists of five items; participants

Need fulfillment was assessed using a German translation of the Balanced Measure of Psychological Needs scale 2

Personality

http://www.psychologie-heute.de/home/ and http://www.forschung-erleben.uni-mannheim.de/

Journal of Individual Differences 2016; Vol. 37(1):56–72

Ó 2016 Hogrefe Publishing

A. B. Neubauer & A. Voss: Validation of the German BMPN

were asked to indicate to what extent they agree with each of the five statements (e.g., ‘‘I am satisfied with my life’’) on a scale ranging from 1 (= not at all) to 7 (= completely). Internal consistency for this measure was high, a = .87.

competence, and relatedness as assessed via the BMPN are related yet different constructs. Further information on descriptive statistics and intercorrelations of the study variables can be found in Table 1, lower diagonal.

Loneliness

Factor Structure of the BMPN

A short 8-item version (Hays & DiMatteo, 1987) of the UCLA Loneliness scale (Russell, Peplau, & Cutrona, 1980) was administered in this study. This scale has shown good psychometric properties (Hays & DiMatteo, 1987; Wu & Yao, 2008). For this study, we used the German translation by Döring and Bortz (1993), but included only the eight items suggested by Hays and DiMatteo (1987). Participants were asked to what extent they agree with each of the statements (e.g., ‘‘I feel left out’’); answers were given on a scale ranging from 1 (= not at all) to 7 (= completely). Internal consistency of this scale was good, a = .87.

The factor structure of the BMPN was analyzed using confirmatory factor analysis (CFA). All tested models are schematically depicted in Figure 1. Specifically, we started with a correlated trait uncorrelated method model (Model 0) which has been reported by Sheldon and Hilpert (2012) as the best fitting model for the BMPN. In a second model (Model 1), we removed one of the method factors (the satisfaction factor) and, hence, arrived at a correlated trait (method-1) model (Eid et al., 2008). It should be noted that this procedure makes the satisfaction component the reference method, which affects parameter estimates as well as model fit and should be considered when interpreting this model (Geiser, Eid, & Nussbeck, 2008). Allowing the loadings of the method factor to vary across the three needs complicates the interpretation of the three subscale measures. Therefore, in the next model (Model 1a, not shown in the figure) the nine loadings of the latent dissatisfaction factor were constrained to be equal. In a last model (Model 2), we empirically tested the claim that the BMPN can be used to assess the six postulated subscales. For all models, the variances of the latent variables were fixed to 1 and all factor loadings were estimated freely (except for the method loadings in Model 1a). No other model constraints were imposed. Models 0, 1, and 1a are nested models, but Model 2 is not nested in the other models. Hence nested model comparisons using chi-square difference tests were performed for the first three models only. Additionally, model fit was determined by several fit indices: The comparative fit index (CFI), the root mean square error of approximation (RMSEA), and the standardized root mean square residual (SRMR). For these fit indices, we applied the conventional cut-off criteria of .90 or higher (CFI) and .08 or less (RMSEA and SRMR) as indication of acceptable model fit. Additionally both the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC) are reported. The information criteria are used for comparisons of the non-nested models, with smaller values indicating better model fit. The robust maximum likelihood estimator (MLR) was used to account for possible violation of the assumption of multivariate normality of the indicators. Therefore, chi-square difference tests for nested models have to be adjusted by a scaling correction factor (Yuan & Bentler, 2000). Previous research (Rhemtulla, Brosseau-Liard, & Savalei, 2012) shows that the robust maximum likelihood estimator performs well with items measured on a 7-point Likert scale.3 Fit indices for all models are reported in Table 2 (upper panel). As expected, the correlated traits uncorrelated

Self-Esteem The Rosenberg Self-Esteem Scale (R-SES; Rosenberg, 1965; for the German version see Ferring & Filipp, 1996) was also administered, which consists of 10 items. Participants had to indicate on a 5-point Likert scale to what extent they agree with each statement (ranging from completely disagree to completely agree). Cronbach’s a in this sample was .91. Depression Depressive symptoms were assessed using the Center for Epidemiological Studies Depression Scale (CES-D; Radloff, 1977; for the German version see Hautzinger, 1988). This scale consists of 20 items inquiring about the respondent’s past week. For each of the statements (e.g., ‘‘During the past week, everything I did was an effort’’), participants were instructed to indicate how often this statement had applied to them in the past week. Response categories ranged from 0 (= rarely or none of the time) to 3 (= most or all of the time). Internal consistency for this scale was a = .92.

Results The three needs were positively correlated (autonomy and competence: r = .53; autonomy and relatedness: r = .52; competence and relatedness: r = .51), all p < .001. These correlations are similar in magnitude to the estimates reported for the US-version (which were between .46 and .49; Sheldon & Hilpert, 2012) and indicate that autonomy, 3

Applying a categorical estimation method (the mean-and variance adjusted unweighted least squares estimator, ULSMV) did not change the conclusions drawn from the MLR results. Hence, will report the results of the latter approach only.

Ó 2016 Hogrefe Publishing

Journal of Individual Differences 2016; Vol. 37(1):56–72

Age Gendera Autonomy Autonomy Sat Autonomy Dis Competence Competence Sat Competence Dis Relatedness Relatedness Sat Relatedness Dis Extraversion Neuroticism Agreeableness Conscientiousness Openness Life satisfaction Self-esteem Loneliness CES-D

26.2 7.3

– .01 .02 .01 .04 .01 .01 .03 .00 .01 .01 .10 .17 .08 .23 .18 .06 .03 .02 .05

4.3 1.2

.06 .05 – .83 .87 .53 .30 .57 .52 .48 .41 .31 .51 .21 .28 .04 .57 .54 .52 .56

.06 – .07 .06 .06 .09 .04 .19 .16 .15 .13 .02 .21 .03 .15 .07 .00 .20 .09 .08

0.78 0.42

4.7 1.3

.06 .10 .85 – .45 .45 .37 .36 .44 .49 .27 .30 .38 .15 .22 .03 .51 .48 .46 .47

4.0 1.4

.04 .02 .89 .52 – .46 .15 .59 .45 .34 .42 .24 .49 .21 .26 .09 .47 .44 .42 .49

4.3 1.3

.17 .05 .48 .49 .35 – .81 .80 .51 .40 .46 .35 .49 .20 .35 .02 .51 .57 .49 .62

4.3 1.6

.18 .09 .24 .36 .07 .75 – .31 .31 .32 .21 .31 .28 .09 .31 .10 .40 .40 .34 .41

3.7 1.6

.08 .02 .49 .38 .46 .77 .15 – .52 .32 .54 .26 .52 .24 .25 .07 .43 .53 .46 .60

5.0 1.2

.12 .03 .38 .34 .33 .37 .11 .45 – .82 .88 .24 .49 .25 .16 .00 .46 .57 .67 .68

5.5 1.2

.06 .06 .26 .35 .12 .31 .15 .32 .71 – .44 .25 .25 .17 .11 .04 .44 .49 .63 .53

3.6 1.5

.12 .09 .32 .18 .36 .25 .03 .35 .80 .15 – .18 .46 .25 .16 .02 .36 .48 .53 .61

3.3 0.9

.07 .19 .13 .19 .05 .25 .19 .19 .11 .20 .02 – .32 .24 .27 .15 .41 .47 .50 .36

3.4 0.9

.09 .15 .41 .40 .33 .42 .21 .42 .35 .18 .33 .25 – .23 .12 .16 .43 .67 .48 .57

3.0 0.9

.07 .06 .08 .10 .05 .10 .06 .10 .13 .10 .10 .13 .16 – .13 .10 .21 .28 .29 .24

3.6 0.8

4.0 0.8

4.6 1.4

3.6 0.9

3.0 1.2

.13 .08 .07 .16 .11 .21 .16 .05 .11 .06 .20 .06 .47 .41 .37 .27 .00 .49 .37 .35 .09 .10 .34 .34 .29 .38 .02 .47 .51 .36 .39 .08 .25 .26 .16 .22 .10 .46 .51 .39 .07 .01 .45 .40 .51 .07 .10 .36 .29 .49 .04 .07 .33 .32 .30 .16 .12 .27 .37 .45 .22 .04 .53 .72 .42 .10 .15 .22 .18 .26 – .06 .32 .35 .20 .10 – .03 .09 .09 .36 .07 – .65 .56 .31 .07 .69 – .61 .27 .02 .58 .67 – .23 .05 .59 .70 .67

