Journal of Individual Differences 2020 by Hogrefe

Volume 41 / Number 1 / 2020

Journal of

Individual Differences Editor-in-Chief Martin Voracek Associate Editors André Beauducel Sam Gosling Jürgen Hennig Philipp Yorck Herzberg Andrea Hildebrandt Anja Leue Alan Pickering Karl-Heinz Renner Willibald Ruch Annekathrin Schacht Astrid Schütz Andrzej Sekowski

32nd International Congress of Psychology

July 19 - 24, 2020 Prague, Czech Republic 5-day Scientific Programme Over 25 State-of-the-Art Lectures Over 100 Keynote Addresses Over 190 Invited Symposia Over 5 Controversial Debates and much more …

Represent your country and join us at the ICP 2020! Follow us on Facebook, Twitter, Instagram!

Journal of

Individual Differences Volume 41 /Number 1/2020

Editor-in-Chief

Prof. Martin Voracek, Department of Basic Psychological Research and Research Methods, University of Vienna, Liebiggasse 5, 1010 Vienna, Austria

Associate Editors

André Beauducel, Germany Sam Gosling, USA Jürgen Hennig, Germany Philipp Yorck Herzberg, Germany Andrea Hildebrandt, Germany Anja Leue, Germany

Alan Pickering, UK Karl-Heinz Renner, Germany Willibald Ruch, Switzerland Annekathrin Schacht, Germany Astrid Schütz, Germany Andrzej Sekowski, Poland

Editorial Board

Philipp L. Ackerman, USA José 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 Häcker, Germany Willem B. Hofstee, The Netherlands

Klaus Kubinger, Austria Bernd Marcus, Germany Robert R. McCrae, USA Carolyn C. Morf, Switzerland Pierre Mormede, France Kurt Pawlik, Germany Robert Plomin, UK Rainer Riemann, Germany Kurt Stapf, Germany Gerhard Stemmler, Germany Jan Strelau, Poland

Publisher

Hogrefe Publishing, Merkelstr. 3, D-37085 Göttingen, 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, 361 Newbury Street, 5th Floor, Boston, MA 02115, USA, Phone (875) 880-2002, E-mail publishing@hogrefe.com

Production

Anne-Lisa Löck, Hogrefe Publishing, Merkelstr. 3, D-37085 Göttingen, Germany, Tel. +49 551 99950-0, Fax +49 551 99950-111, E-mail production@hogrefe.com

Subscriptions

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

Advertising/Inserts

Melanie Beck, Hogrefe Publishing, Merkelstr. 3, D-37085 Göttingen, 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 2020 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 2020: Institutions – from US $308.00/€232.00 (print only; pricing for online access can be found in the journals catalog at hgf.io/journalscatalog); Individuals – US $164.00/€118.00 (print & online). Postage and handling – US $16.00/€12.00. Single copies – US $81.00/€62.50 + postage and handling.

Payment

Payment may be made by check, international money order, or credit card, to Hogrefe Publishing, Merkelstr. 3, D-37085 Göttingen, Germany, or, for North American customers, to Hogrefe Publishing, 361 Newbury Street, 5th Floor, Boston, MA 02115, 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. 2018 Impact Factor 1.138, 5-year Impact Factor 1.521, Journal Citation Reports (Clarivate Analytics, 2019)

Journal of Individual Differences (2020), 41(1)

Ó 2020 Hogrefe Publishing

Contents Original Articles

Ó 2020 Hogrefe Publishing

Core Self-Evaluations Over Time: Predicting Within-Person Variability Michael C. Tocci, Patrick D. Converse, and Nicholas A. Moon

Identiﬁcation of Emotions in Offenders With Antisocial Personality Disorder (ASP): Behavioral and Autonomic Responses Depending on the Reinforcement Scheme Catarina Iria, Fernando Barbosa, and Rui Paixão

A New Money Behavior Quiz Adrian Furnham and Simmy Grover

Reliability of MTurk Data From Masters and Workers Steven V. Rouse

Separating Content and Structure in Humor Appreciation: A Bimodal Structural Equation Modeling Approach Sonja Heintz

Positive and Negative Wayﬁnding Inclinations, Choice of Navigation Aids, and How They Relate to Personality Traits Chiara Meneghetti, Francesco Grimaldi, Massimo Nucci, and Francesca Pazzaglia

Bite the Stress Away? Nail Biting and Smoking Predict Maladaptive Stress Coping Strategies Magdalena Siegel, Eva-Maria Adlmann, Georg Gittler, and Jakob Pietschnig

Journal of Individual Differences (2020), 41(1)

Original Article

Core Self-Evaluations Over Time Predicting Within-Person Variability Michael C. Tocci, Patrick D. Converse, and Nicholas A. Moon School of Psychology, Florida Institute of Technology, Melbourne, FL, USA

Abstract: Core self-evaluations (CSEs) represent a prominent construct with links to a number of important organizational behaviors and outcomes. Previous research on this variable appears to have assumed that CSE is a stable trait. However, very little research has examined this assumption, particularly over longer time periods. This study investigated this issue, focusing on within-person variability in CSE. Drawing from several theoretical perspectives related to self-concept, we examined the extent to which levels of this construct varied over several years as well as potential predictors of this variability. Hierarchical linear modeling analyses indicated there was substantial within-person variance in CSE over time and this variability was related to income and education. These findings shed additional light on the fundamental nature of CSE, contributing to a new perspective on this construct with potential implications for employees, organizations, and researchers. Keywords: core self-evaluations, self-concept, personality, stability, longitudinal

Core self-evaluation (CSE) is a self-concept construct involving individuals’ judgments of themselves. This construct has received substantial attention as it has been linked to a range of important work-related outcomes such as income (Judge & Hurst, 2008), job satisfaction (Converse et al., 2016), and job performance (Judge & Bono, 2001). Although a growing number of studies have focused on CSE, several fundamental questions regarding the nature and implications of this construct have received very little attention. The purpose of this research is to address one of these questions. Specifically, this study examines CSE variability over relatively long time periods (years) and how this variability may be predicted by several work and life variables. Previous research appears to have assumed that CSE is a stable trait but some theoretical and empirical work suggests there may be notable within-person variability in CSE across long time periods. However, little research has directly examined this issue, resulting in significant ambiguity regarding this aspect of the construct. Given the demonstrated importance of CSE in organizational settings, the fundamental nature of this question, and the potential implications of this issue for research and practice, developing a better understanding of this aspect of CSE appears to be useful both theoretically and practically.

Stability of Core Self-Evaluations CSE refers to the “basic conclusions or bottom-line evaluations that individuals hold about themselves” (Judge & Ó 2019 Hogrefe Publishing

Bono, 2001, p. 80). CSE is a higher-order construct comprised of four underlying constructs: locus of control, self-esteem, self-efficacy, and neuroticism. These four constructs have shown convergent validity supporting one higher-order construct – CSE (Judge, Erez, Bono, & Thoresen, 2003). One basic issue with potentially important implications for conceptual understanding and practical use that has received little attention is the potential for within-person change in CSE. The general assumption appears to be that CSE, as a “core” trait, is quite stable. For example, Zacher (2014) explicitly states: “core self-evaluations are a stable personality trait that involves people’s fundamental evaluations of themselves” (p. 23). In addition, Wu and Griffin (2012) note: “most studies have assumed that CSE is a relatively stable personality construct” (p. 331). However, this assumption has not received much direct research attention, as few studies have investigated the extent to which CSE varies over time and what might influence CSE levels. Recently, a few studies have begun to examine this in terms of a state form of CSE (Debusscher, Hofmans, & De Fruyt, 2016; Dóci & Hofmans, 2015; Hofmans, Debusscher, Dóci, Spanouli, & De Fruyt, 2015; Schinkel, van Dierendonck, & Anderson, 2004). Schinkel et al. (2004), for example, showed that CSE decreased after participants received a rejection message with performance feedback. These findings are interesting and informative but also limited insofar as they focus on momentary or short-term fluctuations which may be different from longer-term fluctuations. For instance, the nature of this variability may be different (e.g., the amount of variability in CSE seen in the short-term might be different from that Journal of Individual Differences (2020), 41(1), 1–7 https://doi.org/10.1027/1614-0001/a000314

observed in the longer-term) and the predictors of this variability may be distinct (e.g., momentary experiences and states such as mood may influence CSE in the short-term, whereas broader experiences and accomplishments such as educational attainment may influence CSE in the longer-term). However, very little work has examined variability in CSE over longer time periods. Indeed, in their experience sampling research, Debusscher et al. (2016) recently highlighted this issue, noting: “There is however still a lot to learn about the effect of time in psychological research in general and about time effects in CSEs in particular (see Roe, 2008) and further research should thus investigate to what extent CSEs vary over longer periods of time” (p. 310). Perhaps more relevant to the current research, Wu and Griffin (2012) assessed the linear growth of individuals’ job satisfaction over a four-year period and its relationship to the individuals’ level of CSE in the subsequent year. They examined this relationship twice, based on 10 waves. Although they observed relatively high rank-order stability for CSE from Wave 5 to Wave 10, their findings also suggest that CSE can be influenced by work-related factors. Specifically, they found that growth of job satisfaction was significantly related to subsequent CSE. These findings also represent a useful contribution but have several limitations in terms of implications for CSE variability: The approach used assesses rank-order stability but does not directly address within-person variability over time, CSE was assessed in only two waves, and only one predictor was examined. Given this, the current research focused on within-person variability over time, examined CSE across several data collection waves over several years, and investigated multiple factors that might relate to these changes.

Changes in Core Self-Evaluations: Theoretical Considerations Although personality traits such as those underlying CSE are often thought of as stable, there are several perspectives indicating that personality can change throughout an individual’s life (e.g., Caspi, Roberts, & Shiner, 2005; Roberts, 2006; Roberts, Wood, & Caspi, 2008). This work has also suggested that an individual’s environment can have a significant effect on personality (Roberts, 2006; Roberts et al., 2008). This perspective is summarized in the plasticity principle of personality development: “personality traits are open systems that can be influenced by the environment at any age” (Roberts et al., 2008, p. 384). With respect to CSE in particular, several theoretical models point to factors that may affect CSE (or CSE components), producing variability in levels of this construct over time. One example is the Core Self-Evaluations Job Affect Multilevel model (Judge, Hulin, & Dalal, 2012). This Journal of Individual Differences (2020), 41(1), 1–7

M. C. Tocci et al., Core Self-Evaluations Over Time

model proposes that several aspects of one’s work and life environment (e.g., performing well, achieving outcomes) can influence CSE. Furthermore, these authors specifically note that “Within-individual variation may occur over minutes, hours, days, and even years” (p. 515). Therefore, this model indicates that CSE may fluctuate over time due to changes in a variety of experiences. Other theories provide further guidance regarding specific types of experiences that should be particularly relevant. One major example is the sociometer theory of self-esteem (Leary, 1999). This theory suggests that, given the importance of interpersonal relationships, self-esteem is influenced by information pertaining to social value and acceptance. Therefore, according to this model, self-esteem – and thus CSE – may fluctuate over time due to changes in perceptions of social value and acceptance. Another example is social-cognitive theory (e.g., Bandura, 1986), which proposes that self-efficacy develops based on mastery experiences, social modeling, social persuasion, and physical/ emotional states. Although all of these factors may influence CSE, mastery experiences are most directly relevant to the current research. Mastery experiences refer to overcoming challenges through perseverance (e.g., obtaining a graduate degree). This perspective therefore suggests that self-efficacy – and thus CSE – may fluctuate over time due to mastery experiences. Finally, similar propositions are also included in Roberts’ (2006) neo-socioanalytic topographical model of personality psychology. This model is relevant because self-concept constructs such as CSE fit within the notion of identity and the model suggests that several factors may influence identity. Roberts (2006), for example, suggests that identity is shaped by social interactions that are organized according to social roles. The model also indicates that reputation affects identity, where an individual’s selfperceptions are influenced by how others view that person. This suggests that identity – and thus CSE – may fluctuate over time as social roles and reputation change.

Changes in Core Self-Evaluations: Empirical Evidence Empirical evidence related to the components of CSE also suggests that levels of this construct may vary over time and more specifically that factors related to social value and acceptance, mastery experiences, and reputation may influence CSE. For example, Hughes and Demo (1989) found that “the strongest influences on self-esteem among all respondents and among the currently employed are quality of family and friendship relations” (p. 146). This is generally consistent with sociometer theory. Research has also shown that self-efficacy is influenced by mastery and reputational factors (e.g., Bandura, 1986; Steese et al., 2006). For example, studies have shown that mastery Ó 2019 Hogrefe Publishing

M. C. Tocci et al., Core Self-Evaluations Over Time

experiences produce higher and stronger generalized expectations of personal efficacy (Bandura, Adams, & Beyer, 1977). Finally, research is also consistent with the notion that neuroticism can be influenced by reputation (Robins, Noftle, Trzesniewski, & Roberts, 2005). Hypothesis 1: CSE varies within individuals over time.

Focal Antecedents As discussed, previous work suggests that factors related to social value and acceptance, mastery experiences, and reputation are likely to influence CSE levels. The current research focused on four such factors as potential antecedents to changes in CSE: occupational support, workrelated recognition, income, and education. Occupational support and work-related recognition are considered work values in the Occupational Information Network (O*NET; see National Center for O*NET Development, n.d.) content model and are based on the Theory of Work Adjustment (Dawis & Lofquist, 1984). Occupational support reflects “occupations that. . .offer supportive management that stands behind employees” (National Center for O*NET Development, n.d., p. 6). Given this definition, occupational support may be related to the notion of social value and acceptance because the behaviors associated with supporting someone likely convey value and acceptance. Indeed, definitions of social support include the concept of social value (Sarason, Levine, Basham, & Sarason, 1983) and social support has been linked to social acceptance (Wentzel, 1994). Supporting an individual is also likely an indirect indication of what you think about that person; thus, the individual who is being supported has an indirect indication of his/her reputation. Given the theory and research reviewed previously, this suggests that greater occupational support should be associated with greater CSE. This idea is also supported by Steese et al. (2006) who found that self-efficacy increased after participants attended a 10-week social support development program. Hypothesis 2: Occupational support positively relates to CSE within individuals over time. Work-related recognition reflects “occupations that. . . offer advancement, potential for leadership, and are often considered prestigious” (National Center for O*NET Development, n.d., p. 5). Given this definition, work-related recognition may relate to mastery and reputation. For instance, entering a prestigious occupation often requires overcoming challenges through perseverance and thus may constitute a mastery experience. Consistent with this idea, Spengler et al. (2015) found that self-reported Ó 2019 Hogrefe Publishing

responsible student behaviors and teacher-rated studiousness predicted later occupational prestige/SES over IQ, parental SES, and educational attainment. This suggests prestigious occupations require effort and perseverance, as this responsibility/studiousness entails hard work and persistence. The recognition and prestige associated with one’s occupation also likely has implications for reputation. This suggests work-related recognition may influence CSE. Support for the idea that work-related recognition influences one’s identity has been reported. For example, Hughes and Demo (1989) showed that social status influenced self-efficacy. Hypothesis 3: Work-related recognition positively relates to CSE within individuals over time. Based on similar reasoning and evidence, we argue that income is related to one’s reputation. In our society, an individual’s income influences how others view that person, where higher income is often associated with more positive evaluations. This suggests income may also be related to CSE. Research has supported this idea (e.g., Hughes & Demo, 1989) by showing that an individual’s belief in oneself (e.g., self-efficacy) is related to income. Hypothesis 4: Income positively relates to CSE within individuals over time. Finally, educational attainment fits the notion of mastery experiences, as obtaining a higher educational degree requires perseverance. Similarly, the level of one’s education should also be an indication of one’s reputation. That is, given the value our society places on education, those with higher levels of education are likely seen as having higher status. This suggests that education should be related to one’s CSE. Hypothesis 5: Education positively relates to CSE within individuals over time.

Method Participants and Procedure This research uses data from the National Longitudinal Survey of Youth 1979 (NLSY79) Child and Young Adult database. The NLSY79 Child and Young Adult database was started in 1986 with information about the children of the female participants from the original 1979 study. These children have been assessed at two-year intervals since 1986. The items measuring CSE were available beginning in 1994. These items were administered again to the same sample in 1996, 1998, 2004, and 2008 (i.e., Journal of Individual Differences (2020), 41(1), 1–7

these items were not administered to the same sample in other years). The dataset thus consisted of these five waves. This dataset consisted of 969 participants (Ns for CSE were 857 in 1996, 534 in 1998, 778 in 2004, and 821 in 2008). In this sample, 51% of participants were male and 30% were non-Black/non-Hispanic. Mean age in 2008 was 30.72 years (SD = 1.69).

M. C. Tocci et al., Core Self-Evaluations Over Time

Income At each wave, participants reported income received from wages, salary, commissions, or tips before deductions for taxes for the previous year. Education At each wave, participants reported the highest grade of school they completed and received credit for. This item was reported as years of school completed on a scale from 0 to 20.

Measures Core Self-Evaluations A 10-item measure was used to assess CSE levels at each wave. These items were chosen based on the scale used by Judge and Hurst (2007). These researchers used 12 items from the NLSY79, reporting that this scale had an internal consistency of .80 and demonstrated convergent validity with a previously validated measure of CSE (the CSES; Judge et al., 2003). Although the present study used a different database (Child and Young Adult, rather than the original NLSY79 database), the same or very similar items were available. Specifically, this database includes nine items from Judge and Hurst’s (2007) scale. In addition, another item (“I can do just about anything I really set my mind to”) is very similar to item number 12 used in Judge and Hurst’s research (“When I make plans, I am almost certain to make them work”). Thus, the present research used these 10 items to assess CSE. Nine items were rated from 1 (= strongly disagree) to 4 (= strongly agree) and one item was rated from 0 (= rarely, none of the time) to 3 (= most, all of the time) (recoded 1–4). Within-person α for this scale was .70. Occupational Support and Work-Related Recognition O*NET data were used to measure occupational support and work-related recognition. At each wave, participants reported their current or most recent job. Census codes were then identified for those jobs and included in the database. We used crosswalks between these census codes and O*NET-SOC codes to identify corresponding O*NET-SOC occupations. Within O*NET, the variable “Support” – involving the extent to which occupations offer supportive management that stands behind employees – was used as the index of occupational support. The variable “Recognition” – involving the extent to which occupations offer advancement, potential for leadership, and are often considered prestigious – was used as the index of workrelated recognition. Scores on these variables range from 1 to 7. Previous research has supported the reliability, validity, and use of these work value scores (Converse, Pathak, DePaul-Haddock, Gotlib, & Merbedone, 2012; McCloy et al., 1999; Rounds, Armstrong, Liao, Rivkin, & Lewis, 2008). Journal of Individual Differences (2020), 41(1), 1–7

Results Before examining the hypotheses, longitudinal measurement invariance was investigated for the CSE measure (see Electronic Supplementary Material, ESM 1, for descriptive statistics and correlations). Based on arguments from Little (2013) regarding parceling advantages and invariance evaluation, 2-item parcels were formed using a balancing approach based on item-scale correlations (item with the highest correlation paired with item with the lowest correlation, next highest and lowest items paired, etc.) and invariance was evaluated using the comparative fit index (CFI) change of .01 criterion. Results supported invariance as the models had good fit (all standardized loadings were above .48, ps < .01) and CFI change was less than .01 (Table 1). Hierarchical linear modeling (with HLM software) was used to examine the hypotheses. For Hypothesis 1, a null model with CSE as the outcome variable was examined. The variance component estimates were 0.095 for level-1 and 0.068 for level-2. These results indicate that 58% of CSE variability was within-person and 42% of CSE variability was between-person. This provides support for Hypothesis 1 by demonstrating that more than half of the variability in CSE was within-person. Two additional analyses were conducted to provide more descriptive information regarding CSE variability. First, the amount of CSE change each individual experienced was assessed by obtaining the difference between each participant’s highest CSE score and his/her lowest CSE score. Analysis of this variable revealed that mean change was 0.59 (SD = 0.32). Second, participant standard deviations for CSE over the Table 1. Longitudinal measurement invariance results Model

RMSEA

RMSEA 90% CI

CFI

TLI

Configural

463.20**

215

.035

.030; .039

.967

.953

Weak

481.03**

231

.033

.029; .038

.966

.956

Strong

547.93**

247

.035

.031; .039

.959

.951

Note. CI = Confidence Interval; CFI = Comparative Fit Index; df = Degrees of Freedom; RMSEA = Root Mean Square Error of Approximation; TLI = Tucker-Lewis Index. **p < .01.

Ó 2019 Hogrefe Publishing

M. C. Tocci et al., Core Self-Evaluations Over Time

five waves were calculated. Analysis of this variable indicated a mean standard deviation of 0.28 (SD = 0.14). To examine Hypotheses 2–5, person-mean centering and robust standard errors were used. We first explored for a linear trend in CSE using age as the predictor. Results indicated a significant linear trend (γ = 0.008, SE = 0.001, t = 7.283, p < .01, pseudo R2 = .02). Thus, age was included as a covariate in all subsequent analyses. Results indicated occupational support was not a significant predictor of CSE (γ = 0.006, SE = 0.010, t = 0.543, p = .587), failing to support Hypothesis 2. Results indicated work-related recognition was not a significant predictor of CSE (γ = 0.002, SE = 0.009, t = 0.275, p = .784), failing to support Hypothesis 3. Results indicated income (recoded by dividing by 10,000) was a significant predictor of CSE (γ = 0.020, SE = 0.005, t = 3.601, p < .01, pseudo R2 = .00), supporting Hypothesis 4. Finally, results indicated education was a significant predictor of CSE (γ = 0.020, SE = 0.006, t = 3.391, p < .01, pseudo R2 = .06), supporting Hypothesis 5. Similar results were obtained when examining income and education together (income: γ = 0.022, SE = 0.011, t = 1.920, p = .055; education: γ = 0.023, SE = 0.006, t = 3.542, p < .01). These findings are consistent with Hypotheses 4 and 5 involving CSE as the outcome, but it is also possible that CSE is the antecedent. To explore both possibilities further, lagged analyses were conducted with CSE as the outcome and as the predictor. Lagged occupational support (γ = 0.014, SE = 0.018, t = 0.789, p = .430), lagged work-related recognition (γ = 0.016, SE = 0.011, t = 1.408, p = .159), and lagged education (γ = 0.010, SE = 0.007, t = 1.318, p = .188) were not significant predictors of CSE. Lagged income was a significant predictor of CSE (γ = 0.017, SE = 0.009, t = 1.959, p = .050, pseudo R2 = .06). Similar results were obtained when examining income and education together (income: γ = 0.028, SE = 0.012, t = 2.277, p < .05; education: γ = 0.011, SE = 0.008, t = 1.377, p = .169). Lagged CSE was not a significant predictor of occupational support (γ = 0.022, SE = 0.062, t = 0.353, p = .724), work-related recognition (γ = 0.049, SE = 0.074, t = 0.655, p = .513), or income (γ = 0.031, SE = 0.070, t = 0.436, p = .662). Lagged CSE was a significant predictor of education (γ = 0.245, SE = 0.117, t = 2.088, p < .05, pseudo R2 = .02). Although somewhat mixed, these findings suggest that CSE may function as both an outcome and a predictor.

Discussion Findings and Implications This study obtained several findings relevant to CSE stability and the hypothesized predictor variables. First, results

Ó 2019 Hogrefe Publishing

indicated that there was more within-person variability than between-person variability in CSE. This provides a novel perspective on CSE and suggests researchers might further investigate how and why this trait fluctuates within individuals over long time periods. Note also that the amount of within-person variability in this study is somewhat larger than that observed in research examining state CSE (e.g., 11% between-day and 25% within-day, Debusscher et al., 2016; 23% within-person, Hofmans et al., 2015). This further supports the notion that research should continue to examine both short-term and long-term variability in CSE. Second, results also indicated that two work-related factors (income and education) had positive relationships with CSE over time. These findings begin to highlight factors that might be targeted in models and interventions focusing on within-person changes in CSE. These findings may have several interesting implications for CSE. For instance, this research shows support for the notion that there is a notable amount of variability in CSE within individuals over time. This finding suggests that there are environmental factors that can and do predict an individual’s CSE. This means that further research should seek to gain a more comprehensive understanding as to the nature and relative importance of these variables. This type of information is important in expanding models of CSE and may have notable applied value. The current research also lends some initial insight into what types of work-related factors may predict CSE. Consistent with the theories guiding this research, results indicated CSE was associated with factors relevant to mastery experiences and reputation. These findings are consistent with the notion that positive mastery- and reputationrelated experiences may have relevance to CSE levels. Note, however, that the lagged analyses also suggested that CSE may be a predictor, such that increases in CSE might precede increases in outcomes (e.g., education). Indeed, it may be that these variables have reciprocal relationships, where positive experiences predict increases in CSE and those increases in turn predict more positive outcomes. Researchers might explore this possibility in greater detail in future studies. In addition, given the relatively small effect sizes observed for the focal predictors, it is also clear that there are additional variables that account for variance in CSE. Thus, studies could analyze more work-related factors (e.g., other prominent work characteristics) as well as non-work factors (e.g., social network support) that may relate to CSE. This may help guide future research and theory related to self-concept stability. Insights pertaining to what key factors can facilitate or inhibit the growth of CSE may also help inform guidelines for increasing positive organizational outcomes shown to be associated with CSE (e.g., job satisfaction and job performance).

Journal of Individual Differences (2020), 41(1), 1–7

Limitations and Future Research Several limitations of this research should be noted. For example, this study did not fully address the question of whether CSE is an outcome, a predictor, or both. Some evidence suggests CSE may operate as both and thus future research might pursue this idea further. Missing data may also be an issue (Little’s MCAR test was significant, p < .01). Future studies involving different datasets may help reveal the extent to which this influenced results. In addition, use of archival data often entails measurement limitations. Additional studies involving more elaborate measures may be useful. Also, the current study used O*NET data to characterize participant occupations. O*NET data may provide an accurate representation of the typical support and recognition associated with each job, but this does not necessarily mean that they are an entirely accurate representation of how each individual experiences these variables. Future research could attempt to address this issue by using other measures of job characteristics (e.g., self-reports). The support and recognition hypotheses also required that participants changed occupations during the study. Out of three opportunities for occupation change (1994–1996, 1996–1998, and 2004–2008; 1998–2004 was excluded due to a change in occupational classification), the average number of changes was 1.29 (SD = 0.74). Thus, participants typically experienced some occupation changes but future studies involving longer time frames might provide stronger tests of these ideas. Additionally, the predictors in this study were examined individually rather than in a full model. Future studies testing larger models may be useful. Finally, future research might also build on this study by examining more elaborate ideas. For instance, studies investigating whether relationships between income or education and CSE are stronger for certain groups (e.g., based on CSE levels, age, or ethnicity) might provide further insights.

Electronic Supplementary Material The electronic supplementary material is available with the online version of the article at https://doi.org/ 10.1027/1614-0001/a000314 ESM 1. Descriptive statistics and correlations

References Bandura, A. (1986). Social foundations of thought and action: A social-cognitive view. Englewood Cliffs, NJ: Prentice-Hall. Bandura, A., Adams, N. E., & Beyer, J. (1977). Cognitive processes mediating behavioral change. Journal of Personality and Social Psychology, 35, 125–139. https://doi.org/10.1037/0022-3514. 35.3.125

Journal of Individual Differences (2020), 41(1), 1–7

M. C. Tocci et al., Core Self-Evaluations Over Time

Caspi, A., Roberts, B. W., & Shiner, R. L. (2005). Personality development: Stability and change. Annual Review of Psychology, 56, 453–484. https://doi.org/10.1146/annurev.psych.55. 090902.141913 Converse, P. D., Pathak, J., DePaul-Haddock, A. M., Gotlib, T., & Merbedone, M. (2012). Controlling your environment and yourself: Implications for career success. Journal of Vocational Behavior, 80, 148–159. https://doi.org/10.1016/j.jvb.2011. 07.003 Converse, P. D., Thackray, M., Piccone, K., Sudduth, M. M., Tocci, M. C., & Miloslavic, S. A. (2016). Integrating self-control with physical attractiveness and cognitive ability to examine pathways to career success. Journal of Occupational and Organizational Psychology, 89, 73–91. https://doi.org/10.1111/ joop.12107 Dawis, R. V., & Lofquist, L. H. (1984). A psychological theory of work adjustment. Minneapolis, MN: University of Minnesota Press. Debusscher, J., Hofmans, J., & De Fruyt, F. (2016). The effect of state core self-evaluations on task performance, organizational citizenship behaviour, and counterproductive work behaviour. European Journal of Work and Organizational Psychology, 25, 301–315. https://doi.org/10.1080/1359432X.2015.1063486 Dóci, E., & Hofmans, J. (2015). Task complexity and transformational leadership: The mediating role of leaders’ state core selfevaluations. The Leadership Quarterly, 26, 436–447. https://doi. org/10.1016/j.leaqua.2015.02.008 Hofmans, J., Debusscher, J., Dóci, E., Spanouli, A., & De Fruyt, F. (2015). The curvilinear relationship between work pressure and momentary task performance: The role of state and trait core self-evaluations. Frontiers in Psychology, 6, 1680. https://doi. org/10.3389/fpsyg.2015.01680 Hughes, M., & Demo, D. H. (1989). Self-perceptions of black Americans: Self-esteem and personal efficacy. American Journal of Sociology, 95, 132–159. https://doi.org/10.1086/229216 Judge, T. A., & Bono, J. E. (2001). Relationship of core selfevaluations traits – self-esteem, generalized self-efficacy, locus of control, and emotional stability – with job satisfaction and job performance: A meta-analysis. Journal of Applied Psychology, 86, 80–92. https://doi.org/10.1037/0021-9010.86.1.80 Judge, T. A., Erez, A., Bono, J. E., & Thoresen, C. J. (2003). The Core Self-Evaluations Scale (CSES): Development of a measure. Personnel Psychology, 56, 303–331. https://doi.org/10.1111/ j.1744-6570.2003.tb00152.x Judge, T. A., Hulin, C. L., & Dalal, R. S. (2012). Job satisfaction and job affect. In S. W. J. Kozlowski (Ed.), The Oxford handbook of organizational psychology (pp. 496–525). New York, NY: Oxford University Press. https://doi.org/10.1093/oxfordhb/ 9780199928309.013.0015 Judge, T. A., & Hurst, C. (2007). Capitalizing on one’s advantages: Role of core self-evaluations. Journal of Applied Psychology, 92, 1212–1227. https://doi.org/10.1037/0021-9010.92.5.1212 Judge, T. A., & Hurst, C. (2008). How the rich (and happy) get richer (and happier): Relationship of core self-evaluations to trajectories in attaining work success. Journal of Applied Psychology, 93, 849–863. https://doi.org/10.1037/0021-9010.93.4.849 Leary, M. R. (1999). Making sense of self-esteem. Current Directions in Psychological Science, 8, 32–35. https://doi.org/ 10.1111/1467-8721.00008 Little, T. D. (2013). Longitudinal structural equation modeling. New York, NY: Guilford Press. McCloy, R., Waugh, G., Medsker, G., Wall, J., Rivkin, D., & Lewis, P. (1999). Determining the occupational reinforcer patterns for O*NET occupational units. National Center for O*NET Development Employment Security Commission, Report (1). Retrieved from https://www.onetcenter.org/dl_files/ORP.pdf

Ó 2019 Hogrefe Publishing

M. C. Tocci et al., Core Self-Evaluations Over Time

National Center for O*NET Development. (n.d.). O*NET Resource Center. Retrieved from https://www.onetcenter.org Roberts, B. W. (2006). Personality development and organizational behavior. Research in Organizational Behavior, 27, 1–40. https://doi.org/10.1016/S0191-3085(06)27001-1 Roberts, B. W., Wood, D., & Caspi, A. (2008). The development of personality traits in adulthood. In O. P. John, R. W. Robins, & L. A. Pervin (Eds.), Handbook of personality: Theory and research (3rd ed., pp. 375–398). New York, NY: Guilford Press. Robins, R. W., Noftle, E. E., Trzesniewski, K. H., & Roberts, B. W. (2005). Do people know how their personality has changed? Correlates of perceived and actual personality change in young adulthood. Journal of Personality, 73, 489–522. https://doi.org/ 10.1111/j.1467-6494.2005.00317.x Rounds, J., Armstrong, P. I., Liao, H., Rivkin, D., & Lewis, P. (2008). Second generation Occupational Value Profiles for the O*NET system: Summary. Raleigh, NC: National Center for O*NET Development. Sarason, I. G., Levine, H. M., Basham, R. B., & Sarason, B. R. (1983). Assessing social support: The Social Support Questionnaire. Journal of Personality and Social Psychology, 44, 127–139. https://doi.org/10.1037/0022-3514.44.1.127 Schinkel, S., van Dierendonck, D., & Anderson, N. (2004). The impact of selection encounters on applicants: An experimental study into feedback effects after a negative selection decision. International Journal of Selection and Assessment, 12, 197–205. https://doi.org/10.1111/j.0965-075X.2004.00274.x Spengler, M., Brunner, M., Damian, R. I., Lüdtke, O., Martin, R., & Roberts, B. W. (2015). Student characteristics and behaviors at age 12 predict occupational success 40 years later over and above childhood IQ and parental socioeconomic status.

Ó 2019 Hogrefe Publishing

Developmental Psychology, 51, 1329–1340. https://doi.org/ 10.1037/dev0000025 Steese, S., Dollette, M., Phillips, W., Hossfeld, E., Matthews, G., & Taormina, G. (2006). Understanding Girls’ Circle as an intervention on perceived social support, body image, self-efficacy, locus of control, and self-esteem. Adolescence, 41, 55–74. Wentzel, K. R. (1994). Relations of social goal pursuit to social acceptance, classroom behavior, and perceived social support. Journal of Educational Psychology, 86, 173–182. https://doi. org/10.1037/0022-0663.86.2.173 Wu, C., & Griffin, M. A. (2012). Longitudinal relationships between core self-evaluations and job satisfaction. Journal of Applied Psychology, 97, 331–342. https://doi.org/10.1037/a0025673 Zacher, H. (2014). Career adaptability predicts subjective career success above and beyond personality traits and core selfevaluations. Journal of Vocational Behavior, 84, 21–30. https:// doi.org/10.1016/j.jvb.2013.10.002 History Received October 24, 2017 Revision received November 20, 2018 Accepted November 26, 2018 Published online November 26, 2019 Michael C. Tocci School of Psychology Florida Institute of Technology 150 W. University Blvd. Melbourne, FL 32901 USA tocci.mc@pg.com

Journal of Individual Differences (2020), 41(1), 1–7

Original Article

Identification of Emotions in Offenders With Antisocial Personality Disorder (ASP) Behavioral and Autonomic Responses Depending on the Reinforcement Scheme Catarina Iria1, Fernando Barbosa2, and Rui Paixão1 1

Faculty of Psychology and Education Sciences, University of Coimbra, Portugal

Faculty of Psychology and Education Sciences, University of Porto, Portugal

Abstract: A group of offenders with antisocial personality (ASP) and a control group identified facial expressions of emotion under three conditions: monetary reward, monetary response cost, and no contingency, to explore effects on the antisocial offenders’ deficits commonly reported in these tasks. Skin Conductance Responses (SCRs) indexed emotional arousal. Offenders with ASP performed worse than controls under reward and no contingency conditions, but under the response-cost condition results were similar. The offenders with ASP presented higher SCR than the controls in the two monetary conditions. Findings suggest that offenders with ASP are hypersensitive to monetary contingencies; monetary reward seems to interfere negatively in their performance while monetary response cost improves it. Arousal level seems unable to explain ability to identify facial affects, while results suggest that methodological variations may explain the conflicting results in the literature. Keywords: offenders with antisocial personality disorder, facial affect recognition, skin conductance response, electrodermal activity, monetary reward

Since deficits in the identification of emotions could be related to failures in social competence and the genesis of criminal behavior (Blair, Peschardt, Budhani, Mitchell, & Pine, 2006; Dodge, Laird, Lochman, & Zelli, 2002; van Goozen, Fairchild, Snoek, & Harold, 2007; Howner et al., 2011), psychobiologic models have proposed differences in the ability to identify facial expressions in offenders and non-offenders (Dolan & Fullam, 2006; Marsh & Blair, 2008). In fact, the ability to identify facially expressed emotions seems to be associated to individual or group specificities (Iria & Barbosa, 2009; Pham & Philippot, 2010) and populations with an antisocial lifestyle (people with psychopathy and borderline personality disorder, criminal recidivists, and those suffering from behavioral disorders) frequently exhibit deficits in identifying emotions from facial expressions in general (Hastings, Tangney, & Stuewig, 2008), or negative expressions, such as fear (Blair et al., 2004) and anger (Marsh & Blair, 2008). Nevertheless, some studies have been unable to demonstrate this deficit (Book, Quinsey, & Langford, 2007; Glass & Newman, 2006; Pfabigan, Alexopoulos, & Sailer, 2012). Journal of Individual Differences (2020), 41(1), 8–16 https://doi.org/10.1027/1614-0001/a000298

More recently Iria, Barbosa, and Paixão (2015) have found that offenders with antisocial personality (ASP) could perform as well as controls under experimental conditions where monetary response cost was expected or no contingencies were applied, but not when rewards were in play, suggesting that antisocial offenders seem to have a negative hypersensitivity to monetary reward that may interfere with their ability to recognize emotions. The majority of these studies have only been focused on the cognitive facet of this deficit, simply measuring the ability to perceive the emotional state of others via emotion recognition tasks, using facial expressions (Bons, Scheepers, Rommelse, & Buitelaar, 2010). Nevertheless, a recent investigation has suggested the importance of studying the cognitive and emotional components of empathy in forensic populations (Domes, Hollerbach, Vohs, Mokros, & Habermeyer, 2013). So, it is also necessary to study the processing of facial expressions at the emotional level, evaluating the experience of emotions consistent with, and in response to, those of others by the autonomic response to emotional facial expressions (Bons et al., 2010). Research on emotion Ó 2019 Hogrefe Publishing

C. Iria et al., Facial Affect Identification By Antisocial Offenders

has shown that autonomic responses, among other measures, can offer reliable indices about emotional reactions (Khalfa, Peretz, Blondin, & Manon, 2002). Skin Conductance Response (SCR) is an autonomic index directly under the control of the sympathetic system (Dawson, Schell, & Filion, 2000), and is based on a low current passing through the skin using a bipolar placement of electrodes (Mendes, 2008). This allows us to measure rapid modifications in the sweating activity of the skin, due to the activity of the sweat glands, which results from the cholinergic stimulation of the sympathetic neurons (Khalfa et al., 2002). Facial expressions induce emotional responses in individuals (Marsh, Ambady, & Kleck, 2005), including autonomic arousal (Blair & Cipolotti, 2000) that can be indexed through skin conductance. Moreover, low autonomic nervous system activity is known as a biomarker for aggressive and antisocial behavior (Baker, Shelton, Baibazarova, Hay, & van Goozen, 2013). In fact, a hyporeactive autonomic system regarding emotive material (including facial expressions) may compromise successful social and emotional processing, contributing to a failure in social behavior (Heims, Critchley, Dolan, Mathias, & Cipolotti, 2004). Nevertheless, contradictory data have been found in specific offender samples, for example, suggesting a physiological hyper-reactivity in those who physically abuse children (McCanne & Hagstrom, 1996) and in people with antisocial personality disorder (Iria, Barbosa, & Paixão, 2010). More recently, a neuroimaging study also found contradictory data suggesting that offenders have enhanced neural processing of fearful faces in the amygdala as well as in other facial processing brain areas, compared to controls (Howner et al., 2011). Also, Seidel et al. (2013) found that offenders show reduced physiological responses specifically during the identification of fear and disgust. Raine (2005) suggested that the low autonomic response is linked to social conditions and mostly characterizes antisocial offenders of a high socioeconomic status. In order to clarify these data, further exploration must be done, considering which conditions may modulate the deficit that offenders with ASP seem to show in processing the facial expressions of emotion. To this end, we assessed the performance and the arousal of antisocial offenders and controls in experimental tasks of identification of facial expressions of emotion using three contingencies: monetary reward, response cost, and no contingency. Thus, a poorer accuracy of participants with ASP regarding the identification of emotions in the reward condition was expected, both in comparison with other conditions and controls. It was also hypothesized that offenders with ASP would show higher physiological response under this condition, as a manifestation of their hypersensitivity to monetary reward. Ó 2019 Hogrefe Publishing

Methods Participants Eighty-one adult males were recruited by advertising in four cities in the central region of Portugal and participated in this study. Forty-one were offenders with ASP (Antisocial Group – AG) and 40 participants comprised the control group (Control Group – CG). However, due to technical problems the physiological data of three participants with ASP and four control participants were not recorded (AG: n = 38; CG: n = 36). As general inclusion criteria, participants had to be male, and between 20 and 65 years. Participants with ASP in particular, had to have two or more convictions leading to prison and meet the criteria for Antisocial Personality Disorder (ASPD), as defined by the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV; American Psychiatric Association, 2000). Participants recruited for the control group stated that they had never engaged in law-breaking behavior and the criminal records of all participants were checked after due authorization. Exclusion criteria were: a history of neurological or psychiatric problems; significant sensory or motor impairments; intellectual disabilities; a maladaptive pattern of alcohol or drug use in the previous six months, or the use of medication that could interfere with the results; and the consumption of coffee or cigarettes three hours before the experiment. Exclusion criteria were screened by a semi-structured interview, conducted by a senior clinical psychologist. The two groups had a low socioeconomic status, also assessed during the interview, and were statistically matched in relation to gender (all male), age, number of years of education, and intellectual ability, as shown in Table 1. All participants accepted to take part in the experiment (none was excluded) and provided written informed consent.