Mean

17.2 11.3

.05 25.3 5.1 .07 0.77 0.42 .60 4.5 1.2 .54 4.8 1.3 .51 3.8 1.4 .51 4.3 1.1 .21 4.3 1.5 .56 3.6 1.5 .61 5.3 1.1 .51 5.6 1.3 .44 3.1 1.5 .21 3.3 1.0 .61 3.4 1.0 .15 3.1 0.8 .19 3.6 0.7 .11 3.9 0.7 .69 0.0 1.0 .68 4.4 1.0 .59 2.7 1.2 – 16.3 10.3

Notes. Table depicts product-moment correlations as well as means and standard deviations. The lower diagonal shows the results of Study 1, the upper diagonal the results of Study 2. Sat = Satisfaction; Dis = Dissatisfaction; CES-D = Center for Epidemiological Studies Depression Scale. Correlation coefficients > .17 or < .17 (lower diagonal; Study 1) or > .18 or < .18 (upper diagonal; Study 2) are significant at p < .01 (two-tailed). N = 251 (Study 1); N = 209 (Study 2). a0 = male, 1 = female.

Mean SD

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

Measure

Table 1. Summary of intercorrelations and descriptive statistics

60 A. B. Neubauer & A. Voss: Validation of the German BMPN

Ó 2016 Hogrefe Publishing

A. B. Neubauer & A. Voss: Validation of the German BMPN

Figure 1. Measurement models of the Balanced Measure of Psychological Needs Scale. Left: Model 0 (correlated trait uncorrelated methods); Middle: Model 1 (correlated trait (method-1)); Right: Model 2 (6-factor model).

Table 2. Model fit of the Balanced Measure of Psychological Needs scale measurement models Model

CFI

RMSEA+

SRMR

AIC

BIC

Model Model Model Model

0 1 1a 2

184.54 244.69 281.45 224.47

1.077 1.099 1.090 1.091

114 123 131 120

.953 .919 .900 .931

Study 1 .050 [.036; .062] .063 [.051; .074] .068 [.057; .079] .059 [.047; .071]

.044 .076 .081 .059

16,133.78 16,185.87 16,207.70 16,167.92

16,398.19 16,418.55 16,412.17 16,411.18

Model Model Model Model

0 1 1a 2

242.51 299.81 307.05 228.37

1.041 1.067 1.072 1.078

114 123 131 120

.874 .826 .827 .894

Study 2 .073 [.061; .086] .083 [.071; .095] .080 [.069; .092] .066 [.053; .079]

.102 .086 .087 .064

13,796.32 13,846.16 13,838.98 13,777.96

14,047.00 14,066.75 14,032.83 14,008.58

Notes. c = scaling factor; df = degrees of freedom; CFI = comparative fit index; RMSEA = root mean square error of approximation; SRMR = standardized root mean square residual; AIC = Akaike Information Criterion; BIC = Bayesian Information Criterion. + = 90% confidence interval in brackets.

methods model fitted the data very well. In fact, the parameter estimates (see Table 3) are remarkably similar to the estimates reported by Sheldon and Hilpert (2012). Hence, we were able to replicate the results of the American BMPN with our translation. From the estimates, it can also be seen that the loadings of the latent satisfaction factor were only significant for the three competence items and (although lower in size) for two of the autonomy items. In the next model (Model 1) we removed the satisfaction factor, which resulted in a statistically significant deterioration of model fit, v2(9) = 50.93, p < .001; however, the overall model fit remained in an acceptable range. In Model 1a, the factor loadings of the dissatisfaction factor were constrained to be equal for all nine items; this model fitted worse than Model 1, v2(8) = 39.79, p < .001, but again, the overall model fit remained in an acceptable to good range. The BIC even favors the more parsimonious Model 1a over Model 1. Lastly, we fitted a six-factor model to the data Ă&#x201C; 2016 Hogrefe Publishing

(Model 2). All fit indices favor this model over Model 1 and Model 1a, although they indicate somewhat worse model fit than Model 0. Taken together, these findings suggest that both a three-factor solution (with one dissatisfaction factor) and a six-factor solution are adequate. However, the six-factor solution is superior to the threefactor solution in Models 1 and 1a, and allows a more straightforward interpretation of the subscales than Model 0. Hence, we used the six subscale scores to predict markers of well-being and ill-being in the next step. Parameter estimates for Model 2 are presented in Table 4. Construct Validity Markers of well-being (life satisfaction and self-esteem) were used as dependent variables in two separate linear regression analyses (Table 5). In these analyses, the Journal of Individual Differences 2016; Vol. 37(1):56â&#x20AC;&#x201C;72

A. B. Neubauer & A. Voss: Validation of the German BMPN

Table 3. Parameter estimates for Model 0 (Study 1) Need factors Item I was free to do things my own way. My choices expressed my ‘‘true self.’’ I was really doing what interests me. I had a lot of pressure I could do without. There were people telling me what I had to do. I had to do things against my will. I was successfully completing difficult tasks and projects. I took on and mastered hard challenges. I did well even at the hard things. I experienced some kind of failure, or was unable to do well at something. I did something stupid, that made me feel incompetent. I struggled doing something I should be good at. I felt a sense of contact with people who care for me, and whom I care for. I felt close and connected with other people who are important to me. I felt a strong sense of intimacy with the people I spent time with. I was lonely. I felt unappreciated by one or more important people. I had disagreements or conflicts with people I usually get along with.

Aut

Comp

Method factors Relat

Sat

Dis

.01 .22** .14*

.62*** .63*** .73*** .49*** .44*** .45***

.40*** .39*** .38*** .35*** .41*** .61*** .67***

.72*** .83*** .49*** .51***

.51*** .38***

.38*** .56*** .74***

.06

.88*** .81*** .56*** .38*** .23***

.03 .04 .39*** .46*** .57***

R2 .39 .45 .55 .40 .35 .34 .64 .86 .61 .71 .40 .47 .55 .78 .66 .46 .35 .38

Notes. Table depicts standardized estimates. Aut = Autonomy; Comp = Competence; Relat = Relatedness; Sat = Satisfaction; Dis = Dissatisfaction. Correlations between the latent need factors were significant, p < .001: Relatedness and Competence (r = .58), Relatedness and Autonomy (r = .63), Competence and Autonomy (r = .68). N = 251. *p < .05. **p < .01. ***p < .001 (all two-tailed).

Table 4. Parameter estimates for Model 2 (Study 1) Need factors Item I was free to do things my own way. My choices expressed my ‘‘true self.’’ I was really doing what interests me. I had a lot of pressure I could do without. There were people telling me what I had to do. I had to do things against my will. I was successfully completing difficult tasks and projects. I took on and mastered hard challenges. I did well even at the hard things. I experienced some kind of failure, or was unable to do well at something. I did something stupid, that made me feel incompetent. I struggled doing something I should be good at. I felt a sense of contact with people who care for me, and whom I care for. I felt close and connected with other people who are important to me. I felt a strong sense of intimacy with the people I spent time with. I was lonely. I felt unappreciated by one or more important people. I had disagreements or conflicts with people I usually get along with. Aut_s Comp_s Relat_s Aut_d Comp_d Relat_d

Aut_s Comp_s Relat_s Aut_d Comp_d Relat_d R2 .61** .69** .74** .67** .60** .60** .81** .90** .74** .84** .63** .66** .75** .89** .81** .73** .59** .53**

.37 .47 .55 .45 .36 .36 .66 .81 .55 .70 .39 .44 .56 .78 .66 .54 .35 .28

.47** .62** .36** .65** .16 .44** .49** .39** .42** .84** .45** .30* .64** .64** .77**

Notes. Table depicts standardized factor loadings, as well as correlations of the six latent factors. Aut = Autonomy; Comp = Competence; Relat = Relatedness; _s = Satisfaction; _d = Dissatisfaction. N = 251. p < .10. *p < .05. **p < .001 (all two-tailed). Journal of Individual Differences 2016; Vol. 37(1):56–72