Materials NimStim Dataset Thirty-six pictures from the Portuguese validation of NimStim (NimStim-PT; Iria, Paixão, & Barbosa, 2008) were used. The selected pictures comprised color photographs on a white background showing male and female actors expressing the six basic emotions (Ekman, 1972): Happiness, Sadness, Anger, Fear, Surprise, and Disgust (six photographs per emotional category). Only pictures with 50–60% of correct identification in the Portuguese normative study (Iria et al., 2008) were selected, as a means to control the difficulty of the task between emotional categories.

Journal of Individual Differences (2020), 41(1), 8–16

C. Iria et al., Facial Affect Identification By Antisocial Offenders

Table 1. Demographic characteristics AG (n = 41)

CG (n = 40)

Age

39.1 (9.38)

37.6 (11.5)

0.644

.521

Intellectual Ability (IA test)

16.1 (3.66)

16.2 (3.46)

0.126

.900

Education (years)

7.63 (2.95)

8.25 (3.14)

0.916

.363

Criminal typology (n) Number of convictions Months of imprisonment

t value p value

28 Pp; 36 Pt; 13 St; 32 Dg 2.66 (0.79) 81.8 (42.4)

Notes. AG = antisocial offender group; CG = control group; n = number of participants; Pp = crimes against people (murder, grievous bodily harm, domestic violence, kidnapping, illegal restraint, rape, trafficking of human beings); Pt = crimes against patrimony (theft, organized or armed robbery, swindling, racketeering and extortion, handling stolen goods); St = crimes against society (forgery of administrative documents, counterfeiting of currency, forgery of means of payment, arson); Dg = illicit trafficking in narcotic drugs and psychotropic substances. All data show mean values (± SD) or number of cases.

Psychophysiological Device A computerized polygraph system, model I-330-C2 from JJ Engineering (Poulsbo, WA) was used to record the psychophysiological data. Semi-Structured Interview In the recruitment process a semi-structured interview was performed by a clinical psychologist to assess the socioeconomic status, comorbidities, neurological problems and significant sensory or motor impairments, intellectual disabilities, a maladaptive pattern of alcohol or drug use in the previous 6 months, the use of medication, and the consumption of coffee or cigarettes before the experiment. The following instruments were also applied in this interview: – Structured Clinical Interview for DSM-IV-TR for Axis II Disorders – SCID-II (First, Gibbon, Spitzer, Williams, & Benjamin, 1997). It is considered a standard interview widely used in psychiatric research disorders of Axis II, according to DSM-IV-TR. The Portuguese translation (Pinto-Gouveia, Rijo, Matos, Castilho, & Salvador, 2011) of the module for Antisocial Personality Disorder was used in this study. – Structured Clinical Interview for DSM-IV-TR for Axis I Disorders – SCID-I – Research Version (First, Gibbon, Spitzer, & Williams, 2002). The modules for psychotic disorders and psychotic symptoms were used. – IA Test. This is an intelligence test for abstract reasoning. It is a shortened version of Raven’s Matrices, standardized to the Portuguese population with a sample of 3,228 participants, of which 2,937 are males (Amaral, 1966). It comprises five series (A, B, C, D, and E) of six items each (30 in total). In each item, the participant is asked to identify the missing segment (between six and eight alternative segments) which is required to Journal of Individual Differences (2020), 41(1), 8–16

complete a larger pattern. Each correct response corresponds to one point, to a maximum of 30 points.

Procedure Participants were shown a set of facial expressions of emotions while monitoring the skin conductance under three different monetary contingencies – reward, response cost, no contingency. The pictures were administered on a notebook (14.100 screen) placed 1 m in front of the participants, running Superlab V4.0 (2008, Cedrus Corp., San Pedro, CA). Data were collected in a laboratory room with controlled temperature (20–25 °C). All telephones and others electrical devices were turned off during the electrodermal recordings. Each picture was presented once for 10 s with an equal inter stimuli interval (ISI) of 10 s, during which a fixation point was displayed. Participants had a peripheral keypad with a key for each of the six basic emotions, and were instructed to click on the key where the corresponding emotion was written, as fast as they could. The photographs were presented sequentially, without any kind of feedback. The 3 experimental conditions (reward, response cost, no contingency) were randomly presented in blocks, and 12 equivalent pictures of facial expressions of the emotions (2 pictures for each emotion) were shown per condition. The pictures in each condition were also randomized. Under the reward condition, participants earned €0.50 for every emotion correctly identified, and could totalize an amount of €6.00 at the end of the block. Under the response cost condition, €6.00 were distributed to the participants at the beginning of the block, and €0.50 was removed from that amount for each emotion incorrectly identified. Under the no contingency condition, no money was used. All subjects received detailed instructions about the experimental conditions at the beginning of the task. The instruction for the reward condition was: “Your task is to identify the emotion that each person is feeling in each photo and click the corresponding button as quickly as possible. You will receive €0.50 for each correct answer. At the end of the series you can see how many correct answers you had and the corresponding amount you have won.” For the response cost condition: “Your task is to identify the emotion that each person is feeling in each photo and click the corresponding button as quickly as possible. Before you start you are offered a personal account with €6.00. For every incorrect answer €0.50 will be subtracted from your personal account. At the end of the series you can see how many wrong answers you had and you will get the amount that remains. Notice that if all the answers are wrong, your personal account will be €0.00.” For the no contingency condition: “Your task is to identify the Ó 2019 Hogrefe Publishing

C. Iria et al., Facial Affect Identification By Antisocial Offenders

emotion that each person is feeling in each photo and click the corresponding button as quickly as possible.” Psychophysiological Protocol (Skin Conductance) Skin conductance electrodes were attached to the distal phalanges of the index and middle fingers of the nondominant hand, allowing participants to use their dominant hand to press the keypad. The electrodermal activity (EDA) amplifier was connected to a laptop computer running Physiolab USE3 software (JJ Engineering, Poulsbo, WA) in Windows XP, which was used for the electrodermal recordings. The I-330-cz system contained a built-in optical interface to prevent electrical hazards. After the electrodes had been attached, the participant sat in a resting state for four minutes in order to record the tonic level. Immediately after this initial period, participants performed the facial expression recognition task under the three above-mentioned conditions. Skin conductance was measured throughout the experiment using a lowpass filter of 5 Hz to remove high frequency noise. Data were collected for 12 min after which recording was terminated and the electrodes were removed. The data collected were then stored and recalled for later analysis by means of functions created in Excel. We used a phasic analysis instead a tonic analysis because we were interested in the autonomic responses in a temporal window of 1–10 s after the presentation of the stimulus (Boucsein, 1992). The area under the curve (AUC) in that temporal window was the measure we selected for the phasic analysis. As AUC includes both the response amplitude and the rising/declining time, other researchers suggest that this composite measure may be more effective than any of those parameters alone (Boucsein, 1992; Naqvi & Bechara, 2006) and is more suitable for automatic analysis (Figner & Murphy, 2011). We computed the AUC using the trapezoidal rule for the standardized results and we chose a larger than usual temporal window after the stimulus onset, so as to ensure the inclusion of the time delay of the response (usually less than or equal to 3 s), the rising time (usually 1–3 s, Boucsein et al., 2012) and the period of decline. The formula utilized was:

Area ¼ h=2 y1 þ 2 sum y2 . . . yn 1 þ yn ;

where h is the space between points, and y1 . . . yn all the points considered (where 1 and n indicate the first and last points, respectively). This area is divided by the duration of the time window in seconds, resulting in μS/s units.

Statistical Analysis Repeated measures analysis of variance (ANOVA) were computed with Experimental Conditions (Reward condition, Response Cost condition, No Contingency condition) serving as a within-subject factor, and Group (Antisocial Offenders, Controls) as a between-subject factor. A post hoc Tukey HSD test for multiple comparisons was applied for the analyses of specific effects. Normality and homogeneity assumptions were verified in all cases. Behavioral and physiological data were independently analyzed. The physiological measure was the Mean AUC. The behavioral measure was the total Number of Errors and the Reaction Time, that is, the time delay (in seconds) between the stimulus display and the moment of response. All analyses were performed using the software Statistica, Version 8.0 (2008, StatSoft Inc., Tulsa, OK).

Results Number of Errors and Reaction Time Mean and standard errors were computed for the two groups (Antisocial Group and Control Group) for the total number of errors and for the reaction time. Results can be seen in Table 2. In respect to the total number of errors and taking all the emotions together there is a main effect for Group, F(1, 78) = 4.59, p = .035, ηp2 = .056 with antisocial offenders committing more errors (M = 2.91, SE = 0.20) than controls (M = 2.31, SE = 0.20). Also, there is a main effect for Experimental Condition, F(2, 156) = 3.29, p = .039, ηp2 = .041. Post hoc analyses show that under the No Contingency condition (M = 2.73, SE = 0.21, p = .043) and Reward condition (M = 2.80, SE = 0.16, p = .046) participants committed more errors than under the Response Cost condition (M = 2.30, SE = 0.19). There is also an effect Experimental

Table 2. Mean (M) and standard errors (SE) for number of errors and for reaction time (s) of the antisocial group (AG) and control group (CG) in the reward condition, response cost condition, and no contingency condition Reward condition Measures

Response cost condition

No contingency condition

AG (n = 41) M (SE)

CG (n = 40) M (SE)

AG (n = 41) M (SE)

CG (n = 40) M (SE)

AG (n = 41) M (SE)

CG (n = 40) M (SE)

Total number of errors

3.32 (0.22)

2.25 (0.22)

2.22 (0.26)

2.38 (0.27)

3.17 (0.29)

2.30 (0.29)

Reaction time (in seconds)

3.83 (1.39)

3.67 (1.41)

3.75 (1.42)

3.54 (1.44)

3.88 (1.50)

3.45 (1.42)

Ó 2019 Hogrefe Publishing

Journal of Individual Differences (2020), 41(1), 8–16

C. Iria et al., Facial Affect Identification By Antisocial Offenders

controls (M = 3.56, SE = 1.34), or for Experimental Condition, F (2, 158) = 1.34, p = .160, ηp2 = .017 as reaction times for the Reward condition (M = 3.75, SE = 0.99), Response Cost condition (M = 3.65, SE = 1.01) and No Contingency condition (M = 3.66, SE = 1.07) were all similar. There is no effect for the interaction Group Experimental Conditions, F(2, 158) = 2.10, p = .173, ηp2 = .026.

4.00

3.50

Number of errors

3.00

2.50

Mean AUC of Skin Conductance Response

2.00

1.50

1.00 Monetary Reward

No Contigency

Monetary Response Cost CONDITIONS

Antisocial offenders Controls

Figure 1. Number of errors by group in monetary reward, monetary response cost, and no contingency conditions (vertical bars denote .95 confidence intervals).

Condition Group, F(2, 158) = 4.96, p = .008, ηp2 = .060, that can be seen in Figure 1. Post hoc analyses show there is no within-group effect in the controls. Conversely, offenders with ASP commit more errors under the Reward condition (p = .002) and No Contingency condition (p = .016) than under the Response Cost condition. Otherwise, expected punishment for false responses increased the performance of participants with ASP compared to the other conditions. Finally, between-group analyses show that the antisocial group commits more errors under the Reward condition than the controls under any of the three experimental conditions (all p < .05). Also, the antisocial group under the No Contingency condition performs worse than the controls under the Response Cost condition (p = .037), but the difference to controls under the Reward and No Contingency conditions is marginally not significant (p = .072 and p = .057, respectively). In the response cost condition, the antisocial group scored at the level of the control group, for the same and the other conditions (all p > .05). In relation to the Reaction Time, taking all the emotions together there is no main effect either for Group, F (1, 79) = 1.95, p = .207, ηp2 = .042 with the antisocial group (M = 3.82, SE = 1.32) exhibiting a similar reaction time to the

Mean AUC of SCR in Reward, Response Cost, and No Contingency was computed for the antisocial group (AG) and the control group (CG) considering all the emotions together (Happiness, Sadness, Anger, Fear, Disgust, and Surprise). Results may be seen in Table 3. We found a main effect for Group, F(1, 72) = 9.95, p = .002, ηp2 = .122, with antisocial offenders showing higher mean AUC of the SCR (M = 3.03, SE = 0.23) than controls (M = 2.00, SE = 0.23). Also, there was a main effect for Experimental Condition, F(2, 144) = 12.70, p < .001, ηp2 = .150. Post hoc tests show that the mean AUC under No Contingency condition (M = 2.25, SE = 0.16) is lower than under the Reward (M = 2.66, SE = 0.18, p < .001) and Response Cost (M = 2.64, SE = 0.17, p < .001) conditions. There was also an effect for the interaction Experimental Conditions Group, F(2, 144) = 6.73, p = .002, ηp2 = .085. Post hoc tests clarified that the mean AUC of SCR in antisocial offenders is lower in the No Contingency condition when compared with the Reward and Response Cost conditions (both p < .001), while the controls’ SCR does not seem to be affected by any of the conditions (all p > .05). The mean AUC of SCR of antisocial offenders under the Reward condition and the Response Cost condition are higher than for the controls under the three conditions (all p < .01). No other differences were found. Results could be seen in Figure 2.

Discussion In this study we compared the performance and the physiological response (skin conductance) of antisocial offenders with controls in a task of identifying basic emotions through facial expression under three experimental

Table 3. Mean (M) and standard error (SE) of SCR’s area under the curve (AUC) in reward, response cost and no contingency for antisocial group (AG) and control group (CG) considering all emotions together (Happiness, Sadness, Anger, Fear, Disgust, and Surprise) Reward condition Measures SCR’s mean AUC (μS/s)

Response cost condition

No contingency condition

AG (n = 38) M (SE)

CG (n = 36) M (SE)

AG (n = 38) M (SE)

CG (n = 36) M (SE)

AG (n = 38) M (SE)

CG (n = 36) M (SE)

3.30 (0.25)

2.02 (0.26)

3.21 (0.24)

2.06 (0.24)

2.57 (0.23)

1.92 (0.23)

Journal of Individual Differences (2020), 41(1), 8–16

Ó 2019 Hogrefe Publishing

C. Iria et al., Facial Affect Identification By Antisocial Offenders

4.00

3.50

Number of errors

3.00

2.50

2.00

1.50

1.00 Monetary Reward

No Contigency

Monetary Response Cost

Antisocial offenders Controls

CONDITIONS

Figure 2. Mean magnitude of SCR (AUC) by group in monetary reward, monetary response cost, and no contingency conditions (vertical bars denote .95 confidence intervals).

conditions – monetary reward, monetary response cost, and no contingencies – in order to examine if the monetary contingency modulates the results. We hypothesized that antisocial offenders show hypersensitivity to rewards and that the expectation of getting them negatively interferes with their performance while identifying emotions from facial expressions. This supports the idea that antisocial offenders may have a hypersensitivity to reward (Bjork, Chen, Smith, & Hommer, 2010; Buckholtz et al., 2010), in some cases with a negative effect (Iria et al., 2015) and provides some explanation for the ambiguous results from previous research on antisocial individuals. However, these results only partially confirmed our hypothesis. In this study, the antisocial offenders exhibited a poorer accuracy rate for facial expressions under the Reward condition compared with the controls, and they performed equally badly when there were No Contingencies modulating their responses at the within-group level. Notably, the possibility of losing money was associated with a better accuracy of the antisocial offenders when identifying emotions and their performance raised to the same level of the controls. Simultaneously, the Reward and Response Cost conditions induced similarly higher levels of arousal in the antisocial offenders, as indexed by SCR, when compared to the No Contingency condition and to the Controls under the three conditions. On the other hand, the SCR of the antisocial offenders under the No Contingency condition could not be distinguished from the controls’ SCR under the three experimental conditions. These results seem to indicate that the antisocial offenders were more sensitive to monetary contingencies than the controls and that Ó 2019 Hogrefe Publishing

monetary contingencies lead to a higher arousal in the antisocial offenders. It is possible to consider these results from the point of view of Yerkes-Dodson’s law (Yerkes & Dodson, 1908) that relates performance to arousal, which takes performance as an inverse U-shaped function of the arousal level. The unsuccessful performance of under-aroused people is explained in motivational terms, while the performance of over-aroused people is understood in terms of Easterbrook’s hypothesis, this is, in high arousal, attention tends to be concentrated on the dominant and most obvious aspects of the situation. Further, high arousal makes the capacity to separate relevant from irrelevant features less effective, which increases a predisposition to the allocation policy (Kahneman, 1973). So, it is possible that the high arousal and low performance of the antisocial offenders under the Reward condition could be due to an eventual interference of monetary reward that concentrates their attention on that particular aspect. This may push the task of recognizing facially expressed emotions into the background. However, when the antisocial offenders already have the monetary reward in their possession, this may actually help them focus on the task. This possible explanation is supported by the data showing that the antisocial offenders have a higher SCR under the Response Cost condition than the controls, although they perform in this situation just as well as the Controls under any of the three experimental conditions. If we interpret these results on the basis of an impulsivity pattern due to difficulties in the inhibition process (Fowles, 1987), that characterizes the forensic population in general (He, Cassaday, Howard, Khalifa, & Bonardi, 2011), antisocial offenders under the No Contingency condition do not have enough motivation to inhibit this automatic response, but having the monetary reward in their possession seems to be a sufficient incentive to inhibit the impulsive response, while the prospect of obtaining the monetary reward interferes with the inhibitory process concerning the impulsive response. However, this thesis was not supported by the present results because there are no significant differences in the response reaction time. The absence of differences between groups in reaction times may be due to some characteristics of the experimental procedure, such as the lack of training trials, or the complexity of the response (press one of the keypad’s six keys), which could inhibit automatic responses. These findings may also be considered in light of the revised Reinforcement Sensitivity Theory – RST (Gray, 1981, 1987; Gray & McNaughton, 2000). The physiological results suggest that the Fight/Flight/Freeze System/ Behavioral Inhibition System (FFFS/BIS) of the antisocial participants are reactive to response-costs, and the Behavioral Activation System (BAS) seems very reactive to Journal of Individual Differences (2020), 41(1), 8–16

rewards, but these participants are not as activated or motivated under the No Contingency condition, which may partly explain their poor performance in this case. However, the two monetary conditions produced similar activation with different behavioral consequences: in a conflicting situation, such as the Response Cost condition, antisocial offenders seem to be FFFS/BIS reactive, whereby they succeed in increasing their attention on the facial affect stimulus, performing as well as the controls. While under the Reward condition, and probably because the antisocial offenders are hyper oriented toward the reward, the prospect of obtaining a monetary reward negatively interferes in the task, as their attention is focused on the reward and not the facial expression. These results support the thesis that the BAS, in predisposed individuals, could elicit antisocial tendencies (Gray, 1970, 1972, 1982, 1990) and the results of the previous study by Taylor, Reeves, James, and Bobadilla (2006) that found that Antisocial Personality Disorder is associated with a strong BAS. However, participants with ASP in our study, different to the ones in this previous study, seem to have a normal BIS. Our results seem to indicate that there is not a true correspondence between the psychophysiological response and the quality of the cognitive/behavioral response, giving support to the thesis that variations of autonomic arousal may not totally explain emotional processing (Heims et al., 2004) and emphasizing the importance of monetary contingency as a differential motivator for antisocial individuals and controls. However, our results need to be considered with caution, since they do not take the level and the type of psychopathy of the samples into account. In fact, previous studies have shown that there are different group specificities in the identification of facial emotions in people with psychopathy and criminal behavior (Iria, Barbosa, & Paixão, 2012). The consideration of this variable will be necessary in future research. This study has some limitations, such as the lack of a control group of offenders without antisocial personality disorder and a control group of antisocial individuals that had never been imprisoned. These would be important in order to analyze the specific contributions of antisocial traits and imprisonment within the results that we found in this study. Indeed, because the antisocial individuals are the only ones who have already been arrested, it remains unclear whether personality traits related to antisocial personality, or the fact of having been imprisoned are the main factors influencing the results. The potential mediating effects of imprisonment should be clarified by a study of antisocial individuals who have never been institutionalized. In further developments of this study it is advisable to introduce specific measures of impulsivity, psychopathy,

Journal of Individual Differences (2020), 41(1), 8–16

C. Iria et al., Facial Affect Identification By Antisocial Offenders

and anxiety, since these variables may have some sort of effect on the results. Another limitation is the impossibility of controlling, by chemical analysis, whether the participants had, or not, a maladaptive pattern of alcohol or drug use. Other limitation is the choice of the area under the curve for the measure of psychophysiological responses. In a 10-second time window, several different fluctuations may occur, where phasic responses may be combined with nonspecific or spontaneous fluctuations (Gläscher & Adolphs, 2003). Another limitation is the reduced number of pictures for each emotion under each experimental condition, which prevented us from doing an analysis for each particular emotion. Also, in future studies the No Contingency condition will need to be placed first in the series of conditions to prevent any eventual contamination of the monetary contingency conditions over the one without such contingencies. Whether higher monetary values (under the Monetary Contingency condition) would modulate the performance and physiological response of the controls in the same way as the antisocial offenders’ performance should be assessed in future studies. In addition, the value of the participants’ monthly or annual earnings needs to be included in the variables studied, in order to be able to control this variable more precisely. The sample needs to be increased in order to see if the marginal and non significant differences of the performance of the antisocial group under the No contingencies condition becomes or not significant when compared with the control group under the No contingency and Reward conditions. As general conclusions the data of this exploratory study suggest that the monetary Response Cost had a positive effect on the performance of the antisocial group. However, while the monetary Response Cost seems to raise the performance of the antisocial offenders up to the level of the controls, the monetary Reward seems to have no positive effect on the performance of the antisocial group. The results from this study are important to clarify the mechanism by which monetary conditions may influence the behavior of antisocial individuals in tasks of identification of facial expressions of emotion. They provide a possible explanation for some types of behavior where the recognition of facial expressions and an expectation of monetary reward are involved in the offence. In addition, these results draw attention to the importance of response cost schemes in managing the behavior of offenders with antisocial personality disorder. Also, these results warn researchers of the relevance of different reward/response cost schemes when making methodological decisions concerning studies with these individuals and show the need for further studies on this issue.

Ó 2019 Hogrefe Publishing

C. Iria et al., Facial Affect Identification By Antisocial Offenders

References Amaral, J. (1966). Aferição do Teste IA (Escala Reduzida das Matrizes Progressivas de JC Raven) [Testing IA (Reduced Scale of Progressive Matrices by JC Raven)]. Lisboa, Portugal: Centro de Estudos Psicopedagógicos da Escola Técnica Eugénio dos Santos. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed. text revision). Washington, DC: American Psychiatric Association. Baker, E., Shelton, K. H., Baibazarova, E., Hay, D. F., & van Goozen, S. H. (2013). Low skin conductance activity in infancy predicts aggression in toddlers 2 years later. Psychological Science, 24, 1051–1056. https://doi.org/10.1177/0956797612465198 Bjork, J. M., Chen, G., Smith, A. R., & Hommer, D. W. (2010). Incentive-elicited mesolimbic activation in adolescents with externalizing disorders. Journal of Child Psychology and Psychiatry, 51, 827–837. https://doi.org/10.1111/j.1469-7610.2009. 02201.x Blair, R. J. R., & Cipolotti, L. (2000). Impaired social response reversal: A case of “acquired sociopathy”. Brain, 123, 1122–1141. https://doi.org/10.1093/brain/123.6.1122 Blair, R. J., Mitchell, D. G., Peschardt, K. S., Colledge, E., Leonard, R. A., Shine, J. H., . . . Perrett, D. I. (2004). Reduced sensitivity to others’ fearful expressions in psychopathic individuals. Personality and Individual Differences, 37, 1111–1122. https://doi. org/10.1016/j.paid.2003.10.008 Blair, R. J., Peschardt, K. S., Budhani, S., Mitchell, D. G., & Pine, D. S. (2006). The development of psychopathy. Journal of Child Psychology and Psychiatry, 47, 262–276. https://doi.org/ 10.1111/j.1469-7610.2006.01596.x Bons, D., Scheepers, F., Rommelse, N., & Buitelaar, J. (2010). Motor, emotional and cognitive empathic abilities in children with autism and conduct disorder. Proceeding of Measures Behaviors, 24, 109–113. https://doi.org/10.1007/s10802-012-9689-5 Book, A. S., Quinsey, V. L., & Langford, D. (2007). Psychopathy and the perception of affect and vulnerability. Criminal Justice and Behavior, 34, 531–544. https://doi.org/10.1177/ 0093854806293554 Boucsein, W. (1992). Electrodermal activity. New York, NY: Plenum Press. Boucsein, W., Fowles, D. C., Grimnes, S., Ben-Shakhar, G., Roth, W. T., Dawson, M. E., & Filion, D. L. (2012). Publication recommendations for electrodermal measurements. Psychophysiology, 49, 1017–1034. https://doi.org/10.1111/j.14698986.2012.01384.x Buckholtz, J. W., Treadway, M. T., Cowan, R. L., Woodward, N. D., Li, R., Ansari, M. S., . . . Zald, D. H. (2010). Dopaminergic network differences in human impulsivity. Science, 329, 532. https://doi.org/10.1126/science.1185778 Dawson, M. E., Schell, A. M., & Filion, D. L. (2000). The electrodermal system. In J. T. Cacioppo, L. G. Tassinary, & G. G. Berntson (Eds.), Handbook of psychophysiology (2nd ed., pp. 200–223). New York, NY: Cambridge University Press. Dodge, K. A., Laird, R., Lochman, J. E., & Zelli, A. (2002). Multidimensional latent-construct analysis of children’s social information processing patterns: Correlations with aggressive behaviour problems. Psychological Assessment, 14, 60–73. https://doi.org/10.1037/1040-3590.14.1.60 Dolan, M., & Fullam, R. (2006). Face affect recognition deficits in personality-disordered offenders: Association with psychopathy. Psychological Medicine, 36, 1563–1570. https://doi.org/ 10.1017/S0033291706008634 Domes, G., Hollerbach, P., Vohs, K., Mokros, A., & Habermeyer, E. (2013). Emotional empathy and psychopathy in offenders: An

Ó 2019 Hogrefe Publishing

experimental study. Journal of Personality Disorders, 27, 67–84. https://doi.org/10.1521/pedi.2013.27.1.67 Ekman, P. (1972). Universals and cultural differences in facial expressions of emotion. In J. Cole (Ed.), Nebraska Symposium on Motivation 1971 (Vol. 19, pp. 207–283). Lincoln, NE: University of Nebraska Press. Figner, B., & Murphy, R. O. (2011). Using skin conductance in judgment and decision making research. In M. SchulteMecklenbeck, A. Kuehberger, & R. Ranyard (Eds.), A handbook of process tracing methods for decision research (pp. 163–184). New York, NY: Psychology Press. First, M. B., Gibbon, M., Spitzer, R. L., & Williams, J. B. W. (2002). Structured clinical interview for DSM-IV-TR Axis I Disorders, research version, patient edition with psychothic screen (SCID-I/ P W/PSY SCREEN). New York, NY: Biometrics Research, New York State Psychiatric Institute. First, M. B., Gibbon, M., Spitzer, R. L., Williams, J. B. W., & Benjamin, L. S. (1997). Structured Clinical Interview for DSMIV Axis II Personality Disorders (SCID-II). Washington, DC: American Psychiatric Press. Fowles, D. C. (1987). Application of a behavioural theory of motivation to the concepts of anxiety and impulsivity. Journal of Research in Personality, 21, 417–435. https://doi.org/ 10.1016/0092-6566(87)90030-4 Gläscher, J., & Adolphs, R. (2003). Processing of the arousal of subliminal and supraliminal emotional stimuli by the human amygdala. The Journal of Neuroscience, 23, 10274–10282. https://doi.org/10.1523/JNEUROSCI.23-32-10274.2003 Glass, S. J., & Newman, J. P. (2006). Recognition of facial affect in psychopathic offenders. Journal of Abnormal Psychology, 115, 815–820. https://doi.org/10.1037/0021-843X.115. 4.815 Gray, J. A. (1970). The psychophysiological basis of introversionextraversion. Behaviour Research and Therapy, 8, 249–266. https://doi.org/10.1016/0005-7967(70)90069-0 Gray, J. A. (1972). The psychophysiological basis of introversionextraversion: A modification of Eysenck’s theory. In V. D. Nebylitsyn & J. A. Gray (Eds.), The biological basis of individual behaviour (pp. 182–205). New York, NY: Academic Press. Gray, J. A. (1981). A critique of Eysenck’s theory of personality. In H. J. Eysenck (Ed.), A model for personality (pp. 246–276). Berlin, Germany: Springer. Gray, J. A. (1982). The neuropsychology of anxiety: An enquiry into the functions of the septo-hippocampal system. New York, NY: Oxford University Press. Gray, J. A. (1987). The psychology of fear and stress. New York, NY: Cambridge University Press. Gray, J. A. (1990). Brain systems that mediate both emotion and cognition. Cognition & Emotion, 4, 269–288. https://doi.org/ 10.1080/02699939008410799 Gray, J. A., & McNaughton, N. (2000). The neuropsychology of anxiety: An enquiry into the functions of the septo-hippocampal system (2nd ed.). Oxford, UK: Oxford University Press. Hastings, M., Tangney, J., & Stuewig, J. (2008). Psychopathy and identification of facial expressions of emotion. Personality and Individual Differences, 44, 1474–1483. https://doi.org/10.1016/ j.paid.2008.01.004 He, Z., Cassaday, H. J., Howard, R. C., Khalifa, N., & Bonardi, C. M. (2011). Impaired Pavlovian conditioned inhibition in offenders with personality disorders. The Quarterly Journal of Experimental Psychology, 64, 2334–2351. https://doi.org/10.1080/ 17470218.2011.616933 Heims, H. C., Critchley, H. D., Dolan, R., Mathias, C. J., & Cipolotti, L. (2004). Social and motivational functioning is not critically dependent on feedback of autonomic responses: Neuropsychological evidence from patients with pure

Journal of Individual Differences (2020), 41(1), 8–16

autonomic failure. Neuropsychologia, 42, 1979–1988. https:// doi.org/10.1016/j.neuropsychologia.2004.06.001 Howner, K., Fischer, H., Dierks, T., Federspiel, A., Wahlund, L., Jonsson, T., . . . Kristiansson, M. (2011). Brain processing of fearful facial expression in mentally disordered offenders. Journal of Behavioral and Brain Science, 1, 115–123. https:// doi.org/10.4236/jbbs.2011.13016 Iria, C., & Barbosa, F. (2009). Perception of facial expressions of fear: Comparative research with criminal and non-criminal psychopaths. Journal of Forensic Psychiatry & Psychology, 20, 67–73. https://doi.org/10.1080/14789940802214218 Iria, C., Barbosa, F., & Paixão, R. (2010). Psychophysiological arousal in a task of identification of emotions in facial expressions in antisocials. International Journal of Psychophysiology, 77, 274–274. https://doi.org/10.1016/j.ijpsycho.2010.06.120 Iria, C., Barbosa, F., & Paixão, R. (2012). The identification of negative emotions through a Go/No-Go task. European Psychologist, 17, 291–299. https://doi.org/10.1027/1016-9040/ a000101 Iria, C., Paixão, R., & Barbosa, F. (2008). Avaliação neuropsicológica das emoções no contexto da saúde: contributo do NimStim-Pt [Neuropsychological assessment of emotions in the context of health: the contribution of NimStim-PT]. In I. Leal, J. P. Ribeiro, I. Silva, & S. Marques (Eds.), Actas do 7º Congresso Nacional de Psicologia da Saúde (pp. 9–10). Lisboa, Portugal: Instituto Superior de Psicologia Aplicada. Iria, C., Barbosa, F., & Paixão, R. (2015). Reward expectation modulates the identification of facial expressions of emotion in antisocial offenders. Psychiatry, Psychology and Law, 22, 417–424. https://doi.org10.1080/13218719.2014.960028 Kahneman, D. (1973). Attention and effort. Englewood Cliffs, NJ: Prentice-Hall. Khalfa, S., Peretz, I., Blondin, J.-P., & Manon, R. (2002). Eventrelated skin conductance responses to musical emotions in humans. Neuroscience Letters, 328, 145–149. https://doi.org/ 10.1016/S0304-3940(02)00462-7 Marsh, A. A., Ambady, N., & Kleck, R. E. (2005). The effects of fear and anger facial expressions on approach- and avoidancerelated behaviors. Emotion, 5, 119–124. https://doi.org/ 10.1037/1528-3542.5.1.119 Marsh, A. A., & Blair, R. (2008). Deficits in facial affect recognition among antisocial populations: A meta-analysis. Neuroscience and Biobehavioral Reviews, 32, 454–465. https://doi.org/ 10.1016/j.neubiorev.2007.08.003 McCanne, T. R., & Hagstrom, A. H. (1996). Physiological hyperreactivity to stressors in physical child abusers and individuals at risk or being physically abusive. Aggression and Violent Behavior, 1, 345–358. https://doi.org/10.1016/S1359-1789(96)00004-3 Mendes, W. B. (2008). Assessing autonomic nervous system reactivity. In E. Harmon-Jones & J. Beer (Eds.), Methods in the neurobiology of social and personality psychology (pp. 118–147). New York, NY: Guilford Press.

Journal of Individual Differences (2020), 41(1), 8–16

C. Iria et al., Facial Affect Identification By Antisocial Offenders

Naqvi, N. H., & Bechara, A. (2006). Skin conductance: A psychophysiological approach to the study of decision making. In C. Senior, T. Russell, & M. S. Gazzaniga (Eds.), Methods in mind: The study of human cognition (pp. 103–122). Cambridge, MA: MIT Press. Pfabigan, D. M., Alexopoulos, J., & Sailer, U. (2012). Exploring the effects of antisocial personality traits on brain potentials during face processing. PLoS One, 7, e50283. https://doi.org/ 10.1371/journal.pone.0050283 Pham, T. H., & Philippot, P. (2010). Decoding of facial expressions of emotions in criminal psychopaths. Journal of Personality Disorders, 24, 445–459. https://doi.org/10.1521/pedi.2010.24.4.445 Pinto-Gouveia, J., Rijo, D., Matos, A. P., Castilho, P., & Salvador, M. C. (2011). Entrevista Clínica Estruturada para as Perturbações de Personalidade do Eixo II do DSM-IV [Structured Clinical Interview for DSM-IV Axis II Personality Disorders]. Lisboa, Portugal: Climespi. Raine, A. (2005). The interaction of biological and social measures in the explanation of antisocial and violent behavior. In D. M. Stoff & E. J. Susman (Eds.), Developmental psychobiology of aggression (pp. 13–42). Cambridge, UK: Cambridge University Press. https://doi.org/10.1017/CBO9780511499883.003 Seidel, E. M., Pfabigan, D. M., Keckeis, K., Wucherer, A. M., Jahn, T., Lamm, C., & Derntl, B. (2013). Empathic competencies in violent offenders. Psychiatry Research, 210, 1168–1175. https://doi.org/10.1016/j.psychres.2013.08.027 Taylor, J., Reeves, M., James, L., & Bobadilla, L. (2006). Disinhibitory trait profile and its relation to Cluster B personality disorder features and substance use problems. European Journal of Personality, 20, 271–284. https://doi.org/10.1002/per.585 van Goozen, S. H., Fairchild, G., Snoek, H., & Harold, G. T. (2007). The evidence for a neurobiological model of childhood antisocial behavior. Psychological Bulletin, 133, 149–182. https://doi. org/10.1037/0033-2909.133.1.149 Yerkes, R. M., & Dodson, J. D. (1908). The relation of strength of stimulus to rapidity of habit-formation. Journal of Comparative Neurology of Psychology, 18, 459–482. https://doi.org/10.1002/ cne.920180503 History Received December 16, 2013 Revision received March 16, 2019 Accepted March 19, 2019 Published online June 17, 2019 Catarina Iria Rua Carlos Alberto Pinto Abreu 141, cave direita 3040-245 Coimbra Portugal catarinairia@gmail.com

Ó 2019 Hogrefe Publishing

Original Article

A New Money Behavior Quiz Adrian Furnham1

and Simmy Grover2

Department of Leadership and Organisational Behaviour, Norwegian Business School (BI), Nydalsveien, Oslo, Norway

Research Department of Clinical, Educational and Health Psychology, University College London, UK

Abstract: This study reports on the development of a new questionnaire to measure money behaviors devised by the Financial Times (London). In all, 402 participants from diverse backgrounds, who were recruited online, completed the 29-item questionnaire. Six a priori money types were identified by financial experts, who did not know the salient psychological literature. The internal reliability of the factors was modest and there was some evidence of sex differences. Exploratory factor analyses failed to confirm the six-factor model, but did provide an alternative and interpretable typology. Further step-wise regression analysis showed the simple question: “Are you a spender or a saver?” was strongly related to almost every factor. Gender, age, and self-perceived wealth were also consistently correlated with the money types. Implications and limitations are discussed. Keywords: money, attitudes, questionnaire, saving, spending

Professionals who give money advice have noticed dramatic differences in adults’ attitudes to, interest in, and behavior concerning money. People clearly differ dramatically in their financial literacy, risk taking, and money habits which can have very serious consequences for their well-being (Klontz, Britt, Archuleta, & Klontz, 2012; Klontz, Britt, Mentzer, & Klontz, 2011; Klontz, Seay, Sullivan, & Canale, 2014; Lay & Furnham, 2018). This study set about developing and validating a new measure in conjunction with the Financial Times (January 2017) who were running a large piece and associated quiz on financial literacy. It is different from other measures as the typology was generated by financial experts rather than psychometricians in the academic world. Inevitably there is some overlap between their categorical scheme and those working in the area (Taylor, Klontz, & Britt, 2016), though they maybe much more familiar with nuanced ideas and practices than the pure academic researchers. Second, the study set out to determine whether simple issues like a person’s rating of their overall wealth, general success at work, or whether they were self-categorized as spender versus saver would account for a significant amount of the variance for each money factor/type. There is no shortage of questionnaires attempting to assess money-related attitudes and behaviors (Furnham, 1984; Furnham, Wilson, & Telford, 2012; Klontz et al., 2011; Lay & Furnham, 2018; Lim & Teo, 1997; Rose & Orr, 2007; Tang, 1992; Taylor, Klontz, & Britt, 2016; Yamauchi & Templer, 1982). This study reports on a new test designed by money journalists, and consultants who did this without a knowledge of the psychological literature, but considerable experience of dealing with investors and savers with widely different incomes and financial knowlÓ 2019 Hogrefe Publishing

edge and experience. The central question was whether the measure could provide a robust measure of the “types” as determined by the financial experts. Table 1 shows the various scales used to develop money beliefs and some findings from relevant studies. Studies which have attempted to replicate the factor structure have not always been successful. Further, they have tended to show that only one or two factors (usually referring to power and security) were correlated as hypothesized with other money-related beliefs and behaviors. Table 1 shows that most of the studies used students and that, as always, most (but not all) studies were done in America and Europe, though some recent studies have been reported from other English speaking countries (Furnham & Murphy, 2018).

Money Attitude Correlates Money attitudes have been linked with many demographic variables (Furnham, 1996). Studies have found money attitudes related to gender (Furnham et al., 2012; Furnham, von Stumm, & Fenton-O’Creevy, 2014; Gresham & Fontenot, 1989; Klontz et al., 2011, 2014; Tang, 1992), culture (Burgess, 2005; Lynn, 1991; Medina, Saegert, & Gresham, 1996), education level (Furnham, 1984, Klontz et al., 2011), political and religious values (Furnham et al., 2012; Tang, 1992). Previous results suggest that males tend to associate money with Achievement, Power, and Freedom (Furnham et al., 2012) more than women, who in turn are more inclined to see money as a source of anxiety, as well as associate money with retention (Gresham & Fontenot, 1989) and budgeting (Tang, 1992). Furnham (1984) and

Journal of Individual Differences (2020), 41(1), 17–29 https://doi.org/10.1027/1614-0001/a000299

Journal of Individual Differences (2020), 41(1), 17–29 344, 291, and 328

MBBS

Modified MBBS

Modified MAS

MBBS

MES

Money in the Past and Future Scale

MES

MBBS

MBBS MAS

MAS

Money in the Past and Future Scale Modified MAS

MAS

MBBS

KMSI

Furnham (1984)

Bailey and Gustafson (1986)

Gresham and Fontenot (1989)

Bailey and Gustafson (1991)

Hanley and Wilhelm (1992)

Tang (1992)

Bailey and Lown (1993)

Tang (1995)

Wilhelm, Varese, and Friedrich (1993)

Bailey, Johnson, Adams, Lawson, Williams, and Lown (1994)

Lim and Teo (1997)

Roberts and Sepulveda (1999)

Ozgen and Bayoglu (2005)

Engelberg and Sjoberg (2006)

Christopher, Marek, and Carroll (2004)

Klontz, Britt, Mentser, and Klontz (2011)

Burgess (2005)

559

MAS

Yamauchi and Templer (1982)

422

204

212

221

300

273

200

68 and 249

654

769

143

472

557

256

300

533

MSD

Wernimont and Fitzpatrick (1972)

Scale used

Empirical studies

Adults

Students

Swedish students

Urban South Africans

Turkish students

Adults

Students

Employed adults related to college students

Adult Americans

College students, faculty, managers, etc. College students, their relatives, and other professionals College students

College students and their parents College students

College students

Adults from different professions College students

College students, engineers, religious sisters, etc.