Ó 2016 Hogrefe Publishing

Ó 2016 Hogrefe Publishing Depressiona

Loneliness

.32

.30*** .45*** .06 .33*** .08

.43 16.94***

.31*** (.08) .13 (.07) .23** (.08)

(.08) .20** (.07) (.08) .25** (.08) (.07) .04 (.07) (.07) .25*** (.07) (.07) .06 (.07)

3.56*

.45

.22* (.09) .06 (.09) .04 (.08) .56

.62 12.80***

.08 (.04) .06 (.04) .16*** (.04)

3.26*

.63

.04 (.05) .11* (.05) .06 (.04)

.08 (.04) .04 (.04) .14** (.04)

.36

(.07) .18*** (.04) .14*** (.04) .14*** (.04) 1.82** (.64) (.08) .52*** (.04) .43*** (.04) .39*** (.05) 5.45*** (.63) (.07) .06 (.04) .05 (.04) .03 (.04) .84 (.60) (.07) .14*** (.04) .11** (.04) .10** (.04) 1.33* (.59) (.07) .09* (.04) .09* (.04) .08* (.04) .08 (.60)

.19* (.08) .16* (.08) .24** (.08)

.20** .14 .02 .20** .070

(.58) (.61) (.54) (.55) (.54)

1.14* 1.75** .02 .11 .32

.49 26.25***

23.09***

.60

.41 (.63) 2.21*** (.64) 3.21*** (.59) .38

.56 37.54***

7.53***

.59

.02 (.07) .06 (.07) .26*** (.07)

.08 (.06) .08 (.06) .00 (.06) .00 (.06) .54*** (.06) .45*** (.06)

(.52) .41*** (.07) .33*** (.06) .34*** (.06) (.60) .41*** (.07) .22** (.06) .10 (.07) (.49) .15* (.06) .11* (.05) .08 (.05) (.50) .15* (.06) .12* (.06) .09 (.06) (.48) .00 (.06) .00 (.05) .01 (.05)

1.32* (.63) 1.05 (.58) 1.50* (.59) 1.41** (.54) 3.05*** (.61) 1.84** (.57)

.96 3.80*** .66 .71 .19

4.64*** (.07) 4.64*** (.07) 4.64*** (.06) 3.63*** (.04) 3.63*** (.03) 3.63*** (.03) 17.17*** (.57) 17.17*** (.51) 17.17*** (.45) 3.07*** (.06) 3.07*** (.05) 3.07*** (.05)

Self-esteem

Notes. Table depicts unstandardized regression coefficients (standard errors in brackets). E = Extraversion; N = Neuroticism; A = Agreeableness; C = Conscientiousness; O = Openness; Aut = Autonomy; Comp = Competence; Relat = Relatedness; _s = Satisfaction; _d = Dissatisfaction. N = 251. a = 2 participants had missing values on this measure. The degrees of freedom for FD therefore are (3, 240) and (3, 237), respectively. p < .10. *p < .05. **p < .01. ***p < .001 (all two-tailed).

R2 (adjusted) FD(3, 242) FD(3, 239)

Intercept Block 1 E N A C O Block 2 Aut_s Comp_s Relat_s Block 3 Aut_d Comp_d Relat_d

Life satisfaction

Table 5. Results of the linear regression analyses (Study 1)

A. B. Neubauer & A. Voss: Validation of the German BMPN 63

Journal of Individual Differences 2016; Vol. 37(1):56–72

A. B. Neubauer & A. Voss: Validation of the German BMPN

Big Five personality traits (Extraversion, Neuroticism, Agreeableness, Conscientiousness, and Openness) were entered as predictors in a first block. The satisfaction subscales of autonomy, competence, and relatedness were then entered in a second block, followed by the three dissatisfaction subscales in a third block. All continuous predictors were z-transformed (Wainer, 2000). As expected, all three needs predicted interindividual differences in life satisfaction. After controlling for interindividual differences in core personality traits, need satisfaction explained an additional 11% of the variance in life satisfaction. Including the dissatisfaction scales increased the R2 by another 2%. Specifically, autonomy dissatisfaction predicted life satisfaction, even after controlling for the three need satisfaction subscales. Regarding the analyses on self-esteem, only competence and relatedness, but not autonomy, predicted unique variance. While interindividual differences in personality already explained 56% of the variance in selfesteem, the three satisfaction subscales increased R2 by 6%. Of the three predictors, only relatedness satisfaction was significant. Including the dissatisfaction scale leads to another 1% increase in variance explained; competence dissatisfaction predicted self-esteem over and above the need satisfaction subscales. Results on ill-being constructs also largely confirmed our hypotheses: The three need satisfaction scales predicted additional 13% of variance in depression over and above core personality traits. Crucially, all three needs independently predicted variance in the CES-D. After adding the three dissatisfaction subscales, R2 increased by 11%. Competence dissatisfaction and relatedness dissatisfaction were uniquely associated with depression, while the effect of autonomy dissatisfaction was not statistically significant. Although the size of the regression weight of autonomy satisfaction was similar in Block 2 and Block 3, it was only marginally significant in the final block. As hypothesized, loneliness was predicted by relatedness, but not by autonomy or competence. Only relatedness satisfaction predicted loneliness in Block 2, leading to a substantial increase of R2 of 28%. Including the three dissatisfaction scores increased R2 by another 3%; only the effect of relatedness dissatisfaction was statistically significant.4

Brief Discussion The German translation of the BMPN showed a very similar factor structure as the original American version. The results of Model 0 are remarkably similar to the results reported by Sheldon and Hilpert (2012). Explained variance in the 18 items ranged from .34 to .86, which is similar to the results reported for the original scale (.32–.82; Sheldon

& Hilpert, 2012); the intercorrelations of the latent factors for autonomy, competence, and relatedness in our study (.41, .58, and .62, respectively) were similar to the correlations in the original version (.51, .54, and .59). Also, explained variance in the well-being measures was identical between studies, with 45% variance explained in the SWLS in this study, and the same estimate of explained variance in the aggregate well-being measure (life satisfaction, positive affect, and negative affect) used by Sheldon and Hilpert (2012). From these findings we conclude that the psychometric properties of the BMPN were not altered in the translation process. Next, we tested alternative measurement models and conclude from these analyses that the 18 items of the BMPN are represented best by a six-factor solution, that is, the three needs should be split up into their respective satisfaction and dissatisfaction components. A three-factor solution with a latent ‘‘method’’ factor representing the dissatisfaction items is also acceptable. Hence, we conclude that the BMPN can be used to assess either fulfillment of the three needs, or the three needs split up into their satisfaction and dissatisfaction subscales. Although the sixfactor solution should be preferred, a three-factor solution is acceptable if necessary: For example, with small sample sizes, using the six scores as predictors of an outcome might overload the model. Results on construct validity largely supported our predictions: As expected, all three needs independently predicted life satisfaction and depression, but only relatedness predicted loneliness. Effect sizes (in terms of explained variance over and above interindividual differences in the Big Five personality traits) were substantial in these predictions and ranged from 13% (life satisfaction) to 24% (depression). Unexpectedly, self-esteem was only predicted by relatedness and competence, but not by autonomy, with only 7% of variance accounted for by the three needs. This finding supports theoretical accounts that tie self-esteem specifically to the current level of belongingness such as the sociometer hypothesis (Leary, Tambor, Terdal, & Downs, 1999), or that assume self-esteem to be an additional psychological need such as terror management theory (Greenberg et al., 1986). The results of Study 1 leave open several questions that will be addressed in Studies 2 and 3: Firstly, as outlined above it is unclear why self-esteem was not predicted by autonomy. One possible explanation regards the rather narrow operationalization of self-esteem by means of the Rosenberg self-esteem scale. To test whether this specific measure for self-esteem accounts for the null findings, we employ a more comprehensive measure in Study 2. For this purpose, the Multidimensional Self-Esteem Scale (MDSES; Schütz & Sellin, 2006), a German adaptation of the Multidimensional Self-Concept Scale (Fleming & Courtney, 1984),

We reran the analyses with the three need scales (i.e., not separately for satisfaction and dissatisfaction), and this did not change our main conclusions: Life satisfaction and depression were predicted by all three needs, self-esteem was predicted by relatedness and competence, and loneliness was predicted by relatedness only.