Sample

American

Sweden

Major Metropolitan Cities

Ankara, Turkey

Mexican

Singapore

USA, Australia, Canada

USA

Taiwan

Western US States

Phoenix, Tucson, Denver, and Detroit Middle Tennessee City

US South-western City

US South-western Cities

US South-western City

England, Scotland and Wales

Los Angeles and Fresno, CA

Large US Midwestern City

Location

Table 1. Empirical studies: methodological characteristics and demographic and personality factors that do and do not influence money attitudes

(Continued on next page)

Sex, age, race Education, gross income

Materialism

Emotional intelligence

Values and culture

Gender, age, family type

Compulsive buying

Gender differences

Geographical location

Gender, financial progress

Age, income, work ethic, social, political, and religious values Age

Compulsive behavior

Sensitivity and emotional stability

Gender

Income, gender, age, and education

Work experience, socio-economic level and gender

Factors that influence money attitudes

18 A. Furnham & S. Grover, A New Money Behavior Quiz

Ó 2019 Hogrefe Publishing

KMSI = Klontz Money Script Inventory; KMSI-R = Klontz Money Script Inventory – Revised; MSD = Modified Semantic Differential; MAS = Money Attitude Scale; MBBS = Money Beliefs and Behavior Scale; MES = Money Ethic Scale; NMAQ = New Money Attitudes Questionnaire; SMTM = Short Money Type Measure.

Sex, age, ideology, work success British

Sex, age, education, etc. American

Adults 268 NMAQ Lay and Furnham (2018)

Ó 2019 Hogrefe Publishing

326 KMSI-R Taylor, Klontz, and Britt (2016)

Students

Education, income, financial habits British 109,472 SMTM von Stumm, Fenton-O’Creevy, and Furnham (2013)

Adults

Age, ethnicity, salary, education, politics English Adults

Social capital Russian Adults

400 SMTM

634 MPPS

Furnham, Wilson, and Telford (2012)

Location

Tatarko and Schmidt (2012)

Table 1. (Continued)

N Scale used Empirical studies

Sample

Factors that influence money attitudes

A. Furnham & S. Grover, A New Money Behavior Quiz

Furnham et al. (2012) both found associations between money beliefs and socio-political ideology. While for political and religious values, Furnham et al. (2012) found those who are more affiliated with right wing are more likely to endorse power and freedom related emotions toward money. There is also considerable evidence that other factors correlate with money attitudes and beliefs (Furnham, 2014). Some like financial anxiety have been extensively researched by Klontz and colleagues (Klontz et al., 2011; Taylor et al., 2017). Others include subjective ratings of success and happiness, as well as the simple categorization of whether one is a spender or a saver. In this study we use a number of other questions like political beliefs, psychological stability and a simple statement about whether people are “money worriers” to further examine attitudinal correlates. Each has been identified as relating to money attitudes in the literature (Furnham, 2014) but there is a paucity of data on these issues. The data suggest that the happier and healthier a person rates themselves to be the more they associate money with Achievement and Power and the less they are concerned with Savings and poor Financial Literacy. Similarly more religious people have been shown to be more “conservative” in their money attitudes seeing money as an index of achievement and means of power.

This Study This study uses a (radical) new measure which was not informed by the academic literature but focus groups and the insights of financial advisors. It examines the psychometric properties of this new test as well as individual correlates. Lay people tend to typologize others with rich and popular descriptions, very unlike the approach of differential psychologists interested in the topic (Furnham, 2014). This study posed the question of whether this approach would yield a psychologically valid and different test, compared to the others available (see Table 1). The first aim of the present study is to devise and evaluate a new measure of money attitudes in an effort to understand people’s relationship with money. The measure was designed for the Financial Times for a large feature on money (Rovnick, 2017). Rather than relying on the academic literature the items and types were generated by financial journalists and consultants whose jobs involved giving people advice about their money. Inevitably the types are described more in current journalistic language rather than more traditional academic terminology. Further, in some instances they seem to contradict the academic literature on money attitudes, though the differences are very subtle. Journal of Individual Differences (2020), 41(1), 17–29

They use their typologies to be able to quickly differentiate between various individuals they deal with in a number of different ways. The following were their summary types, described in their own language. (1) The Fitbit Financier, who checks their balance twice a day, always switches to best deals and only buys goods online. They obsess over credit card points, use many comparison sites and apps that track budget and mortgages; (2) The Anxious Trader, who tries always to buy low and sells high and thinks more trading makes things better. They trade frequently and hence have high levels of charges; (3) The Social Value Spender, often a woman, who buys expensive gifts for herself and others as a way of feeling like a better person, more accepted, and so forth. They make purchases to boost their selfesteem often with debt problems; (4) The Cash Splasher, a close cousin of the social value spender, who pays for everyone’s meal in a restaurant and boasts about the value of home/car in order to feel appreciated. Often male, they view themselves as generous but use money primarily to make others think more highly of them; (5) The Hoarder, who wants to have £100 k in the bank and will possibly use pension freedoms to take money out of pension in order to get cash, which is then put into a current account. They do not like to invest in stock market as would rather earn interest on cash, which is small but guaranteed, instead of risking a loss; (6) The Ostrich, who never opens a bill or bank-statement and finds doing nothing much more palatable to making a decision. These types were derived from various interviews held in London. The author took the descriptions of each type to develop statements. These were piloted for their clarity and “approved” by some. (a) Achievement and Success is the idea that money is an obvious and comparable sign of achievement. Money is a measure of success and self-worth: earning lots of money gives people a sense of achievement. The amount of money one has is simple but powerful index of success in life. Those who score high will be very motivated to make a lot of money. (b) Power and Prestige concerns the idea that money is something that many people strive for, and are happy to show off, if and when they acquire it. Money gives social power and capital and is a major source of social status. Moreover, people look up to those who have lots of money. It is the idea that people enjoy respect by displaying their wealth: money is a major source of power and also prestige. This attitude is different from the former in the sense that this is a more other-oriented attitude, while the former is more of a self-focused attitude. (c) Mindful and Responsible. This idea concerns the fact that many people think of saved and invested money Journal of Individual Differences (2020), 41(1), 17–29

A. Furnham & S. Grover, A New Money Behavior Quiz

as a form of protection against the vicissitudes of life. Money can protect one because people can buy things they really need (a home, healthcare) and not be beholden to others who might want to control them in some way. People with these beliefs tend to save money through means like budgeting and paying bills to avoid being fined. They are, and pride themselves in, being money responsible. They do not “flash” their wealth and money; and tend to be well informed. (d) Savings Concerns. Money is a source of various types of stress, the main one being if there is not enough of it. This attitude is characterized by worries and anxieties, primarily in the association with savings. People with this money attitude are worried about savings; they fear that their savings are never enough or that they will run out of money. The whole topic of finance and money is associated with anxiety and depression. They ask for and seek out help. (e) Financial Literacy Worries. This refers to feeling ignorant about financial affairs. The main source of worry and anxiety for individuals with this money attitude is their low comprehension of financial issues. They often feel anxious and ashamed to talk to others about financial affairs. However, they know they need help, often seeking out professionals. (f) Money Denial and Duplicity. People with these beliefs try to deal with their money worries by denying them. Rather than confront their money problems they try to block them out and avoid all discussions about the topic. This can take a more serious form when people lie about their finances to friends and relations or commit “sins” of omission or co-mission when providing financial data to others. Of those who were interviewed to confirm they represented the beliefs and behaviors of those hypothetical types. The second aim is to investigate the demography, ideology, biographical information, and subjective well-being correlates of money attitudes. There is scarce evidence examining the possible associations of money attitudes in terms of biographical information. Hence, we set out to explore the relationship between money habits (being a spender vs. a saver) and one’s money type.

Method Sample 268 participants (148 male, 120 female) participated in the study. Their mean age was 37.43 (SD = 12.75 years, range of Ó 2019 Hogrefe Publishing

A. Furnham & S. Grover, A New Money Behavior Quiz

18–77). 59.3% of the participants were from the US (N = 159), 36.6% were from India (N = 98), the rest were from Canada and the United Kingdom. In terms of ethnicity, 47.4% were White (N = 127), 43.7% Asian (N = 117), 3.4% Black (N = 9), the rest identified themselves as other ethnicities. With regard to education, 14.2% completed high school (N = 38), 14.2% obtained a diploma equivalent level (N = 38), 50.0% have a Bachelor’s degree (N = 134), and 21.6% achieved a Master’s degree or a PhD (N = 58). This means they are a highly educated sample.

and this would ensure the test could be used in many different countries. There was no overall significant difference in the responses from the different countries. In all, 300 people were sampled but the number was reduced because of missing data and checks on the time participants took to complete the survey, as an index of careless responding.

Measures

Gender Differences

Money Attitudes Measure This measure consists of 29 items with questions regarding participant’s attitudes toward money. All responses were answered on a 7-point Likert scale ranging from 1 (= strongly disagree) to 7 (= strongly agree). The measure is proposed to have six factors: Fitbit Financier (four items; α = .61); Anxious Trader (four items; α = .65); Social Value Spender (six items; α = .56); Cash Splasher (five items; α = .73); Hoarder (five items; α = .53); and Ostrich (five items; α = .65). Items can be seen in Table 2.

A one-way ANOVA was conducted using IBM SPSS statistics (Version 21) to examine whether gender differences were present in the typologies but also the underlying 29 items.

Biographical Information, Ideology, and Subjective Well-Being Various single-item questions were designed to assess participants’ biographical information, ideologies and wellbeing. Participants rated on a 100-point Likert scale ranging from 1 (= not at all) to 100 (= extremely wealthy) how rich they were (M = 42.79, SD = 23.32) as well as their work success from 1 (= not at all) to 100 (= very successful) (M = 57.28, SD = 26.42). Participants were also asked on a binary (forced choice) scale whether they are a spender (34.2%) or a saver (65.8%), and whether they had a happy childhood (Yes = 73.9%; No = 26.1%).

Analysis

Factor Structure To begin with a Confirmatory Factor Analysis (CFA) was conducted to examine whether the a priori hypothesized six-factor solution was a good fit for the data. This analysis was conducted using the lavaan package in R. The lavaan package in R has a cfa command and when the inputted data are specified as ordered lavaan treats the data as ordinal endogenous variables and estimates polychoric correlations and uses a more robust estimator for the test statistics. The robust estimator “WLSM” was utilized, as this is an appropriate estimator when dealing with data that violate multivariate normality. This estimator uses Diagonally Weighted Least Squares (DWLS) estimation with robust standard errors and a scaled test-statistic (Satorra–Bentler scaled).

Results Demographic Questions Participants provided information regarding their gender, age, ethnicity, and education level.

Procedure The participants were recruited via Amazon Mechanical Turk (MTurk), an online market for enlisting workers to participate in research and surveys. Data from MTurk have been found to be comparable with traditional recruitment methodologies in terms of reliability, while the diversity of the samples surpasses those from standard Internet surveys and student samples (Buhrmester, Kwang, & Gosling, 2011; Paolacci, Chandler, & Ipeirotis, 2010). We chose to sample people primarily from India and the United States Ó 2019 Hogrefe Publishing

Correlations and Descriptive Statistics The Appendix shows the full correlations between all 29 items. Table 3 shows the correlation matrix of the typologies. It indicated modest correlations between the scales with two r > .60. Fitbit Financier was correlated r = .63 with Anxious Trader; while Social Value Spender was correlated r = .62 with Cash Splasher.

Gender Differences A one-way ANOVA suggested that males scored higher than females on Anxious Trader (Mmale = 17.66, SD = 4.60; Mfemale = 16.68, SD = 4.09; F[1, 399] = 4.96, p < .05) and Cash Splasher (Mmale = 17.52, SD = 5.25; Mfemale = 15.56, Journal of Individual Differences (2020), 41(1), 17–29

A. Furnham & S. Grover, A New Money Behavior Quiz

Table 2. The results of factor analysis and one-way ANOVA M (SD) Factor

Fitbit financier

Anxious trader

Social value spender

The Cash Splasher

The Hoarder

The Ostrich

Items

.61

I get a real kick out of the business of managing my money

4.11 (1.73)

3.86 (1.73)

2.07

I think I check my financial affairs more than other people

4.42 (1.62)

4.31 (1.73)

0.42

There are lots of money bargains if you are prepared to search for them I spend a lot of time trying to find money bargains

5.24 (1.38)

5.55 (1.10)

6.11*

4.52 (1.67)

5.17 (1.36)

18.47***

I follow the trends about money management

3.98 (1.68)

3.50 (1.54)

8.87**

I am constantly re-evaluating all my investments

4.27 (1.72)

4.03 (1.60)

2.00

You have to be vigilant about all money matters

5.30 (1.35)

5.49 (1.18)

2.25

I believe investing time in watching money programs is worth it When it comes to spending money on myself its “because I am worth it” I often demonstrate my love to people by buying them presents I am very generous with the people I love

4.11 (1.65)

3.70 (1.67)

6.07*

4.08 (1.71)

3.79 (1.63)

3.07

3.71 (1.55)

3.84 (1.71)

0.63

4.93 (1.38)

5.58 (1.15)

The best present you can give to someone is money

3.31 (1.73)

3.01 (1.69)

Money can help you be accepted by others

4.42 (1.48)

4.34 (1.67)

I love retail therapy: shopping to cheer me up

3.01 (1.67)

3.75 (1.81)

17.13*** 5.73*

.65

.56

.73

.53

.65

Male

Female

25.94*** 3.02 0.20

Having a lot of money is a sign of success

4.67 (1.64)

4.28 (1.60)

I rather enjoy letting people know how well-off I am

2.69 (1.51)

2.26 (1.40)

8.36**

I use money to persuade people to do things for me

2.77 (1.64)

2.13 (1.41)

17.65***

I admit that I buy things to impress others

2.80 (1.59)

2.55 (1.52)

2.73

You get respect from others when you have lots of money

4.58 (1.55)

4.45 (1.57)

0.74

I feel safe and secure if I have a lot of money saved

5.46 (1.36)

5.59 (`.37)

I prefer to be safe rather than a gambler when it comes to money I value having a lot of easy-to-access money in the bank

5.05 (1.42)

5.81 (1.20)

4.92 (1.49)

4.86 (1.46)

0.15

I am much more of a saver than a spender

4.78 (1.61)

4.65 (1.68)

0.64

Essentially I am risk-averse when it comes to money investments Thinking about money makes me anxious.

4.49 (1.52)

4.65 (1.51)

1.09

4.12 (1.65)

4.76 (1.69)

I dither a lot over money decisions.

4.01 (1.55)

4.22 (1.52)

1.75

I am really not interested in money matters.

3.09 (1.71)

3.00 (1.57)

0.28

I prefer to let others I trust make my important money decisions I feel anxious and defensive about my personal finances

2.82 (1.65)

2.73 (1.60)

0.37

4.07 (1.80)

4.33 (1.67)

2.26

0.86 33.95***

14.42***

Note. *p < .05, **p < .01, ***p < .001.

SD = 5.28; F[1, 399] = 13.29 p < .001). Indeed the size of three of the correlations is close to the size of the Cronbach’s α, which indicates that all the systematic construct relevant variance is shared by these scales. Table 2 shows the items for each scale and sex differences for each of them. Eleven of the 29 items showed a sex difference. The greatest sex difference was found for the following item: “I prefer to be safe rather than a gambler when it comes to money”. The model fit statistics for the six-factor model were poor: CFI = .728, TLI = .694, RMSEA = .173 (lower bound Journal of Individual Differences (2020), 41(1), 17–29

= .169 and upper bound = .177). Additionally some items loaded poorly on their expected factors. Consequently, an exploratory factor analysis (EFA) was conducted on the data. To determine the number of factors to extract two tests were used: Velicer’s minimum average partial (MAP) criteria and a parallel analysis on the polychoric correlation matrix. Both were conducted in R using the psych package (Revelle, 2017). Velicer’s MAP suggested a four-factor solution, however parallel analysis suggested a six-factor solution. Consequently, an exploratory factor analysis was conducted extracting four, five, and six factors. Table 4 Ó 2019 Hogrefe Publishing

A. Furnham & S. Grover, A New Money Behavior Quiz

Table 3. Intercorrelations between the six proposed factors

1. Fitbit financier 2. Anxious trader

.63***

.38***

.30***

.43***

.03

.37***

.41***

.35***

.05

.62***

.17***

.20***

.12*

.30***

3. Social value spender 4. The Cash Splasher 5. The Hoarder

.06

6. The Ostrich Note. *p < .05, **p < .01, ***p < .001.

shows the factor loadings of each item from the four, five, and six-factor EFA. Regardless of the number of factors some items were found to load poorly (< .40) on any of the factors and others had significant cross-loading (loading .30 or greater on another factor). These items were dropped and a CFA was conducted for a four, five, and six-factor solution. For each factor solution modification indices were inspected and any items with significant cross variance (high modification indices) were dropped. The model fit statistics for each solution were as follows: Four factor: CFI = .975, TLI = .966, RMSEA = .076 (.067–.086) and Five Factor: CFI = .961, TLI = .947, RMSEA = .085 (.077–.094). The Six Factor was not identified and the correlations between the six factors were found to be greater than one. This suggests that a six-factor model is a poor fit for the underlying data (Bentler, 1990; Bentler & Bonnet, 1980). We decided to do two sets of regressions. In the first group we used the original scales despite the poor alpha reliability and the poor model fit statistics from the CFA for this model. In the second group of regressions we used the four factors derived from the four-factor model as this model showed the best fit for the underlying data (see Figure 1). Additionally the four-factor model aligns from a theoretical perspective in to four factors that describe the following typologies: Fitbit, Status, Splasher, and Anxious. The items that form the fifth factor in the five-factor solution do not link with one another theoretically: “There are lots of money bargains if you are prepared to search for them” and “I am very generous with the people I love”. From Figure 1, it is clear the items on the first factor (Fitbit) seemed to be concerned with money obsessionality and focus, the second (Status) is about the respect and what money represents to others, the third (Splasher) is related to a desire to flash cash and with using money to influence others, and the fourth (Anxious) encompasses the nervousness and discomfort that is associated with financial decision making. Based on the factor analysis items loading > .30 on all four factors were combined to form new typologies. Alpha (α) for these new factors was: Fitbit = .568; Status = .771; Splasher = .799; and Anxious = .749. Ó 2019 Hogrefe Publishing

Six stepwise regressions were then computed with the six money types as the criterion variable and three groups of predictor variables: sex, age, and class, ratings of wealth and work success and two general questions (spender vs. saver) and happy childhood. The final step is shown in Tables 5 and 6. The results showed that some variables like education, success at work, and happy childhood were unrelated to any of the six types. On the other hand, sex, age, wealth, and spender versus saver were significant. The regression which accounted for most of the variance (15%) indicated that younger males, who felt more wealthy, and were self-confessed spenders, were more likely to be Cash Splashers. Two variables predicted whether one was a Hoarder: female savers. Similarly, for Ostriches: these were younger spenders. Social Value spenders were younger females, who considered themselves wealthy, spenders. Only one variable was related to being an Anxious Trader: people who rated themselves more wealthy tended to be anxious traders. Finally, Fitbit Financiers were more likely to be young, female savers who gave higher ratings for their wealth.

Discussion As Table 1 indicates there exist around half a dozen questionnaires designed to measure money beliefs and behaviors some of which have been used in many studies. They are similar in many ways though there is still debate about the number and naming of the underlying dimensions. This study tested a new measure designed not by psychometricians but financial advisors/consultants/journalists using focus group methods. It resulted in a major article in the Financial Times which attracted a good deal of attention. The resulting questionnaire had six money types but the α failed to reach the generally acceptable .70 threshold except in one case. Indeed, various factor analyses confirmed the fact that the money types were not as coherent as they might be yielding four identifiable typologies. There could be many reasons for this including non-professionals’ relative lack of ability to write items Journal of Individual Differences (2020), 41(1), 17–29

A. Furnham & S. Grover, A New Money Behavior Quiz

Table 4. Item loadings from EFA with four, five, and six factors F1

.287

.805

.008

.293

.244

.767

.029

.268

.045

.807

.210

.008

.229

.018

.026

.015

.565

.120

.022

.014

.472

.110

.026

.185

.530

.050

.054

.047

.099

.040

.318

.366

.037

.217

.254

.185

.057

.172

.472

.217

.285

.059

.139

.394

.117

.083

.493

.157

.290

.054

.359

.215

.259

.354

.377

.045

.207

.206

.294

.134

.375

.740

.143

.031

.304

.751

.092

.002

.052

.759

.340

.086

.019

.056

.078 .058

.233

.680

.012

.035

.209

.614

.011

.030

.127

.665

.159

.055

.036

.058

.393

.321

.189

.166

.382

.215

.197

.157

.200

.260

.412

.136

.214

.084

.078

.291

.551

.027

.041

.233

.561

.024

.013

.065

.582

.238

.007

.010

.090

.013

.402

.292

.206

.089

.449

.205

.128

.126

.227

.250

.355

.132

.121

.241

.129

.299

.197

.148

.127

.378

.069

.028

.074

.364

.083

.342

.082

.045

.440

.034

.218

.181

.039

.002

.093

.037

.158

.092

.652

.028

.153

.109

.226

.722

.074

.463

.140

.153

.112

.388

.236

.239

.155

.284

.222

.450

.261

.097

.219

.086

.122

.163

.720

.054

.131

.175

.718

.052

.033

.168

.122

.758

.039

.044

.107

.395

.059

.235

.123

.534

.226

.089

.045

.426

.171

.393

.107

.037

.431

.200

.187

.027

.714

.040

.153

.008

.794

.060

.177

.019

.143

.787

.035

.142

.072

.790

.230

.088

.008

.751

.276

.089

.001

.113

.288

.730

.112

.021

.044

.104

.774

.070

.040

.027

.714

.176

.079

.054

.262

.170

.721

.099

.041

.187

.098

.722

.126

.102

.044

.740

.114

.043

.022

.060

.145

.667

.055

.026

.098

.182

.137

.040

.698

.057

.109

.026

.760

.076

.143

.026

.136

.804

.025

.068

.175

.190

.231

.431

.103

.178

.143

.436

.095

.140

.190

.214

.398

.113

.080

.079

.443

.035

.030

.236

.434

.021

.044

.230

.111

.077

.292

.145

.000

.185

.481

.061

.286

.510

.001

.038

.184

.492

.012

.178

.255

.140

.429

.075

.095

.064

.182

.537

.208

.146

.272

.597

.120

.093

.151

.579

.079

.032

.319

.085

.504

.221

.114

.113

.350

.247

.103

.163

.362

.062

.192

.052

.308

.086

.056

.553

.040

.298

.075

.815

.043

.316

.079

.783

.051

.303

.048

.029

.819

.034

.009

.196

.169

.190

.602

.152

.176

.150

.606

.019

.187

.183

.200

.651

.105

.023

.264

.379

.292

.301

.270

.321

.319

.285

.058

.422

.429

.202

.055

.121

.238

.521

.022

.199

.301

.484

.078

.189

.303

.100

.024

.606

.101

.127

.028

.178

.059

.121

.153

.686

.037

.130

.190

.680

.016

.090

.003

.084

.857

.176

.092

Note. Loadings below .20 have been grayed out and those above .45 are in bold to ease reading of the table.

and understand the psychological dynamics of money beliefs. Analyses of sex differences in Table 3 confirm many previous findings namely that females are more concerned with compulsive buying behavior (retail therapy) and present buying while males are more risky and likely to show off their wealth. Nearly every study on money beliefs and behaviors shows systematic and occasionally large sex differences (Furnham, von Stumm, & Fenton-O’Creevy, 2014; von Stumm et al., 2013). However, perhaps the most interesting feature of the study lies in the regression because of the way it highlights certain factors that seem consistently related to money beliefs. As noted in Tables 5 and 6 three variables seemed unrelated to money beliefs and practices. The first was education which was not related to money beliefs and behaviors. This perhaps counter-intuitive finding has been found before and can surprise financial experts and advisors that often very well-educated people are surprisingly Journal of Individual Differences (2020), 41(1), 17–29

ignorant about their money and vice versa (Furnham, 2014). Whereas specifically financial education is related to money beliefs general education is not (Lay & Furnham, 2018). However, this was a relatively homogenous convenience sample and it may well be that if there were a wider range of educational attainment, some significant differences would become manifest. The second factor was self-rated success at work which seemed unrelated to the money beliefs. It has been observed by many in this area that “money madness” and irrationality seems not to be associated with more or less successful people, whose success may be in part due to education (see above) (Furnham, 2014) A third factor seeming unrelated to money beliefs was a report of a happy childhood. There is a vast literature inspired by psychoanalysis that suggests that money attitudes and beliefs are formed in childhood and often as a result of inappropriate parenting (Furnham, von Stumm, & Milner, 2014). Many writers have noted that those with Ó 2019 Hogrefe Publishing

A. Furnham & S. Grover, A New Money Behavior Quiz

Figure 1. Four factor CFA model.

2. I think I check my financial affairs more than other people 4. I spend a lot of time trying to find money bargains

Fitbit

6. I am constantly re-evaluating all my investments 13. Money can help you be accepted by others 15. Having a lot of money is a sign of success

Status

19. You get respect from others when you have lots of money

16. I rather enjoy letting people know how well-off I am 17. I use money to persuade people to do things for me

Splasher

18. I admit I buy things to impress others

25. Thinking about money makes me anxious 26. I dither a lot over money decisions

Anxious

29. I feel anxious and defensive about my personal finances

Table 5. Results for the hierarchical regression Fitbit financier

Anxious trader

Social value spender

The Cash Splasher

The Hoarder

The Ostrich

F = 3.83***, adj R2 = .05

F = 2.80**, adj R2 = .03

F = 8.32***, adj R2 = .12

F = 10.39***, adj R2 = .15

F = 8.26***, adj R2 = .12

F = 7.36***, adj R2 = .10

Gender

.13

2.61**

.07

1.34

.10

2.06*

.14

2.84**

.12

2.35*

.09

1.85

Education

.02

0.41

.06

1.08

.07

1.35

.01

0.20

.02

0.44

.03

0.64

Age

.11

2.10*

.04

0.70

.10

2.11*

.20

4.23***

.07

1.32

.18

3.60***

Wealth

.13

2.08*

.13

2.03*

.27

4.44***

.24

4.01***

.05

0.81

.10

1.74

Success at work

.02

0.30

.01

0.21

.03

0.50

.09

1.51

.05

0.89

.11

1.77

Happy childhood

.04

0.73

.03

0.60

.08

1.61

.08

1.65

.02

0.42

.09

1.81

Spender or saver

.15

2.92**

.08

1.61

.21

4.19***

.19

3.88***

.34

6.92***

.17

3.48***

Note. *p < .05, **p < .01, ***p < .001.

Table 6. Results for the regression onto the factor scores F1 – Fitbit

F2 – Status

F3 – Splasher

F4 – Anxious

F = 3.42**, adj R2 = .04

F = 1.22, adj R2 = .00

F = 15.92***, adj R2 = .21

F = 10.29***, adj R2 = .14

Gender

.11

2.11*

.08

1.53

.12

2.59*

.14

Education

.01

.19

.00

.06

.01

.21

.02

2.90**

Age

.15

2.87**

.04

.68

.25

5.37***

.18

3.75***

Wealth

.11

1.83

.08

1.19

.28

4.96***

.21

3.50***

Success at work

.05

.77

.09

1.48

.05

.94

.10

1.66

Happy childhood

.01

.19

.02

.41

.09

1.92

.07

1.51

Spender or saver

.11

2.07*

.08

1.62

.22

4.74***

.14

2.87**

.40

Note. *p < .05, **p < .01, ***p < .001.

Ó 2019 Hogrefe Publishing

Journal of Individual Differences (2020), 41(1), 17–29

self-reported unhappy childhoods blamed that experience on their poor money management. However, this study failed to confirm this suggestion. On the other hand, we did not get details about the childhood, particularly parents’ economic socialization which could be crucial. What the regression results (Tables 5 and 6) did confirm was the importance of four factors to all money beliefs. The first was sex which has been demonstrated many times before. The results suggested that females were more likely to be Fitbit Financiers, Social Value Spenders and Hoarders and males Cash Splashers. Similarly, females scored higher on the factors Fitbit and Anxious, while males scored higher on Splasher, and there were no effects of gender on Status. As noted, many times before males associate money more with Power and females with Love (Furnham, 2014). It seems that money is associated more with affect in females. The results also confirmed the associations between the money beliefs and age and wealth which are themselves positively associated (r = .18, p < .01). Younger people tended to be associated with being a FitBit Financier and a Cash Splasher but also a Social Value Spender and an Ostrich indicting perhaps the more technologically savvy but brash attitudes of young people. They were also more likely to associate money with influence and respect. Actual wealth was also associated with a number of beliefs. Thus, the wealthier people were more likely to be FitBit Financiers, Anxious Traders, Cash Splashers, and Social Values Spenders. This is clearly an interesting finding because they are such very different types, yet the participant’s wealth was associated with all of them, particularly the latter two. However, perhaps one of the most interesting findings to emerge from this study was the predictive power of the single question: “Are you a spender or saver?” Many studies of financially distressed couples noted at the heart of their financially difficulties was that one was a Spender and the other a Saver and that they could not resolve these two opposite instincts (Furnham, 2014). Indeed, this single item may be for financial advisors the single best place to try to start diagnosing the beliefs and behaviors of their clients. Of all the regressions shown in Tables 5 and 6, the regression with the Splasher resulting from the four-factor solution accounted for most of the variance (R2 = 21). Four demographics showed this to be the younger, males who rated themselves relatively wealthy and a spender. Equally interesting was the regression with the factor Status which was nonsignificant. Like all studies this had limitations. It had a relatively small convenience sample and all the data were self-report allowing no causal analysis. Replicating this on a bigger national sample would always allows for not only more confidence in the results but also the possibility of more sophisticated analysis like SEM. Further, both a strength and Journal of Individual Differences (2020), 41(1), 17–29

A. Furnham & S. Grover, A New Money Behavior Quiz

weakness of this study is to take ideas and observation from those familiar with financial advice rather than the psychology of money. While academic test developers might do well to seek the help and advice of those who attempt to understand, categorize, and label the types of individuals they deal with, it may also help the latter to consult the academic literature. The question is whether this scale is different from, or better than other measures in the area (Lay & Furnham, 2018). Ideally any new scale would have to show convergent, discriminant and predictive validity of the new measure which is a considerable research undertaking. One issue of interest is how financial advisors would react to the essential lack of empirical support for their typology. A few who took part in this study did not express great surprise and seemed happy to take into consideration the types that resulted from the factor analysis.

References Bailey, W., & Gustafson, W. (1986). Gender and gender-role orientation differences in attitudes and behavior towards money. Proceedings of the 4th Annual Conference of the Association of Financial Counselling and Planning Educators, 1, 11–20. Bailey, W., & Gustafson, W. (1991). An examination of the relationship between personality factors and attitudes to money. In R. Frantz, H. Singh, & J. Gerber (Eds.), Handbook of behavioural economics (pp. 271–285). Greenwich, CT: JAI Press. Bailey, W., Johnson, P., Adams, C., Lawson, R., Williams, P., & Lown, J. (1994). An exploratory study of money beliefs and behaviours scale using data from three nations. Consumer Interests Annual, 40, 178–185. Bailey, W., & Lown, J. (1993). A cross-cultural examination of the aetiology of attitudes toward money. Journal of Consumer Studies and Home Economics, 17, 391–402. https://doi.org/ 10.1111/j.1470-6431.1993.tb00181.x Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238–246. https://doi.org/ 10.1037//0033-2909.107.2.238 Bentler, P. M., & Bonnet, D. C. (1980). Significance tests and goodness of fit in the analysis of covariance structures. Psychological Bulletin, 88, 588–606. https://doi.org/10.1037/ 0033-2909.88.3.588 Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon’s mechanical Turk a new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science, 6, 3–5. https:// doi.org/10.1177/1745691610393980 Burgess, S. M. (2005). The importance and motivational content of money attitudes: South Africans with living standards similar to those in industrialised Western countries. South African Journal of Psychology, 35, 106–126. https://doi.org/10.1177/ 008124630503500107 Christopher, A. N., Marek, P., & Carroll, S. M. (2004). Materialism and attitudes toward money: An exploratory investigation. Individual Differences Research, 2, 109–117. Engelberg, E., & Sjoberg, L. (2006). Money attitudes and emotional intelligence. Journal of Applied Social Psychology, 36, 2027– 2047. https://doi.org/10.1111/j.0021-9029.2006.00092.x Furnham, A. (1984). Many sides of the coin: The psychology of money usage. Personality and Individual Differences, 5, 501– 509. https://doi.org/10.1016/0191-8869(84)90025-4

Ó 2019 Hogrefe Publishing

A. Furnham & S. Grover, A New Money Behavior Quiz

Furnham, A. (1996). Attitudinal correlates and demographic predictors of monetary beliefs and behaviours. Journal of Organizational Behaviour, 17, 375–388. https://doi.org/10.1002/(SICI) 1099-1379(199607)17:4<375::AID-JOB767>3.0.CO;2-8 Furnham, A. (2014). The new psychology of money. London, UK: Routledge. Furnham, A., & Murphy, T.-A. (2018). Money types, money beliefs and financial worries: An Australian study. Australian Journal of Psychology. Advance online publication. https://doi.org/ 10.1111/ajpy.12219 Furnham, A., von Stumm, S, & Fenton-O’Creevy, M. (2014). Sex differences in money pathology in the general population. Social Indicators Research, 123, 701–713. https://doi.org/ 10.1007/s11205-014-0756-x Furnham, A., von Stumm, S., & Milner, R. (2014). Moneygrams: Recalled childhood memories about money and adult money pathology. Journal of Financial Therapy, 5, 40–54. https://doi. org/10.4148/1944-9771.1059 Furnham, A., Wilson, E., & Telford, K. (2012). The meaning of money: The validation of a short money-types measure. Personality and Individual Differences, 52, 707–711. https:// doi.org/10.1016/j.paid.2011.12.020 Gresham, A., & Fontenot, G. (1989). The differing attitudes of the sexes toward money: An application of the money attitude scale. Advances in Marketing, 8, 380–384. Hanley, A., & Wilhelm, M. (1992). Compulsive buying: An exploration into self-esteem and money attitudes. Journal of Economic Psychology, 13, 5–18. https://doi.org/10.1016/01674870(92)90049-d Klontz, B. T., Britt, S. L., Archuleta, K., & Klontz, P. T. (2012). Disordered money behaviors: Development of the Klontz Money Behavior Inventory. Journal of Financial Therapy, 3, 17–42. https://doi.org/10.4148/jft.v3i1.1485 Klontz, B., Britt, S. L., Mentzer, J., & Klontz, T. (2011). Money beliefs and financial behaviors: Development of the Klontz Money Script Inventory. Journal of Financial Therapy, 2, 1. https://doi.org/10.4148/jft.v2i1.451 Klontz, B. T., Seay, M. C., Sullivan, P., & Canale, A. (2014). The psychology of wealth: Psychological factors associated with high income. Journal of Financial Planning, 27, 46–53. Lay, A., & Furnham, A. (2018). A new money attitudes questionnaire. European Journal of Psychological Assessment. Advance online publication. https://doi.org/10.1027/1015-5759/ a000474 Lim, V. K., & Teo, T. S. (1997). Sex, money and financial hardship: An empirical study of attitudes towards money among undergraduates in Singapore. Journal of Economic Psychology, 18, 369–386. https://doi.org/10.1016/s0167-4870(97)00013-5 Lynn, R. (1991). The secret of the miracle economy: Different national attitudes to competitiveness and money. London, UK: Social Affairs Unit. Medina, J. F., Saegert, J., & Gresham, A. (1996). Comparison of Mexican-American and Anglo-American attitudes toward money. Journal of Consumer Affairs, 30, 124–145. https://doi. org/10.1111/j.1745-6606.1996.tb00728.x Ozgen, O., & Bayoglu, A. (2005). Turkish college students’ attitudes towards money. International Journal of Consumer Studies, 29, 493–501. https://doi.org/10.1111/j.1470-6431.2005.00417.x Paolacci, G., Chandler, J., & Ipeirotis, P. G. (2010). Running experiments on Amazon Mechanical Turk. Judgment and Decision Making, 5, 411–419. Retrieved from https://ssrn.com/abstract=1626226 Revelle, W. R. (2017). psych: Procedures for personality and psychological research. (R package version 1.7.8) [Computer software]. Retrieved from https://CRAN.R-project.org/ package=psych

Ó 2019 Hogrefe Publishing

Roberts, J., & Sepulveda, C. (1999). Money attitudes and compulsive buying. Journal of International Consumer Marketing, 11, 53–79. https://doi.org/10.1300/j046v11n04_04 Rose, G. M., & Orr, L. M. (2007). Measuring and exploring symbolic money meanings. Psychology and Marketing, 24, 743–761. https://doi.org/10.1002/mar.20182 Rovnick, N. (2017, January 12). Six financial personality types – which one are you? Financial Times. Retrieved from https:// www.ft.com/content/5e8da24c-bb09-11e6-8b45b8b81dd5d080 Tang, T. L. P. (1992). The meaning of money revisited. Journal of Organizational Behavior, 13, 197–202. https://doi.org/10.1002/ job.4030130209 Tang, T. L. P. (1995). The development of a short money ethic scale: Attitudes toward money and pay satisfaction revisited. Personality and Individual Differences, 19, 809–816. https://doi. org/10.1016/S0191-8869(95)00133-6 Tatarko, A., & Schmidt, P. (2012). Social capital and attitudes towards money. Higher School of Economics Research Paper, No. WP BRP, 7. https://doi.org/10.2139/ssrn.2011044 Taylor, C., Klontz, B., & Britt, S. (2016). Internal consistency and convergent validity of the Klontz Money Behavior Inventory (KMBI). Journal of Financial Therapy, 6, 14–31. https://doi.org/ 10.4148/1944-9771.1101 Taylor, C., Klontz, B.T., & Lawson, D. (2017). Money disorders and locus of control: Implications for assessment and intervention. Journal of Financial Therapy, 8, 124–137. https://doi.org/ 10.4148/1944-9771.1121 von Stumm, S., Fenton-O’Creevy, M., & Furnham, A. (2013). Financial capability, money attitudes and socioeconomic variables. Personality and Individual Differences, 54, 344–349. https://doi.org/10.1016/j.paid.2012.09.019 Wernimont, P., & Fitzpatrick, S. (1972). The meaning of money. Journal of Applied Psychology, 56, 248–261. https://doi.org/ 10.1037/h0033107 Wilhelm, M. S., Varcoe, K., & Fridrich, A. H. (1993). Financial satisfaction and assessment of financial progress: Importance of money attitudes. Financial Counselling and Planning, 4, 181– 199. Yamauchi, K. T., & Templer, D. J. (1982). The development of a money attitude scale. Journal of Personality Assessment, 46, 522–528. https://doi.org/10.1207/s15327752jpa4605_14 History Received March 16, 2018 Revision received March 25, 2019 Accepted March 25, 2019 Published online June 17, 2019 Conflict of Interest There is no conflict of interest in the paper. ORCID Adrian Furnham http://orcid.org/0000-0001-7545-8532

Adrian Furnham Department of Leadership and Organisational Behaviour Norwegian Business School (BI) Nydalsveien 37 0484 Oslo Norway adrian@adrianfurnham.com

Journal of Individual Differences (2020), 41(1), 17–29

I think I check my financial affairs more than other people

There are lots of money bargains if you are prepared to search for them I spend a lot of time trying to find money bargains

I follow the trends about money management

I am constantly re-evaluating all my investments

You have to be vigilant about all money matters

I believe investing time in watching money programmes is worth it When it comes to spending money on myself its “because I am worth it” I often demonstrate my love to people by buying them presents I am very generous with the people I love

Journal of Individual Differences (2020), 41(1), 17–29

Money can help you be accepted by others

I love retail therapy: shopping to cheer me up

Having a lot of money is a sign of success

I rather enjoy letting people know how well-off I am

I use money to persuade people to do things for me

I admit that I buy things to impress others

You get respect from others when you have lots of money

I feel safe and secure if I have a lot of money saved

I prefer to be safe rather than a gambler When it comes to money I value having a lot of easy-to-access money in the bank

I am much more of a saver than a spender

Essentially I am risk-averse when it comes to money investments Thinking about money makes me anxious.

I dither a lot over money decisions.

I am really not interested in money matters.

I prefer to let others I trust make my important money decisions I feel anxious and defensive about my personal finances

The best present you can give to someone is money

I get a real kick out of the business of managing my money.