Journal of Individual Differences 2016; Vol. 37(1):56–72

Ó 2016 Hogrefe Publishing

A. B. Neubauer & A. Voss: Validation of the German BMPN

is used. The scale assesses six subdimensions of self-esteem which can be combined in one total selfesteem score. A second open question regards the relatedness dissatisfaction scale. This scale is conceptually similar to the construct loneliness: The observed correspondence between relatedness dissatisfaction and loneliness might – at least in part – be driven by an overlap in item content. To avoid an artificial inflation of correlations, we replaced item 2 of the relatedness scale (‘‘I was lonely’’) by a new item (‘‘I was excluded or ostracized’’).5 Thirdly, it is also important to assess the stability of the BMPN subscales. The BMPN is designed as a state measure. Therefore, test-retest correlations should not be exceedingly large, even over a short measurement interval. Study 2 addresses the first questions; Study 3 will investigate the stability of the BMPN.

Study 2 Method Sample and Procedure Again, data was collected using an online questionnaire. The link to this questionnaire was presented at the end of a questionnaire assessing ecological behavior which was unrelated to the current study. The link to the questionnaire was distributed via Facebook groups and mailing lists of student groups of different German universities. A total of 276 participants started the survey. Only complete questionnaires with a maximum of 5% missing values were retained (246 questionnaires). To avoid overlap with Study 1, participants were asked whether they had already participated in an earlier study on this topic. Only those participants negating this question were included in the final sample; this resulted in a final sample of 209 participants (Mage = 25.3 years, SD = 5.1, range = 18–47; 77% female).

Measurements Need Fulfillment The BMPN (Sheldon & Hilpert, 2012) was used again, but the item ‘‘I was lonely’’ was replaced by ‘‘I was excluded or ostracized.’’ Internal consistencies were satisfactory for all three needs, a = .78 (autonomy), a = .64 (competence), and a = .68 (relatedness). For the six subscales solution, the estimates were a = .73 (autonomy satisfaction), a = .75 (competence satisfaction), a = .84 (relatedness

satisfaction), a = .70 (autonomy dissatisfaction), a = .65 (competence dissatisfaction), and a = .68 (relatedness dissatisfaction).

Personality The same short form of the Big Five Inventory (Rammstedt & John, 2005) as in Study 1 was used. Internal consistencies were a = .86 (Extraversion), a = .80 (Neuroticism), a = .59 (Agreeableness), a = .71 (Conscientiousness), and a = .66 (Openness).

Life Satisfaction Two measures of life satisfaction were used in this study. First, we used a single item measure asking participants: ‘‘How satisfied are you with your life, all things considered,’’ and asked to respond on an 11-point Likert scale ranging from 0 (= completely dissatisfied) to 10 (= completely satisfied). This item is used as measurement of life satisfaction in the socioeconomic panel (Wagner, Frick, & Schupp, 2007). Additionally, the SWLS (Diener et al., 1985) was used again as a second measure of life satisfaction. Internal consistency of the SWLS was a = .86. Since the two measures were substantially correlated, r = .79, p < .001, they were z-transformed and averaged into one indicator of life satisfaction. Loneliness The same 8-item version of the UCLA Loneliness scale (Russell et al., 1980) was administered in this study (a = .86). Self-Esteem The Multidimensional Self-Esteem Scale (MDSES; Schütz & Sellin, 2006) was administered to assess self-esteem. It contains a total of 32 items capturing six sub-facets of self-esteem (self-regard, social confidence: social contact, social confidence: dealing with criticism, performance related self-esteem, physical appearance, physical ability); for 15 of the items, participants are instructed to rate to what extent these statements apply to them (ranging from 1 = not at all to 7 = very much); for the remaining 17 items, they are asked to rate how often they apply to them (ranging from 1 = never to 7 = always). The 32 items are combined into one general self-esteem score (Schütz & Sellin, 2006). Internal consistency was a = .95.

In addition to these theoretical concerns, there is also an empirical reason to replace this item. In a daily-diary design, we (Neubauer & Voss, in press) found that the item ‘‘I was lonely’’ had the lowest loading on the relatedness dissatisfaction factor and suggested that this item might need to be replaced by an alternative item.

Ó 2016 Hogrefe Publishing

Journal of Individual Differences 2016; Vol. 37(1):56–72

A. B. Neubauer & A. Voss: Validation of the German BMPN

Table 6. Parameter estimates for Model 2 (Study 2) Need factors Item

Aut_s

I was free to do things my own way. My choices expressed my ‘‘true self.’’ I was really doing what interests me. I had a lot of pressure I could do without. There were people telling me what I had to do. I had to do things against my will. I was successfully completing difficult tasks and projects. I took on and mastered hard challenges. I did well even at the hard things. I experienced some kind of failure, or was unable to do well at something. I did something stupid, that made me feel incompetent. I struggled doing something I should be good at. I felt a sense of contact with people who care for me, and whom I care for. I felt close and connected with other people who are important to me. I felt a strong sense of intimacy with the people I spent time with. I was excluded or ostracized. I felt unappreciated by one or more important people. I had disagreements or conflicts with people I usually get along with. Aut_s Comp_s Relat_s Aut_d Comp_d Relat_d

Comp_s Relat_s Aut_d Comp_d Relat_d R2

.62** .78** .67**

.38 .61 .44 .50 .45 .37 .38 .52 .66 .52

.71** .67** .61** .62** .72** .82** .72** .54** .57**

.29 .33 .63

.79** .81** .81** .69** .63** .60**

.58** .49** .23* .67** .16 .58** .36* .29* .07

.66 .66 .48 .40 .36

.16 .40** .71** .22* .53** .58**

Depression Again, the CES-D (Radloff, 1977) was administered. Internal consistency was a = .91.

Results The same models as in Study 1 were tested in CFA. Model fit indices (see Table 2, lower panel) support the conclusions drawn in Study 1: The six-factor model was superior to the alternative models; estimates for model parameters of Model 2 are presented in Table 6. To analyze the effects of need satisfaction and need dissatisfaction on well-being and ill-being, life satisfaction, general self-esteem, loneliness, and depression were entered as dependent variables in separate multiple regression analyses. In three steps, first the Big Five personality traits, then the three need satisfaction subscales, and, finally, the three need dissatisfaction subscales were entered in the analyses. As can be seen from Table 7, life satisfaction was predicted by autonomy satisfaction, relatedness satisfaction, and relatedness dissatisfaction; the effect of competence dissatisfaction was marginally significant only, p = .054. Self-esteem was predicted by competence dissatisfaction and (though Journal of Individual Differences 2016; Vol. 37(1):56–72

only marginally significant, p = .062) relatedness satisfaction. Depression was predicted by all three needs, and loneliness by relatedness, only. By and large, these results replicate the findings from Study 1.

Study 3 In Study 3, we investigated the test-retest correlations of the BMPN over a measurement interval of 1 week. Moderate relations are expected, because the BMPN aims at measuring states rather than traits.

Method Sample and Procedure A total of 106 participants were assessed twice with a time lag of 1 week. Participants completed different reaction time tasks and filled in the BMPN at the end of each session. Data from three participants had to be discarded because they did not complete both sessions. Thus, results Ó 2016 Hogrefe Publishing

A. B. Neubauer & A. Voss: Validation of the German BMPN

are based on a sample of 103 participants (Mage = 22 years, SD = 2.9; 81% female). Measurements Participants completed the revised BMPN (from Study 2) at both measurement occasions. Instructions were adapted to account for the 1-week measurement interval. Specifically, participants were asked to rate to what degree each statement applies to them with respect to their last week.

Results Internal consistencies and test-retest correlations can be found in Table 8. As can be seen from these estimates, all scales but one (relatedness dissatisfaction) showed medium sized test-retest correlations in the expected range, corroborating the assumption of moderate stability of these scales. To further explore the low test-retest correlation of the relatedness dissatisfaction subscale, test-retest correlations were computed on the item level. These analyses revealed that correlations on the item level were statically significant for two of three items of the relatedness dissatisfaction subscale, with r > .30, p < .01; however, no significant test-retest correlation was observed for the item ‘‘I had disagreements or conflicts with people I usually get along with.’’ (r = .15, p = .13).6

General Discussion This study aimed at developing and validating a German version of the BMPN (Sheldon & Hilpert, 2012). Specifically, we addressed the following objectives: (1) The factor structure of the questionnaire was analyzed; (2) the construct validity of the questionnaire was tested by exploring the relationships of the scales with indicators of well-being and ill-being; and, (3) the test-retest stability of the questionnaire was assessed. In a CFA, we first replicated the very good model fit for the three-dimensional BMPN model with two uncorrelated method factors reported by Sheldon and Hilpert (2012). However, we argue that this solution complicates the interpretation of the three need factors as the loadings of the method factors are not equal across all items. This indicates

that items (and, hence, the three subscales) are differentially influenced by the two method factors. A six-factor solution, measuring satisfaction and dissatisfaction for each need separately, allows a more straightforward interpretation of the need factors while still resulting in a good model fit. Thus, although model fit indices were somewhat worse for the six-factor model as compared to the model reported by Sheldon and Hilpert (2012), we conclude that the former solution should be preferred. Furthermore, in Study 2, where one item was replaced, we cross-validated the six-factor solution which bolsters the credibility of the results in Study 1. These results suggest that the BMPN should – whenever possible – not be used as a three-dimensional measure of autonomy, competence, and relatedness, but rather split up into the six dimensions autonomy satisfaction, autonomy dissatisfaction, competence satisfaction, competence dissatisfaction, relatedness satisfaction, and relatedness dissatisfaction. We further tested in two studies, whether the BMPN also relates to markers of well-being and ill-being as predicted by SDT. In both studies, we found evidence for construct validity: Life satisfaction was predicted by all three needs in Study 1, and by autonomy and relatedness in Study 2 (the effect of competence was marginally significant). A similar picture emerged for depression, which was predicted by all three needs in Study 2, and by competence and relatedness (and, tentatively, by autonomy) in Study 1. The expectations regarding loneliness were fully supported: In both studies, only relatedness satisfaction and dissatisfaction predicted loneliness, while the effects of autonomy and competence were not significant. This finding is particularly important regarding the discriminant validity of the needs. As to self-esteem, our results suggest that need fulfillment explains only little variance over and above the Big Five personality traits. Only competence dissatisfaction and – although only marginally significant in Study 2 – relatedness satisfaction emerged as significant predictors. Overall, these results suggest that well-being and ill-being are to a large extent predicted by satisfaction and dissatisfaction of the three needs for autonomy, competence, and relatedness, hence supporting prediction by SDT. Self-esteem is less affected by situational circumstances such as need fulfillment and largely predicted by stable interindividual differences in personality traits. These findings corroborate earlier findings on the high stability of self-esteem over the life span (Trzesniewski, Donnellan, & Robins, 2003). The overall pattern of our results suggests good construct validity of the BMPN. Finally, stability of

Since participants in Studies 2 and 3 filled in the same questionnaires, this allowed us to assess the stability of the factor structure by means of multigroup structural equation modeling. Specifically, we added t1 data of Study 3 to the Study 2 data and reestimated the best fitting model (Model 2) under varying levels of measurement invariance (Meredith, 1993). Under weak factorial invariance (same number of factors and same factor loadings in both samples), model fit remained in a comparable range (v2[258] = 436.97, c = 1.051, RMSEA = .067, CFI = .894, SRMR = .079) as model fit based solely on the Study 2 sample (Table 2). Further constraining the model to strong measurement invariance (additionally constraining the item intercepts to equality) did not decrease model fit significantly, v2(12) = 15.71, p = .205. Imposing restrictions of strict measurement invariance (additionally constraining the error variances to equality across samples) decreased model fit significantly, v2(18) = 66.85, p < .001. Hence, the measurement model (Model 2) shows evidence of strong factorial measurement invariance across samples.

Ó 2016 Hogrefe Publishing

Journal of Individual Differences 2016; Vol. 37(1):56–72

Journal of Individual Differences 2016; Vol. 37(1):56–72 Depressiona

Loneliness

.60 2.90*

.00 (.05) .01 (.05) .13** (.05)

.66 5.96*** .62 .61 1.21*

7.14***

.09 (.06) .17** (.06) .05 (.05) .63 .38

.08 (.06) .03 (.04) .10 (.05)

(.05) .14** (.05) .15** (.05) (.05) .64*** (.05) .56*** (.05) (.05) .03 (.05) .02 (.04) (.05) .18*** (.05) .17** (.05) (.05) .07 (.05) .10* (.04)

(.60) (.61) (.58) (.59) (.59)

(.50) (.54) (.49) (.52) (.49)

.13 3.41*** .33 .15 .96*

.57 37.87***

15.28***

1.90*** (.56) 1.25* (.55) 1.45** (.49) .65 .32

.45 17.61***

4.81**

.48

.13 (.08) .04 (.07) .16* (.07)

.08 (.07) .01 (.08) .06 (.07) .03 (.07) .41*** (.06) .40*** (.07)

(.46) .40*** (.07) .34*** (.06) .36*** (.06) (.52) .35*** (.07) .27*** (.07) .18** (.07) (.44) .18** (.07) .16** (.06) .15* (.06) (.48) .08 (.07) .08 (.07) .08 (.06) (.45) .04 (.07) .01 (.06) .03 (.06)

2.47*** (.57) 1.17* (.59) .20 (.53) .16 (.49) 3.61*** (.52) 3.30*** (.49)

.06 4.54*** .44 .21 1.32**

(.05) 4.37*** (.04) 4.37*** (.04) 16.27*** (.57) 16.33*** (.47) 16.33*** (.43) 2.65*** (.07) 2.65*** (.06) 2.65*** (.06)

Self-esteem

Notes. Table depicts unstandardized regression coefficients (standard errors in brackets). E = Extraversion; N = Neuroticism; A = Agreeableness; C = Conscientiousness; O = Openness; Aut = Autonomy; Comp = Competence; Relat = Relatedness; _s = Satisfaction; _d = Dissatisfaction. N = 209. a = 1 participant had missing values on this measure. The degrees of freedom for FD therefore are (3, 199) and (3, 196), respectively. p < .10. *p < .05. **p < .01. ***p < .001 (all two-tailed).

Intercept .00 (.01) .00 (.05) .00 (.05) 4.37*** Block 1 E .11* (.06) .07 (.05) .08 (.05) .17*** N .42** (.06) .32*** (.06) .25*** (.06) .66*** .10 (.05) .09 (.05) .03 A .11 (.06) C .18** (.06) .14* (.05) .13* (.05) .19*** O .01 (.05) .01 (.06) .02 (.05) .08 Block 2 Aut_s .21*** (.06) 18** (.07) Comp_s .01 (.06) .00 (.06) Relat_s .18** (.05) .13* (.06) Block 3 Aut_d .01 (.06) Comp_d .12 (.06) Relat_d .12* (.06) .42 .45 .59 R2 (adjusted) .34 16.90*** FD(3, 200) 4.12** FD(3, 197)

Life satisfaction

Table 7. Results of the linear regression analyses (Study 2)

68 A. B. Neubauer & A. Voss: Validation of the German BMPN

Ó 2016 Hogrefe Publishing

A. B. Neubauer & A. Voss: Validation of the German BMPN

Table 8. Internal consistency (Cronbach’s alpha) and test-retest correlation of the BMPN subscales Test-retest correlation

Measurement occasion

Autonomy

.53*

Autonomy satisfaction

.48*

Autonomy dissatisfaction

.47*

Competence

.46*

Competence satisfaction

.39*

Competence dissatisfaction

.49*

Relatedness

.40*

Relatedness satisfaction

.48*

Relatedness dissatisfaction

.16

t1 t2 t1 t2 t1 t2 t1 t2 t1 t2 t1 t2 t1 t2 t1 t2 t1 t2

.77 .79 .70 .68 .61 .73 .65 .63 .83 .77 .71 .67 .77 .77 .88 .89 .71 .70

Scale

Mean (SD) 4.73 4.82 4.73 4.75 3.28 3.10 4.38 4.59 3.91 3.88 3.17 2.70 5.63 5.70 5.72 5.80 2.47 3.37