Table A1. Inter-item correlations

Appendix

.101*

.003

.025

.003

.170**

0.079

0.014

0.093

0.027

.236** .272**

.086

.188** .003

.113*

.328** .184**

.328** .391**

.088

.291** .273**

.243** .184**

.250** .132**

.268** .127*

.392** .184**

.178** .198**

.093

.171** .136**

.258** .157**

.084

.181** .205**

.357** .254**

.383** .189**

.130** .318**

.518** .399**

.532** .320**

.198** .237**

.181** .243**

.430**

.278**

.320** .242** .019

.130** .102*

.127*

.058

.094

.146** .223** .153** .086

.169**

.075 .087

.201** .214** .228** .182**

.271** .224** .134** .229**

.276** .237** .158** .261**

.351** .302** .145** .379**

.075

.222** .020

.118*

.009

.044 .142** .068

.089

.206** .177** .204** .128** .166**

.149** .071

.019

.112*

0.089

.181** 0.075 .204** .203** 0.035

0.068

.202** 0.048

.167** .147**

.175** .142** .157** .238** .298** .194**

.126*

.002

.176** .267** .204** .166** .153** .216** .063

.071

.344** .210** .281** .279** .339** .279**

.250** .147** .026

.279** .184** .194** .204** .326** .210**

.174** .108*

.071

.150** .012

.004

.140** .165** .188** .187** .182** .242**

.077

.153**

.261**

.183** .028

.133** .328** .177** .017

.168** 0.073

.045

.121*

.149** .195** .252** .085

.098*

.273** .163** .045

.102*

.139** .197** .526** .366** .084

.411** .287** .131** .208**

.210** .272** .457**

.123*

.391**

0.011

.171**

.052

.049

.037

.064

.055

.265**

.028

.165**

.161**

.329**

.328**

.399**

.237**

.301**

.182**

.212**

.044

.271**

.115*

.201**

.375** .256** .341**

.304** .476** .199**

.055

.214**

.035

.071

.006

.028

.054

.110*

.107*

.104*

.019

.056

0.057

.026

.029

–.250** .153** -.050

.055

.120*

.094 .143**

.085

(Continued on next page)

.172** .206** .131**

.220** .100*

.014

.133** .082

.137** .194** .129**

.008

.080

.136** .316** .114*

.106*

.141** .248** .132**

.277** .581** .177**

.271** .197** .452**

.175** .404** .184** .238**

.077

.075

.089

.107*

.063

.160** 0.021

.003

.168**

.126*

.055

0.03

.227**

.032

.140**

.252**

.338**

.245**

.297**

.184**

.309**

.169**

.226**

.311**

28 A. Furnham & S. Grover, A New Money Behavior Quiz

Ó 2019 Hogrefe Publishing

I think I check my financial affairs more than other people

There are lots of money bargains if you are prepared to search for them I spend a lot of time trying to find money bargains

I follow the trends about money management

I am constantly re-evaluating all my investments

You have to be vigilant about all money matters

Ó 2019 Hogrefe Publishing

I dither a lot over money decisions.

I am really not interested in money matters.

I prefer to let others I trust make my important money decisions I feel anxious and defensive about my personal finances

Note. *p < .05, **p < .01, ***p < .001.

Essentially I am risk-averse when it comes to money investments Thinking about money makes me anxious.

You get respect from others when you have lots of money

I am much more of a saver than a spender

I admit that I buy things to impress others

I use money to persuade people to do things for me

I rather enjoy letting people know how well-off I am

I prefer to be safe rather than a gambler When it comes to money I value having a lot of easy-to-access money in the bank

Having a lot of money is a sign of success

I love retail therapy: shopping to cheer me up

I feel safe and secure if I have a lot of money saved

Money can help you be accepted by others

The best present you can give to someone is money

I get a real kick out of the business of managing my money.

Table A1. (Continued)

.283**

.214**

.099*

.224**

.105*

.162**

.106*

.010

.403**

.016

.408**

.541**

.265**

.168**

.111*

.399**

.104*

.220**

.077

.065

.024

.118*

.215**

.075

.212**

.621**

.593**

.151**

.365**

.181**

.186**

.091

.053

.039

.226**

.007

.183**

.502**

.322**

.179**

.139**

.135**

.179**

.009

.402**

.028

.133**

.246**

.385** 0.071

.088

.183**

.060

.065

.084

.124*

.205**

.035

.196**

.291**

.165**

.150**

.139**

.155**

.460**

.205** .100*

.011

.071

.109*

.386**

.174**

.107*

.199**

0.063

.335**

.162**

0.024

.098*

.156** 0.022

0.057

.145**

.223** 0.035

0.058

.229** 0.056

0.026

.185**

.105*

0.077

.120*

.261**

.620**

.208**

.110*

.447** .326**

.421** 0.029

.266**

0.096

.154**

A. Furnham & S. Grover, A New Money Behavior Quiz 29

Journal of Individual Differences (2020), 41(1), 17–29

Original Article

Reliability of MTurk Data From Masters and Workers Steven V. Rouse Social Sciences Division, Pepperdine University, Malibu, CA, USA

Abstract: Previous research has supported the use of Amazon’s Mechanical Turk (MTurk) for online data collection in individual differences research. Although MTurk Masters have reached an elite status because of strong approval ratings on previous tasks (and therefore gain higher payment for their work) no research has empirically examined whether researchers actually obtain higher quality data when they require that their MTurk Workers have Master status. In two different online survey studies (one using a personality test and one using a cognitive abilities test), the psychometric reliability of MTurk data was compared between a sample that required a Master qualification type and a sample that placed no status-level qualification requirement. In both studies, the Master samples failed to outperform the standard samples. Keywords: Amazon’s Mechanical Turk, MTurk, online data collection, psychometrics, score reliability

Amazon’s Mechanical Turk (MTurk; Amazon, 2011) has quickly become a valuable resource for psychological researchers (Buhrmester, Talaifar, & Gosling, 2018). As an Internet-based platform that allows “Requestors” to hire “Workers” for brief Human Intelligence Tasks (HITs), researchers have often used MTurk as a means of gaining survey data for online research on personality and individual difference variables. A PsycINFO search conducted in March 2019 using “MTurk” or “Mechanical Turk” as subject terms resulted in the identification of 1 publication in 2010, 6 in 2011, 17 in 2012, 34 in 2013, 59 in 2014, 125 in 2015, 192 in 2016, 258 in 2017, and 333 in 2018, demonstrating the speed with which empirical psychological research has incorporated this data acquisition resource. Goodman and Paolacci (2017) highlighted several strengths of MTurk data for research purposes that may be reasons for this rapid growth. First, MTurk data are less expensive than data gained through many other methods, allowing for more democratization in the research process (as researchers who do not have large budgets are able to be involved in high-quality research) and for more exploratory and more confirmatory research than can be conducted using more expensive means. Second, MTurk samples (representing a wide range of ages, socioeconomic backgrounds, racial and ethnic groups, and countries of origin) tend to be more diverse than samples obtained using many traditional methods, allowing researchers to focus specifically on subgroups that might be hard to study using traditional data collection methods. Third, MTurk provides a high level of flexibility, allowing for cross-cultural research, Journal of Individual Differences (2020), 41(1), 30–36 https://doi.org/10.1027/1614-0001/a000300

longitudinal research, and even research in which Workers engage in online real-time interactions with others. Fourth, we can presume high levels of quality for MTurk data because of their built-in incentive structure; since Workers who submit poor quality work can be rejected or given low approval ratings (which can then prevent a Worker from being eligible for additional HITs) and Workers are incentivized to be attentive to detail and conscientious in their work. Balanced against the benefits of using MTurk for data collection purposes, Buhrmester et al. (2018) noted several factors that could have a negative effect on the quality of the data. For example, plausible concerns might be raised about inattentiveness while participating in research, dishonesty in affirming qualifications to participate in studies for which one is not actually qualified, and the non-naivete and familiarity with research measures that results when a person completes a large number of research studies. Although Goodman and Paolacci (2017) presented the incentive structure as a reason for confidence in the quality of MTurk data, one might reasonably wonder if the opposite were true. It is plausible that because Workers are aware that a Requester might reject their work or provide low approval ratings, there may be an implicit pressure to provide socially desirable responses, and this impression management could affect the validity of a measure. In addition, if people are paid for task completion, this might increase the likelihood of rushing through tasks, resulting in inattentiveness. As a general rule, this does not appear to be the case. For example, Ramsey, Thompson, McKenzie, Ó 2019 Hogrefe Publishing

S. V. Rouse, Reliability of MTurk Data from Masters and Workers

and Rosenbaum (2016) showed a greater level of attentiveness to research instructions for an MTurk sample than for an undergraduate sample. Likewise, Thomas and Clifford (2017) found that inattentiveness in an experimental procedure was no more problematic for MTurk samples than for other convenience samples, and MTurk samples were found to be engaged in experimental procedures and to be committed to producing high-quality data. Thus, as a general rule, the external incentivization of MTurk payment does not appear to lead to poor or invalid data. Research on the quality of MTurk data has especially focused on the question of score reliability, or the consistency of the data. Coefficients exceeding .70 have been documented for MTurk data for temporal consistency (Holden, Dennie, & Hicks, 2013; Kim & Hodgins, 2017), inter-rater consistency (Tosti-Kharas & Conley, 2016), and internal consistency (Bates & Lanza, 2013; Buhrmester, Kwang, & Gosling, 2011; Miller, Crowe, Weiss, Lynam, & Maples-Keller, 2017). For most of these MTurk reliability studies, the documentation of reliability took the form of contrasting reliability estimates obtained for MTurk samples against those obtained for standard samples. For example, Johnson and Borden (2012) calculated a Cronbach’s α of .86 for a measure of empathy taken in a faceto-face laboratory setting and a Cronbach’s α of .83 when this same test was taken by an MTurk sample. Rouse (2015) used a hypothesis testing procedure developed by Bonett (2003) to evaluate the statistical significance of the differences in reliability estimates for MTurk data and traditional data. Rouse showed that the inclusion of “best practices” (such as asking respondents to Opt-in or Opt-out of the study) resulted in reliability estimates for MTurk samples that were not statistically different from those calculated for community samples. Although a large number of articles have documented the psychometric quality of MTurk responses, very little research has been conducted to evaluate the effect of a “best practice” advocated by Amazon (2011) – hiring MTurk Masters rather than non-Master Workers. “Master” refers to a designation given to an MTurk Worker who has demonstrated superior performance on a large number of HITs for several Requesters. Requesters can choose to limit their HITs to only being completed by MTurk Masters; presumably, this would result in high-quality data because “Masters must maintain this high level of performance or risk losing this distinction” (Amazon, 2011, p. 8). However, there is an additional financial cost; Amazon charges a 5% fee for HITs that are designated exclusively for Masters (above and beyond the 40% fee charged for all HITs). “Because Masters have demonstrated accuracy, they can command a higher reward for their HITs” (p. 8). Lovett, Bajaba, Lovett, and Simmering (2017) offered a supportive

rationale based on Equity Theory; since Masters receive higher payment than non-Master Workers, one may presume that they will put forth more effort and more attention into the research task. However, no studies have empirically examined whether or not Masters provide more accurate or reliable survey data. A literature search only identified one publication that addressed the quality of data from MTurk Masters. Lovett et al. (2017) conducted a survey of 40 MTurk Masters, asking both open-ended and forced-choice questions about MTurk compensation, their motivation while completing MTurk HITs, the settings in which they complete MTurk HITs, and their perceptions of the quality of their responses. In general, these MTurk Masters believed that their work was high-quality, but that it could be affected by factors such as the fairness of compensation and attention-checks that encourage careful responding. While the responses given by these MTurk Masters are aligned with Amazon’s (2011) expectation that they provide especially accurate data, Lovett et al. only sought self-report perceptions of their own work and did not attempt to empirically contrast the quality of work produced by Masters and non-Master Workers. The present studies sought to begin exploring this deficit in the research literature, to determine whether Masters generate more reliable data and, if so, whether the increased quality justifies a higher research expense. For the present studies, the quality of the data was assessed based on internal consistency reliability; this decision is aligned with Viswanathan’s (2005) taxonomy of measurement error and the methods best suited for identifying such error. For example, Viswanathan noted that internal consistency estimates are well-suited for identifying generic random error that may be present within the administration of a measure. Therefore, if one subject sample is more attentive to a survey than is another sample (and therefore obtains scores that have less random measurement error), this difference in attentiveness and data quality will likely be seen in the internal consistency estimates of reliability. Thus, Amazon’s claim of superior data produced by Masters could be evaluated by statistically comparing the reliability estimates obtained for the same measures when restricting data collection to Masters and when foregoing such a restriction. In the present studies (one using a measure of personality and one using a measure of cognitive ability), independent datasets were collected from MTurk Masters and non-Master Workers, with internal consistency reliability estimates calculated for each sample. The hypothesis was that Masters would generate more reliable data than non-Master Workers; support for this hypothesis would justify a higher payment for these elite surveyrespondents.

Ó 2019 Hogrefe Publishing

Journal of Individual Differences (2020), 41(1), 30–36

S. V. Rouse, Reliability of MTurk Data from Masters and Workers

Reliability for MTurk Masters and MTurk Workers on a Personality Test

“Altruism (helping others in need)” and “Compassion (caring for others, displaying kindness)” for Communion. Respondents were asked to indicate the salience of each value to their lives on a 9-point scale from 1 (= Not at all important to me) to 9 (= Extremely important to me). Trapnell and Paulhus reported internal consistency reliability estimates of .83 for both scales. This measure was selected for this present study because it is relatively brief, and yet it provides scale scores for two distinctly different personality traits. It was also a viable scale because of consistently high reliability estimates across a wide range of samples. Finally, it was selected because a literature search conducted in January 2017 suggested that none of the publications citing Trapnell and Paulhus (2012) included the terms “MTurk” or “Mechanical Turk” in their abstracts. This leads to the assumption that MTurk samples would not have been exposed often to these questions (unlike some measures that are frequently used in MTurk studies), making it unlikely that the respondents would be influenced by familiarity with the measure.

The purpose of the first study was to determine whether data obtained from MTurk Masters were more reliable than data obtained from non-Master Workers on a personality measure. The data collection and data analysis plans were preregistered at https://osf.io/j3x4k. The analysis strategy was to collect two independent datasets and calculate Cronbach’s alphas for each. Using Bonett’s (2003) process for statistically evaluating the difference between reliability estimates, a one-tailed hypothesis was set with a .05 p-value: The Null Hypothesis was that the Masters’ data were not more reliable than the non-Masters’ data, and the Alternate Hypothesis was that the Masters’ data were more reliable. The study was approved by the Seaver College (Pepperdine University, Malibu, CA, USA) Institutional Review Board prior to data collection.

Method Materials Two nearly identical surveys were created, with informed consent information and the questions themselves administered online in a single window within the MTurk platform. To follow the principle of Open Materials advocated by the Open Science Collaboration (2015), the survey has been publicly archived at https://osf.io/e2387/. Pilot testing of the survey suggested that it would be reasonable to expect that this 30-item survey could be completed in 10 min. Based on this, a payment of $1.50 was selected (for an effective wage of $9.00/hr, which exceeded the $8.25/hr median US state minimum wage at the time of data collection). Demographic Information Both versions of the survey began with questions to assess age, gender, and race/ethnicity. The version of the survey that was not restricted to Masters also included a question to assess whether or not the respondent held Master status. Agentic and Communal Values Scale (ACV; Trapnell & Paulhus, 2012) With 24-point Likert scale, the ACV provides independent measures of two main dimensions of personality and motivation – Agency (i.e., “getting things done” and “getting ahead”) and Communion (i.e., “getting along”). Respondents are asked about the importance of 24 values as guiding principles in their lives, with 12 items such as “Achievement (reaching lofty goals)” and “Status (high rank, wide respect)” for Agency and 12 items such as

Journal of Individual Differences (2020), 41(1), 30–36

Opt-In/Opt-Out Question The final question on the survey followed the recommendation of Rouse (2015), asking respondents to indicate whether or not their data should be included in the analyses. Subjects were assured that their response was confidential, it would not affect the MTurk approval rating they would receive, and it would not affect their payment. They were asked to either Opt-In (i.e., “You should keep my data; I paid attention and answered honestly”) or Opt-Out (i.e., “You should delete my data; honestly, I wasn’t really taking this seriously.”) from inclusion in the data analyses. Master Sample To be eligible for inclusion in the first sample, MTurk Workers had to be in the United States and have earned Master status. With these criteria set, responses were collected from 80 respondents; this sample size was selected to meet to Bonett’s (2003) recommendation for testing the statistical significance of the difference between two reliability estimates (specifying a one-tailed significance of .05, power of .80, and an anticipated range of reliability estimates from .70 to .85). All of the respondents answered the final Opt-In/Opt-Out question by indicating that their responses should be included for data analysis. Ages ranged from 22 to 63 years (M = 36.88, SD = 9.18). The sample included 50 men and 30 women, and race/ethnicity selfidentifications included European American (n = 67), Latinx/Hispanic (n = 9), African American (n = 6), and Asian American (n = 4); the total number of race/ethnicity self-identifications exceeded 100% because respondents were allowed to select multiple self-identifications.

Ó 2019 Hogrefe Publishing

S. V. Rouse, Reliability of MTurk Data from Masters and Workers

Worker Sample To be eligible for the second sample, MTurk workers had to be in the United States and not be included in the first sample. All respondents indicated in the Opt-In/Opt-Out question that their data should be included in analyses. Of the 80 respondents, 51 were men and 29 were women, and the race/ethnicity self-identifications included European American (n = 56), African American (n = 11), Asian American (n = 8), Latinx/Hispanic (n = 7), Native American (n = 1), and Other (n = 1). Their ages ranged from 18 to 74 (M = 34.11, SD = 11.80). The majority (n = 71) did not have MTurk Master status. A small proportion of the sample (n = 9) did have Master status; however, excluding these participants from the sample would result in data that do not reflect typical MTurk research. Although many MTurk studies do not restrict inclusion to MTurk Masters, a literature search conducted in January 2017 could not identify any MTurk studies that specifically excluded Masters, and the MTurk system itself only permits excluding non-Masters. By comparing responses from a Masters-only sample with a sample that did not specify status, this study was able to examine the effect of following a process that Amazon considers a “best practice” (i.e., Masters-only data collection).

MTurk Masters produce data with higher levels of reliability on a well-established personality measure.

Results and Discussion To follow the principle of Open Data advocated by the Open Science Collaboration (2015), the survey data have been publicly archived at https://osf.io/e2387/. The two samples did not differ on mean scores for either of the two scales. On the overall score for Agency, the Masters (M = 54.81, SD = 16.88) and the non-Master Workers (M = 59.08, SD = 16.23) were not significantly different (t = 1.61, p = .11, d = 0.26). Similarly, on the overall score for Communion, the Masters (M = 76.23, SD = 17.60) and the non-Master Workers (M = 77.34, SD = 17.39) were not significantly different (t = 0.90, p = .69, d = 0.06). The correlations between Agency and Communion scales for the Masters (r = .01, p = .92, 95% CI [ .23, .21]) and the non-Master Workers (r = .07, p = .54, 95% CI [ .29, .15]) suggests that these two scores are largely uncorrelated, which is consistent with the two-factor structure proposed by Trapnell and Paulhus (2012). The reliability estimates did not differ significantly across samples for either scale. On the Agency scale, Cronbach’s α for the Masters was .89, and Cronbach’s α for the nonMaster Workers was .87. The difference (z = 0.68, onetailed p = .25) did not meet criteria to reject the Null Hypothesis. On the Communion scale, Cronbach’s α for the Masters was .91, and Cronbach’s α for the non-Master Workers was .90. The difference (z = 0.67, one-tailed p = .25) did not meet criteria to reject the Null Hypothesis. Thus, the results of the first study did not suggest that Ó 2019 Hogrefe Publishing

Reliability for MTurk Masters and MTurk Workers on a Cognitive Ability Test The purpose of the second study was to determine whether data obtained from MTurk Masters on a measure of cognitive ability were more reliable than data obtained from MTurk non-Master Workers. The data collection and data analysis plans were preregistered at https://osf.io/j3x4k. The analysis strategy was to collect two independent data sets and calculate a Cronbach’s alpha for each. Using Bonett’s (2003) process for statistically evaluating the difference between reliability estimates, a one-tailed hypothesis test was set with a .05 p-value: The Null Hypothesis was that the Masters’ data were not more reliable than the nonMaster Workers’ data, and the Alternate Hypothesis was that the Masters’ data were more reliable. The study was approved by the Seaver College (Pepperdine University, Malibu, CA, USA) Institutional Review Board prior to data collection.