(1.10) (1.08) (1.17) (1.15) (1.31) (1.32) (1.03) (0.97) (1.43) (1.35) (1.39) (1.28) (1.02) (1.02) (1.17) (1.19) (1.34) (0.89)

Note. N = 103. *p < .001 (two-tailed).

the six BMPN dimensions over the course of 1 week was investigated. Consistent with the conceptualization of the BMPN as a state measure, test-retest correlations were of moderate size. As to the dissociation into satisfaction and dissatisfaction components, this finding – while surprising at first glance – dovetails with assumptions made by Sheldon’s (2011) two process model. In this reasoning, need satisfaction and dissatisfaction can take effect at different time points of an action sequence: Sheldon (2011) assumes that need dissatisfaction triggers motivation to restore the dissatisfied need, while need satisfaction rewards a successful restoration process. This reasoning also explains why Facebook use (which is hypothesized to be a relatedness restoration process; Sheldon et al., 2011) correlates positively with relatedness dissatisfaction and relatedness satisfaction. Thus, treating need satisfaction and dissatisfaction as merely psychometric opposites would be unwarranted by both theoretical expectations and empirical evidence gathered in this work. Promising avenues for future research on this dissociation include experimental and intensive longitudinal designs to capture the temporal dynamics of need restoration processes and to investigate the hypothesized differential role of need satisfaction and need dissatisfaction.

possibility that correlations between the measures were in part driven by common method variance. Additional behavioral data should be collected to explore the criterionrelated validity of the BMPN. Third, exploring the stability of the BMPN by means of simple test-retest correlations precludes conclusions about trait consistency and state specificity: Low test-retest correlations could result from either low stability (and hence, high occasion specificity) or low reliability of the measurement. To disentangle these two sources, we suggest that future research investigate trait consistency and state specificity of the BMPN subscales using latent state-trait models (Steyer, Schmitt, & Eid, 1999). Fourth, when exploring the nomological network of the BMPN subscales, we were primarily interested in whether the three needs predicted ill-being and well-being, regardless of whether the effect was driven by the satisfaction or the dissatisfaction components. We had no a priori expectations as to whether the effects should be driven by need satisfaction or need dissatisfaction, which is why we did not focus on the differential effects of satisfaction and dissatisfaction subscales. Although experimental data (Sheldon & Filak, 2008) suggests that need frustration has a larger impact on well-being than need satisfaction, this finding was not apparent in our data. Future research – possibly using the BMPN – should further investigate the differential effects of need satisfaction and dissatisfaction.

Limitations and Directions for Future Research

Conclusions

In the interpretation of our results, some caveats should be noted: First, our samples were convenience samples which are not representative. Future research needs to replicate these findings using a more heterogeneous sample. Second, all measures are based on self-reports which leaves open the

Our results suggest that the revised German BMPN is a useful tool to assess need fulfillment in a German-speaking population. The questionnaire is easy to administer, exhibits good internal consistency, fits the proposed factorial structure, and predicts several markers of well-being and ill-being.

Ó 2016 Hogrefe Publishing

Journal of Individual Differences 2016; Vol. 37(1):56–72

A. B. Neubauer & A. Voss: Validation of the German BMPN

In accord with previous research (Neubauer & Voss, in press; Sheldon et al., 2011; Sheldon & Filak, 2008) our data support the notion that need satisfaction and need dissatisfaction are more than only psychometric opposites and can differentially affect self-reports and behavior. We advise researchers interested in assessing need fulfillment to use the revised version of the BMPN, that is, replacing the original item ‘‘I was lonely’’ with our alternative item ‘‘I was rejected or ostracized’’ to avoid too high content overlap between relatedness dissatisfaction and loneliness.

References Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and selfdetermination in human behavior. New York, NY: Plenum. Deci, E. L., & Ryan, R. M. (2000). The ‘‘What’’ and ‘‘Why’’ of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11, 227–268. doi: 10.1207/ S15327965PLI1104_01 DeWall, C. N., Twenge, J. M., Koole, S. L., Baumeister, R. F., Marquez, A., & Reid, M. W. (2011). Automatic emotion regulation after social exclusion: Tuning to positivity. Emotion, 11, 623–636. doi: 10.1037/a0023534 Diener, E. (1984). Subjective well-being. Psychological Bulletin, 95, 542–575. doi: 10.1037/0033-2909.95.3.542 Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The Satisfaction with Life Scale. Journal of Personality Assessment, 49, 71–75. doi: 10.1207/s15327752jpa4901_13 Döring, N., & Bortz, J. (1993). Psychometrische Einsamkeitsforschung. Deutsche Neukonstruktion der UCLA Loneliness Scale [Psychometric Loneliness Research: German Newconstruction of the ULCA Loneliness Scale]. Diagnostica, 39, 224–239. Eid, M., Nussbeck, F. W., Geiser, C., Cole, D. A., Gollwitzer, M., & Lischetzke, T. (2008). Structural equation modeling of multitrait-multimethod data: Different models for different types of methods. Psychological Methods, 13, 230–253. doi: 10.1037/a0013219 Ferring, D., & Filipp, S.-H. (1996). Messung des Selbstwertgefühls: Befunde zu Reliabilität, Validität und Stabilität der Rosenberg-Skala [Measurement of self-esteem: Findings on reliability, validity, and stability of the Rosenberg Scale]. Diagnostica, 42, 284–292. Fleming, J. S., & Courtney, B. E. (1984). The dimensionality of self-esteem: II. Hierarchical facet model for revised measurement scales. Journal of Personality and Social Psychology, 46, 404–421. doi: 10.1037/0022-3514.46.2.404 Gagné, M. (2003). The role of autonomy support and Autonomy orientation in prosocial behavior engagement. Motivation and Emotion, 27, 199–223. doi: 10.1023/A:1025007614869 Geiser, C., Eid, M., & Nussbeck, F. W. (2008). On the meaning of the latent variables in the CT-C(M-1) model: A comment on Maydeu-Olivares and Coffman (2006). Psychological Methods, 13, 49–57. doi: 10.1037/1082-989X.13.1.49 Glaesmer, H., Grande, G., Braehler, E., & Roth, M. (2011). The German version of the Satisfaction With Life Scale (SWLS). European Journal of Psychological Assessment, 27, 127–132. doi: 10.1027/1015-5759/a000058 Greenberg, J., Pyszczynski, T., & Solomon, S. (1986). The causes and consequences of a need for self-esteem: A terror management theory. In R. F. Baumeister (Ed.), Public self and private self (pp. 189–212). New York, NY: Springer.

Journal of Individual Differences 2016; Vol. 37(1):56–72

Hautzinger, M. (1988). Die CES-D Skala: Ein Depressionsmesssinstrument für Untersuchungen in der Allgemeinbevölkerung [The CES-D sclae. A measurement for depression in the general population]. Diagnostica, 34, 167–173. Hays, R. D., & DiMatteo, M. R. (1987). A short-form measure of loneliness. Journal of Personality Assessment, 51, 69–81. doi: 10.1207/s15327752jpa5101_6 Johnston, M. M., & Finney, S. J. (2010). Measuring basic needs satisfaction: Evaluating previous research and conducting new psychometric evaluations of the Basic Needs Satisfaction in General Scale. Contemporary Educational Psychology, 35, 280–296. doi: 10.1016/j.cedpsych.2010.04.003 Leary, M., Tambor, E., Terdal, S., & Downs, D. (1999). SelfEsteem as as Interpersonal Monitor: The Sociometer Hypothesis. In R. F. Baumeister (Ed.), Key readings in social psychology. The self in social psychology (pp. 87– 104). Philadelphia, PA: Psychology Press. Meredith, W. (1993). Measurement invariance, factor analysis and factorial invariance. Psychometrika, 58, 525–543. doi: 10.1007/BF02294825 Muthén, L. K., & Muthén, B. O. (2015). Mplus User’s Guide (7th ed). Los Angeles, CA: Muthén & Muthén. Neubauer, A. B., & Voss, A. (in press). The structure of need fulfillment: Separating need satisfaction and dissatisfaction on between- and within-person level. In European Journal of Psychological Assessment. Manuscript submitted for publication. R Core Team. (2015). R: A language and environment for statistical computing. Retrieved from https://www.Rproject.org/. Vienna, Austria: R Foundation for Statistical Computing. Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385–401. doi: 10.1177/ 014662167700100306 Rammstedt, B., & John, O. P. (2005). Kurzversion des Big Five Inventory (BFI-K). Diagnostica, 51, 195–206. doi: 10.1026/ 0012-1924.51.4.195 Reis, H. T., Sheldon, K. M., Gable, S. L., Roscoe, J., & Ryan, R. M. (2000). Daily well-being: The role of autonomy, competence, and relatedness. Personality and Social Psychology Bulletin, 26, 419–435. doi: 10.1177/ 0146167200266002 Rhemtulla, M., Brosseau-Liard, P. É., & Savalei, V. (2012). When can categorical variables be treated as continuous? A comparison of robust continuous and categorical SEM estimation methods under suboptimal conditions. Psychological Methods, 17, 354–373. doi: 10.1037/a0029315 Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press. Russell, D., Peplau, L. A., & Cutrona, C. E. (1980). The revised UCLA Loneliness Scale: Concurrent and discriminant validity evidence. Journal of Personality and Social Psychology, 39, 472–480. Ryan, R. M., & Brown, K. W. (2003). Why we don’t need selfesteem: On fundamental needs, contingent love, and mindfulness. Psychological Inquiry, 14, 71–76. doi: 10.2307/ 1449046 Ryan, R. M., & Deci, E. L. (2001). On happiness and human potentials: A review of research on hedonic and eudaimonic well-being. Annual Review of Psychology, 52, 141–166. doi: 10.1146/annurev.psych.52.1.141 Schütz, A., & Sellin, I. (2006). Multidimensionale Selbstwertskala [Multidimensional Self-Concept Scale]. Göttingen, Germany: Hogrefe.

Ó 2016 Hogrefe Publishing

A. B. Neubauer & A. Voss: Validation of the German BMPN

Sheldon, K. M., Ryan, R., & Reis, H. T. (1996). What makes for a good day? Competence and autonomy in the day and in the person. Personality and Social Psychology Bulletin, 22, 1270–1279. doi: 10.1177/01461672962212007 Sheldon, K. M. (2011). Integrating behavioral-motive and experiential-requirement perspectives on psychological needs: A two process model. Psychological Review, 118, 552–569. doi: 10.1037/a0024758 Sheldon, K. M., Abad, N., & Hinsch, C. (2011). A two-process view of Facebook use and relatedness need-satisfaction: Disconnection drives use, and connection rewards it. Journal of Personality and Social Psychology, 100, 766–775. doi: 10.1037/a0022407 Sheldon, K. M., & Filak, V. (2008). Manipulating autonomy, competence, and relatedness support in a game-learning context: New evidence that all three needs matter. British Journal of Social Psychology, 47, 267–283. doi: 10.1348/ 014466607X238797 Sheldon, K. M., & Gunz, A. (2009). Psychological needs as basic motives, not just experiential requirements. Journal of Personality, 77, 1467–1492. doi: 10.1111/j.14676494.2009.00589.x Sheldon, K. M., & Hilpert, J. C. (2012). The balanced measure of psychological needs (BMPN) scale: An alternative domain general measure of need satisfaction. Motivation and Emotion, 36, 439–451. doi: 10.1007/s11031-012-9279-4 Steyer, R., Schmitt, M., & Eid, M. (1999). Latent state-trait theory and research in personality and individual differences. European Journal of Personality, 13, 389–408. Trzesniewski, K. H., Donnellan, M. B., & Robins, R. W. (2003). Stability of self-esteem across the life span. Journal of Personality and Social Psychology, 84, 205–220. doi: 10.1037/0022-3514.84.1.205 Vansteenkiste, M., Lens, W., Soenens, B., & Luyckx, K. (2006). Autonomy and relatedness among Chinese sojourners and applicants: Conflictual or independent predictors of wellbeing and adjustment? Motivation and Emotion, 30, 273–282. doi: 10.1007/s11031-006-9041-x Wagner, G. G., Frick, J. R., & Schupp, J. (2007). The German Socio-Economic Panel Study (SOEP) – Scope, evolution and enhancements. Schmollers Jahrbuch, 127, 139–169.

Ó 2016 Hogrefe Publishing

Wainer, H. (2000). The centercept: An estimable and meaningful regression parameter. Psychological Science, 11, 434–436. Wei, M., Shaffer, P. A., Young, S. K., & Zakalik, R. A. (2005). Adult attachment, shame, depression, and loneliness: The mediation role of basic psychological needs Satisfaction. Journal of Counseling Psychology, 52, 591–601. doi: 10.1037/0022-0167.52.4.591 Williams, K. D. (2001). Ostracism: The power of silence. Emotions and social behavior. New York, NY: Guilford Press. Wu, C.-H., & Yao, G. (2008). Psychometric analysis of the short-form UCLA Loneliness Scale (ULS-8) in Taiwanese undergraduate students. Personality and Individual Differences, 44, 1762–1771. doi: 10.1016/j.paid.2008.02.003 Yuan, K.-H., & Bentler, P. M. (2000). Three likelihood-based methods for mean and covariance structure analysis with nonnormal missing data. Sociological Methodology, 30, 165–200. doi: 10.1111/0081-1750.00078

Date of acceptance: May 12, 2015 Published online: February 29, 2016

Andreas B. Neubauer Institute of Psychology University of Heidelberg Hauptstr. 47-51 69117 Heidelberg Germany Tel. +49 622 154-7345 Fax +49 622 154-7273 E-mail andreas.neubauer@psychologie.uni-heidelberg.de

Journal of Individual Differences 2016; Vol. 37(1):56–72

A. B. Neubauer & A. Voss: Validation of the German BMPN

Appendix The Balanced Measure of Psychological Needs Scale: Items and translations Item I felt a sense of contact with people who care for me, and whom I care for. I was lonely. I was excluded or ostracized.* I felt close and connected with other people who are important to me. I felt unappreciated by one or more important people. I felt a strong sense of intimacy with the people I spent time with. I had disagreements or conflicts with people I usually get along with. I was successfully completing difficult tasks and projects. I experienced some kind of failure, or was unable to do well at something. I took on and mastered hard challenges. I did something stupid, that made me feel incompetent. I did well even at the hard things. I struggled doing something I should be good at. I was free to do things my own way. I had a lot of pressure I could do without. My choices expressed my ‘‘true self.’’ There were people telling me what I had to do. I was really doing what interests me. I had to do things against my will.

Translation

Scale

Ich hatte das Gefühl in Kontakt mit Menschen zu sein, die mir nahe stehen. Ich habe mich einsam gefühlt. Andere Menschen haben mich zurückgewiesen oder ausgegrenzt.* Ich habe mich anderen Menschen, die mir wichtig sind, nahe und verbunden gefühlt. Ich habe mich von einem oder mehreren mir wichtigen Menschen nicht wertgeschätzt gefühlt. Ich habe eine starke Vertrautheit mit den Menschen gespürt, mit denen ich Zeit verbracht habe. Ich hatte Unstimmigkeiten oder Konflikte mit Menschen, mit denen ich normal gut zu Recht komme. Ich habe erfolgreich eine schwierige Aufgabe oder ein schwieriges Projekt abgeschlossen. Ich hatte das Gefühl, bei irgendetwas versagt zu haben oder nicht gut in etwas zu sein. Ich habe große Herausforderungen angenommen und gemeistert. Ich habe etwas Dummes gemacht und mich deshalb inkompetent gefühlt. Ich war erfolgreich, selbst bei schwierigen Dingen. Ich habe mich mit etwas schwer getan, das ich eigentlich gut kann. Ich hatte den Freiraum Dinge so zu tun, wie ich es wollte. Ich habe viel Druck gespürt, auf den ich lieber verzichtet hätte. Meine Handlungen waren Ausdruck meines ‘‘wahren Ichs.’’ Andere Menschen haben mir vorgeschrieben, was ich tun soll. Ich habe wirklich das getan, was mich interessiert. Ich musste Dinge gegen meinen Willen tun.

Relatedness (+) Relatedness ( ) Relatedness ( ) Relatedness (+) Relatedness ( )

Relatedness (+)

Relatedness ( )

Competence (+)

Competence ( ) Competence (+) Competence ( ) Competence (+) Competence ( ) Autonomy (+) Autonomy ( ) Autonomy (+) Autonomy ( ) Autonomy (+) Autonomy ( )

Notes. (+) = satisfaction subscale; ( ) = dissatisfaction subscale. Item removed in revised scale. *Item not included in the original version.

Journal of Individual Differences 2016; Vol. 37(1):56–72

Ó 2016 Hogrefe Publishing

Instructions to Authors The Journal of Individual Differences publishes manuscripts dealing with individual differences in behavior, emotion, cognition, and their developmental aspects. This includes human as well as animal research. The The Journal of Individual Differences is conceptualized to bring together researchers working in different areas ranging from, for example, molecular genetics to theories of complex behavior. Moreover, it places emphasis on papers dealing with special methodological and conceptual issues in basic science as well as in their applied fields (assessment of personality and intelligence). Journal of Individual Differences publishes the following types of articles: Regular Research Articles, Extended Research Articles, Meta-Analyses, and Reviews. Manuscript submission: All manuscripts should in the first instance be submitted electronically at http://www.editorialmanager.com/jindivdiff. Detailed instructions to authors are provided at http://www.hogrefe.com/periodicals/journal-ofindividual-differences/advice-for-authors/ Copyright Agreement: By submitting an article, the author confirms and guarantees on behalf of him-/herself and any coauthors that the manuscript has not been submitted or published elsewhere, and that he or she holds all copyright in and titles to the submitted contribution, including any figures, photographs, line drawings, plans, maps, sketches, tables, and electronic supplementary material, and that the article and its contents do not infringe in any way on the rights of third parties. ESM will be published online as received from the author(s) without any conversion, testing, or reformatting. They will not be checked for typographical errors or functionality. The author indemnifies and holds harmless the publisher from any third-party claims. The author agrees, upon acceptance of the article for publication, to transfer to the publisher the exclusive right to reproduce and distribute the article and its contents, both physically and in nonphysical, electronic, or other form, in the journal to which it has been submitted and in other

Ó 2016 Hogrefe Publishing

independent publications, with no limitations on the number of copies or on the form or the extent of distribution. These rights are transferred for the duration of copyright as defined by international law. Furthermore, the author transfers to the publisher the following exclusive rights to the article and its contents: 1. The rights to produce advance copies, reprints, or offprints of the article, in full or in part, to undertake or allow translations into other languages, to distribute other forms or modified versions of the article, and to produce and distribute summaries or abstracts. 2. The rights to microfilm and microfiche editions or similar, to the use of the article and its contents in videotext, teletext, and similar systems, to recordings or reproduction using other media, digital or analog, including electronic, magnetic, and optical media, and in multimedia form, as well as for public broadcasting in radio, television, or other forms of broadcast. 3. The rights to store the article and its content in machinereadable or electronic form on all media (such as computer disks, compact disks, magnetic tape), to store the article and its contents in online databases belonging to the publisher or third parties for viewing or downloading by third parties, and to present or reproduce the article or its contents on visual display screens, monitors, and similar devices, either directly or via data transmission. 4. The rights to reproduce and distribute the article and its contents by all other means, including photomechanical and similar processes (such as photocopying or facsimile), and as part of so-called document delivery services. 5. The right to transfer any or all rights mentioned in this agreement, as well as rights retained by the relevant copyright clearing centers, including royalty rights to third parties. Online Rights for Journal Articles: Guidelines on authors’ rights to archive electronic versions of their manuscripts online are given in the Advice for Authors on the journal’s web page at www.hogrefe.com. February 2016

Journal of Individual Differences 2016; Vol. 37(1)

Alternatives to traditional self-reports in psychological assessment “A unique and timely guide to better psychological assessment.” Rainer K. Silbereisen, Research Professor, Friedrich Schiller University Jena, Germany Past-President, International Union of Psychological Science

Tuulia Ortner / Fons J.R. van de Vijver (Editors)

Behavior-Based Assessment in Psychology Going Beyond Self-Report in the Personality, Affective, Motivation, and Social Domains Series: Psychological Assessment – Science and Practice – Vol. 1 2015, vi + 234 pp. US $63.00 / € 44.95 ISBN 978-0-88937-437-9 Also available as an eBook Traditional self-reports can be an unsufficiant source of information about personality, attitudes, affect, and motivation. What are the alternatives? This first volume in the authoritative series Psychological Assessment – Science and Practice discusses the most influential, state-of-the-art forms of assessment that can take us beyond self-report. Leading scholars from various countries describe the

www.hogrefe.com

theoretical background and psychometric properties of alternatives to self-report, including behavior-based assessment, observational methods, innovative computerized procedures, indirect assessments, projective techniques, and narrative reports. They also look at the validity and practical application of such forms of assessment in domains as diverse as health, forensic, clinical, and consumer psychology.

European Journal of Psychological Assessment

nline free o issue le samp

Editor-in-Chief Matthias Ziegler Humboldt University Berlin, Germany Editorial Assistant Doreen Bensch Humboldt University Berlin, Germany

ISSN-Print 1015-5759 ISSN-Online 2151-2426 ISSN-L 1015-5759 4 issues per annum (= 1 volume)

Subscription rates (2016) Libraries / Institutions US $434.00 / € 334.00 Individuals US $208.00 / € 152.00 Postage / Handling US $16.00 / € 12.00

www.hogrefe.com

About the Journal The main purpose of the EJPA is to present important articles which provide seminal information on both theoretical and applied developments in this field. Articles reporting the construction of new measures or an advancement of an existing measure are given priority. The journal is directed to practitioners as well as to academicians: The conviction of its editors is that the discipline of psychological assessment should, necessarily and firmly, be attached to the roots of psychological science, while going deeply into all the consequences of its applied, practice-oriented development. Psychological assessment is experiencing a period of renewal and expansion, attracting more and more attention from both academic and applied psychology, as well as from political, corporate, and social organizations. The EJPA provides a meeting point for this movement, contributing to the scientific development of psychological assessment and to communication between professionals and researchers in Europe and worldwide.

Associate Editors Martin Bäckström, Sweden Gary N. Burns, USA Laurence Claes, Belgium Marjolein Fokkema, The Netherlands David Gallardo-Pujol, Spain Samuel Greiff, Luxembourg Christoph Kemper, Germany Lena Lämmle, Germany Carolyn MacCann, Australia René T. Proyer, Germany Sebastian Sauer, Germany Marit Sijbrandij, The Netherlands

Manuscript Submissions All manuscripts should be submitted online at www.editorialmanager.com/ejpa, where full instructions to authors are also available. Electronic Full Text The full text of the journal – current and past issues (from 1995 onward) – is available online at http://econtent.hogrefe.com/ loi/jpa (included in subscription price). A free sample issue is also available here. Abstracting Services The journal is abstracted/indexed in Current Contents/Social and Behavioral Sciences (CC/S&BS), Social Sciences Citation Index (SSCI), Social SciSearch, PsycINFO, Psychological Abstracts, PSYNDEX, ERIH, and Scopus. Impact Factor (Journal Citation Reports®, Thomson Reuters): 2014 = 1.973

State-of-the-art in developing and constructing psychometric tests ming Co l Apri 2016

Karl Schweizer / Christine DiStefano

Principles and Methods of Test Construction Standards and Recent Advances

Series: Psychological Assessment – Science and Practice – Vol. 3 2016, ca. vi + 329 pp. ca. US $63.00 / € 44.95 ISBN 978-0-88937-449-2 Also available as an eBook This latest volume in the series Psychological Assessment – Science and Practice describes the current state-of-the-art in test development and construction. The past 10-20 years have seen substantial advances in the methods used to develop and administer tests. In this volume many of the world’s leading authorities collate these advances and provide information about current practices, thus equipping researchers and students to successfully construct new tests using the best modern standards and tech-

www.hogrefe.com

niques. The first section explains the benefits of considering the underlying theory when designing tests, such as factor analysis and item response theory. The second section looks at item format and test presentation. The third discusses model testing and selection, while the fourth goes into statistical methods that can find group-specific bias. The final section discusses topics of special relevance such as multi-trait multi-state analyses and development of screening instruments.