Method Materials Two nearly identical surveys were created to be administered online with informed consent and survey questions on a single window within the MTurk platform. Because the usage agreement for the cognitive ability test used in this study prohibits researchers from presenting items in a publicly accessible location, this study does not follow the Open Materials principle advocated by the Open Science Collaboration (2015). Pilot testing of the survey suggested that it would be reasonable to expect that this 22-item survey could be completed in 13 min. Based on this, a payment of $2.00 was selected (for an effective wage of $9.23/hr, which exceeded the $8.25/hr median US state minimum wage at the time of data collection). Demographic Information Both versions of the survey began with questions to assess age, gender, and race/ethnicity. The version of the survey that was not restricted to Masters also included a question to assess whether or not the respondent held Master status. International Cognitive Ability Resource (ICAR) – Verbal Reasoning (Condon & Revelle, 2014) The ICAR is a set of public domain cognitive ability tests, assessing a range of abilities similar to those included on Journal of Individual Differences (2020), 41(1), 30–36

S. V. Rouse, Reliability of MTurk Data from Masters and Workers

many tests of intelligence. The Verbal Reasoning scale includes 16 multiple-choice items that assess knowledge of word meanings, logical relationships, and general factual knowledge. Condon and Revelle reported an internal consistency reliability estimate of .76 for data collected during scale development. This scale was selected for the present study because (despite its brevity) it is a very demanding measure of cognitive ability. Because all items are textbased (unlike some of the visual items included in other ICAR scales), it lends itself well to be used for an online survey. Finally, it was selected because a literature review did not yield any instances in which MTurk data were collected using this measure, suggesting that it was unlikely for MTurk Workers to have a high level of familiarity with the measure (which could lead to inflation of scores through the practice effect).

The reliability estimate obtained for the Masters (Cronbach’s α = .67) was much lower than the reliability estimate obtained for the non-Master Workers (Cronbach’s α = .81). Because a one-tailed significance test was performed to test whether Masters’ data are more reliable, the difference (z = 2.33, one-tailed p = .99) did not meet the criteria to reject the Null Hypothesis. Thus, the results of the second study did not suggest that MTurk Masters produce data with higher levels of reliability on a well-established cognitive ability measure. If a two-tailed significance test had been performed, the results would have been strong enough to reject the Null Hypothesis. However, a one-tailed test was used in order to evaluate the claim by Amazon (2011) that hiring MTurk Masters is a “best practice” that enhances the likelihood of higher quality data.

Opt-In/Opt-Out Question The surveys ended with the same Opt-In/Opt-Out question used for Study 1.

General Discussion

Master Sample The 80 respondents in the Masters sample met the qualification criteria (i.e., being in the United States and having been grated Master status); all indicated in the Opt-In/ Opt-Out question that their data should be included in analyses. The 44 men and 36 women had ages ranging from 23 to 62 years (M = 36.78, SD = 8.70), and self-identified as European American (n = 63), Latinx/Hispanic (n = 9), African American (n = 7), and Asian American (n = 4). Worker Sample The 80 respondents in the Worker sample were in the United States and none had participated in the previous survey. All respondents indicated in the Opt-In/Opt-Out question that their data should be included in analyses. The 50 men and 30 women had ages ranging from 18 to 63 (M = 32.78, SD = 9.77), though one miscoded his age and two opted not to provide a response. Respondents selfidentified as European American (n = 68), Asian American (n = 6), African American (n = 5), and Latinx/Hispanic (n = 1), though one participant opted not to self-identify. The majority (n = 73) did not have Master status.

Results and Discussion To follow the principle of Open Data advocated by the Open Science Collaboration (2015), the survey data have been publicly archived at https://osf.io/e2387/. The mean scores on the ICAR Verbal Reasoning scale did not differ significantly (t = 1.75, p = .08, d = 0.28) between the Masters (M = 10.50, SD = 3.46) and nonMaster Workers (M = 9.65, SD = 2.61). Journal of Individual Differences (2020), 41(1), 30–36

Previous research has been supportive of the use of MTurk for research data collection. Although there are some differences that can be expected between MTurk samples and general community samples, a wide range of psychological phenomena have been replicated using MTurk samples, and many studies have documented that the psychometric quality of the data is as strong as would be expected for more traditional means of data collection. Amazon (2011) recommended hiring MTurk Masters (with an additional incremental fee) if one seeks higher quality data. However, no research had empirically examined this claim prior to the present project. In this project, two different studies failed to support this claim, with no evidence of higher reliability for MTurk Masters relative to data sets that did not require Master status. In fact, contrary to Amazon’s claim, the reliability of data provided by Masters was substantially lower than data provided by a general MTurk sample for one of the two studies. Additional research should be conducted to determine whether this finding is replicable. If researchers gather additional replication data to explore the reliability differences between Masters’ data and nonMaster Workers’ data, it is very possible that the results will continue to be contradictory. After all, in the present project one study yielded no clear difference in the psychometric quality between the two samples, while the other suggested that (if anything) the Masters’ data were psychometrically weaker. If this pattern of contradictory results continues, the results should be evaluated within the context of research by Hamby and Taylor (2016) on the effect of “survey satisficing” and “survey optimizing” on the quality of MTurk data. Optimizing refers to instances of careful and effortful consideration of questions before providing “the best” responses, whereas Satisficing refers to situations Ó 2019 Hogrefe Publishing

S. V. Rouse, Reliability of MTurk Data from Masters and Workers

in which a respondent provides quick, “good enough” responses. Hamby and Taylor demonstrated that these response patterns may affect the quality of MTurk data, such that Optimizing may be more common among MTurk respondents when motivation is high and task difficulty is low, but that Satisficing may be more common among MTurk respondents as task difficulty rises or motivation falls. In the present project, reliability levels were very similar on the personality measure (which is a relatively effortless survey to complete), but the reliability levels differed for the cognitive ability measure (which poses a much greater cognitive demand on the respondent). Perhaps a complex interaction effect exists between MTurk Master/ Non-master status and Optimizing/Satisficing. If MTurk Workers seek a Master status, it is plausible that they may put an emphasis on producing high quality work regardless of the cognitive demand of the HIT; it is plausible, however, that MTurk Masters might simply be satisfied with producing “good enough” work when the cognitive demand of the HIT gets high. In addition, as noted by Lovett et al. (2017), MTurk Masters acknowledge high levels of familiarity and expertise with self-report surveys and acknowledge working quickly on research measures in order to maximize compensation. These same Masters reported the belief that the quality of their data was strong, and acknowledged being observant of possible “attention checks” that might affect their compensation and approval ratings. Nevertheless, it is possible that Masters (many of whom are working as a primary source of income) become more inattentive than they realize when completing complex tasks like cognitive abilities tests. If this is the case, on some tasks Masters might actually provide lower quality data relative to non-Master Workers due to test-wiseness and the desire to finish work quickly. However, substantial additional research will be needed to determine whether or not this pattern replicates. Although the present studies did not find higher reliability estimates for data obtained by Masters, undermining Amazon’s (2011) claim that the employment of Masters is a “best practice,” additional psychometric statistics could provide information about other aspects of data quality. For example, factor analytic procedures can assess the consistency of the structure of a measure and tests of validity can assess the existence of systematic cross-construct error (Viswanathan, 2005). Central to the question of the quality of data obtained by Masters, additional research could examine the number of missing responses, or the length and complexity of answers to free-response open-ended survey items; within the MTurk environment, both of these could be affected by a Worker’s concern for having a HIT approved and receiving a positive evaluation. However, while the present studies did not refute Amazon’s claim,

the burden of responsibility lies with Amazon if the company wishes to promote hiring Masters as a “best practice”.

Ó 2019 Hogrefe Publishing

References Amazon. (2011). Requester best practices guide. Retrieved from http://mturkpublic.s3.amazonaws.com/docs/MTURK_BP.pdf Bates, J. A., & Lanza, B. A. (2013). Conducting psychology student research via the Mechanical Turk crowdsourcing service. North American Journal of Psychology, 15, 385–394. Bonett, D. G. (2003). Sample size requirements for comparing two alpha coefficients. Applied Psychological Measurement, 27, 72– 74. https://doi.org/10.1177/0146621602239477 Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon’s Mechanical Turk: A new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science, 6, 3–5. https:// doi.org/10.1177/1745691610393980 Buhrmester, M., Talaifar, S., & Gosling, S. D. (2018). An evaluation of Amazon’s Mechanical Turk, its rapid rise, and its effective use. Perspectives on Psychological Science, 13, 149–154. https://doi.org/10.1177/1745691617706516 Condon, D. M., & Revelle, W. (2014). The international cognitive ability resource: Development and initial validation of a publicdomain measure. Intelligence, 43, 52–64. https://doi.org/ 10.1016/j.intell.2014.01.004 Goodman, J. K., & Paolacci, G. (2017). Crowdsourcing consumer research. Journal of Consumer Research, 44, 196–210. https:// doi.org/10.1093/jcr/ucx047 Hamby, T., & Taylor, W. (2016). Survey satisficing inflates reliability and validity measures: A experimental comparison of college and Amazon Mechanical Turk samples. Educational and Psychological Measurement, 76, 912–932. https://doi.org/10.1177/ 0013164415627349 Holden, C. J., Dennie, T., & Hicks, A. D. (2013). Assessing the reliability of the M5-120 on Amazon’s Mechanical Turk. Computers in Human Behavior, 29, 1749–1754. https://doi.org/ 10.1016/j.chb.2013.02.020 Johnson, D. R., & Borden, L. A. (2012). Participants at your fingertips: Using Amazon’s Mechanical Turk to increase student-faculty collaborative research. Teaching of Psychology, 39, 245–251. https://doi.org/10.1177/0098628312456615 Kim, H. S., & Hodgins, D. C. (2017). Reliability and validity of data obtained from alcohol, cannabis, and gambling populations on Amazon’s Mechanical Turk. Psychology of Addictive Behaviors, 31, 85–94. https://doi.org/10.1037/adb0000219 Lovett, M., Bajaba, S, Lovett, M., & Simmering, M. J. (2017). Data quality from crowdsourced surveys: A mixed method inquiry into perceptions of Amazon’s Mechanical Turk Masters. Applied Psychology: An International Review, 66, 1–28. https://doi.org/ 10.1111/apps.12124 Miller, J. D., Crowe, M., Weiss, B., Lynam, D. R., & Maples-Keller, J. L. (2017). Using online, crowdsourcing platforms for data collection in personality disorder research: The example of Amazon’s Mechanical Turk. Personality Disorders: Theory, Research, and Treatment, 8, 26–34. https://doi.org/10.1037/ per0000191 Open Science Collaboration. (2015). Estimating the reproducibility of psychological science. Science, 349, 943–951. https://doi. org/10.1126/science.aac4716 Ramsey, S. R, Thompson, K. L., McKenzie, M., & Rosenbaum, A. (2016). Psychological research in the Internet age: The quality of web-based data. Computers in Human Behavior, 58, 354– 360. https://doi.org/10.1016/j.chb.2015.12.049

Journal of Individual Differences (2020), 41(1), 30–36

S. V. Rouse, Reliability of MTurk Data from Masters and Workers

Rouse, S. V. (2015). A reliability analysis of Mechanical Turk data. Computers in Human Behavior, 43, 304–307. https://doi.org/ 10.1016/j.chb.2014.11.004 Thomas, K. A., & Clifford, S. (2017). Validity and Mechanical Turk: An assessment of exclusion methods and interactive experiments. Computers in Human Behavior, 77, 184–197. https:// doi.org/10.1016/j.chb.2017.08.038 Tosti-Kharas, J, & Conley, C. (2016). Coding psychological constructs in text using Mechanical Turk: A reliable, accurate, and efficient alternative. Frontiers in Psychology, 7, 1–9. https://doi. org/10.3389/fpsyg.2016.00741 Trapnell, P. D., & Paulhus, D. L (2012). Agentic and communal values: Their scope and measurement. Journal of Personality Assessment, 94, 39–52. https://doi.org/10.1080/00223891. 2011.627968 Viswanathan, M. (2005). Measurement error and research design. Thousand Oaks, CA: Sage.

Conflict of Interest There were no conflicts of interest in conducting this research.

History Received January 7, 2019 Revision received March 10, 2019 Accepted May 9, 2019 Published online September 10, 2019 Acknowledgments Both studies were presented at the 2019 conference of the Society for Personality and Social Psychology.

Journal of Individual Differences (2020), 41(1), 30–36

Publication Ethics This research was approved by the Seaver College Institutional Review Board (protocol #17-11-654) and followed the ethical standards of the American Psychological Association. Open Data The materials, data, and preregistration for Study 1 are archived at https://osf.io/e2387/. The data and preregistration for Study 2 are also archived at https://osf.io/e2387/; the materials for this second study are not archived because doing so would violate the usage agreement for the measure used. Funding This research was supported by an internal grant from Pepperdine University’s Seaver Research Council. ORCID Steve Rouse https://orcid.org/0000-0002-1080-5502 Steve Rouse Social Sciences Division Pepperdine University Malibu CA 90263 USA steve.rouse@pepperdine.edu

Ó 2019 Hogrefe Publishing

A concise guide to the assessment and treatment of insomnia for busy professionals “This book provides an excellent concise introduction to insomnia and its treatment with cognitive behavioral therapy. A great addition to any therapist’s bookshelf!” Philip Gehrman, PhD, CBSM, Associate Professor, Department of Psychiatry, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA

William K. Wohlgemuth / Ana Imia Fins

Insomnia (Series: Advances in Psychotherapy – Evidence-Based Practice – Volume 42) 2019, viii + 94 pp. US $29.80 / € 24.95 ISBN 978-0-88937-415-7 Also available as eBook About 40% of the population experiences difficulty falling or staying asleep at some time in a given year, while 10% of people suffer chronic insomnia. This concise reference written by leading experts for busy clinicians provides practical and upto-date advice on current approaches to assessment, diagnosis, and treatment of insomnia. Professionals and students learn to correctly identify and diagnose insomnia and gain hands-on information on how to

www.hogrefe.com

carry out treatment with the best evidence base: cognitive behavioral therapy for insomnia (CBT-I). The American Academy of Sleep Medicine (AASM) and the American College of Physicians (ACP) both recognize CBT-I as the first-line treatment approach to insomnia. Appendices include useful resources for the assessment and treatment of insomnia, which readers can copy and use in their clinical practice.

So that’s how my mind works – Now I get it! “As a fan of PSI theory for more than 20 years, I am very happy seeing it translated for popular consumption! The book makes the theory reasonably simple, with lots of fun illustrations.” Kennon M. Sheldon, PhD, Professor of Psychological Sciences, University of Missouri, Columbia, MO

Johannes Storch / Corinne Morgenegg / Maja Storch / Julius Kuhl

Now I Get It!

Understand Yourself and Take Charge of Your Behavior 2018, vi + 248 pp. US $34.80 / € 27.95 ISBN 978-0-88937-541-3 Also available as eBook Using the example of four colleagues working together in a small company, the book shows us the main personality types and their strengths and weakness in such a way that we gain real “now I get it!” insights into what is going on in our own and others’ subconscious. How does my mind work and what kind of personality do I have? When we can answer these questions and have come to terms with who we are, then the solutions to many issues that arise in everyday life will fall into place. What sort of people do I get on with best and how can I best deal

www.hogrefe.com

with the others? Are there recurring stressful situations in my professional or private life, and how do I resolve them? This humorously written and illustrated book, by the world’s leading experts in personality systems interaction (PSI) theory and the Zurich Resource Model (ZRM), gives us profound insights into our and other people’s subconscious thoughts – so we can adapt our own behavior and interactions to improve our quality of life. Cartoons and worksheets help us on our way.

Original Article

Separating Content and Structure in Humor Appreciation A Bimodal Structural Equation Modeling Approach Sonja Heintz Department of Psychology, University of Zurich, Switzerland

Abstract: The present study empirically tests a bimodal model of humor appreciation. In this model, individual differences in evaluating jokes and cartoons are attributed to their contents (sexual and aggressive) and structures (incongruity-resolution and nonsense). A total of 225 participants (64% women, Mage = 32.37 years) rated 50 jokes and cartoons on funniness, offensiveness, and boringness. They also completed a standard humor appreciation measure (the 3 WD). Using structural equation modeling, the bimodal model was found to be superior to alternative models. Regressions predicting the 3 WD categories supported the convergent and discriminant validity of the content and structure factors. In line with previous research, gender differences were found for the content factors, but not for the structure factors. Overall, this study is the first to show the viability of bimodal models of humor appreciation. They allow capturing the sources of individual differences in humor appreciation more adequately than previous models, thus providing a basis for future theories, research, and applications. Keywords: humor appreciation, individual differences, structural equation modeling, construct validity, behavior tests

Humor is a ubiquitous everyday phenomenon, which broadly entails everything funny (see Martin & Ford, 2018; Ruch, 2007, 2012). Individual differences in humor can be studied in different domains, including humor comprehension (how people understand humor), humor appreciation (how people evaluate humorous stimuli and events), and humor production (how people show humor). The present study focuses on humor appreciation, which has long been of interest to personality researchers (e.g., Cattell & Tollefson, 1966; Eysenck, 1942, 1944; Ruch, 1992; Ruch & Hehl, 2007). This research has shown that reactions to humorous stimuli (e.g., funniness ratings of jokes) are stable individual differences that relate to other personality traits, attitudes, intelligence, and art preferences. They can thus serve as indirect (or in Cattell’s sense objective) measures of personality (Cattell & Tollefson, 1966). Furthermore, the properties of humorous stimuli (mainly content and structure) were found to influence these individual differences. The present study extends this research by empirically testing a bimodal model of humor appreciation (as proposed by Ruch & Hehl, 2007; Ruch & Platt, 2012). In this bimodal model, both contents (sexual and aggressive) and structures (incongruity-resolution and nonsense) of humorous stimuli are considered simultaneously.

Ó 2019 Hogrefe Publishing

Content and Structure in Humor Appreciation The content and structure of humorous stimuli represent two sources that account for a large amount of variance in individual differences in humor appreciation (Carretero-Dios, Pérez, & Buela-Casal, 2010; Ruch, 1992; Ruch & Hehl, 2007). The content of humorous stimuli is an affective stimulus property with many variations; for example, a stimulus can be innocent/harmless, benevolent, sexual, satirical, sick, dark, scatological, or disparaging/ aggressive/hostile. The content is the most salient stimulus property and has been incorporated in early classifications (e.g., Mindess, Miller, Turek, Bender, & Corbin, 1985) and humor theories (for an overview, see Ferguson & Ford, 2008). Empirically, however, the structure of humorous stimuli was found to be more relevant for explaining individual differences (Ruch, 1992; Ruch & Hehl, 2007). Structure is a cognitive stimulus property and refers to how the joke “works” (i.e., the mechanisms underlying the punchline). This includes linguistic techniques (e.g., aggregation, repetition, and contradiction), incongruities (i.e., punchlines that are surprising and unexpected), and incongruity-resolution (i.e., making sense of punchlines). These different

Journal of Individual Differences (2020), 41(1), 37–44 https://doi.org/10.1027/1614-0001/a000301

mechanisms were formalized in detail in several models and theories of humor appreciation (e.g., Freud, 1905; Raskin, 1985; Suls, 1983). A similar distinction of cognitive (structure) and affective properties (contents) was also found in ratings of humorous materials. For example, Wicker, Thorelli, Barron and Ponder (1981) had two sets of jokes rated on 13 scales. They found one cognitive response dimension (originality, which included funniness) and two affective dimensions (emotionality and superiority). Later, Ruch and Rath (1993) had jokes and cartoons rated on 17 scales. They found one positive reaction to humor (funniness), which combined both affective and cognitive elements, and two negative reactions, which were either cognitive (boringness) or affective (offensiveness). It is thus important to go beyond “funniness” ratings to adequately capture different reactions to humorous stimuli.

Integrating Content and Structure Different theoretical and empirical approaches aimed at integrating both contents and structures (e.g., CarreteroDios et al., 2010; Eysenck, 1942; Freud, 1905; Godkewitsch, 1976; Ruch, 1992; Wicker et al., 1981). These approaches often derived content and structure dimensions based on principal component or exploratory factor analyses of ratings of humorous stimuli. Ruch (1992) had large sets of jokes and cartoons rated for funniness and aversiveness, resulting in the 3 WD (3 Witz-Dimensionen) humor test. He distinguished three stimulus categories: Incongruityresolution (INC-RES), which completely resolves incongruities, nonsense (NON), which contains partially or completely unresolved incongruities, and sexual humor (SEX), which contains sexual contents. Extending this approach, Carretero-Dios et al. (2010) added the three content categories of black, man-disparagement, and woman-disparagement humor. Thus, two structures (incongruity-resolution and nonsense) and several contents of humorous stimuli (mostly related to sex and aggression; see also Ferguson & Ford, 2008; Freud, 1905; Godkewitsch, 1976) were found to influence individual differences in humor appreciation. Extending these approaches, it was suggested that content and structure can contribute simultaneously to the appreciation of humorous stimuli with varying significance or salience (Eysenck, 1942; Freud, 1905; Godkewitsch, 1976; Ruch & Hehl, 2007; Ruch & Platt, 2012). However, existing classifications and measures usually only consider one of these properties in each stimulus; that is, a joke that has an incongruity-resolution structure and sexual content might be assigned to the SEX category because the content might be more dominant; conversely, a cartoon with a nonsense structure and aggressive content might be assigned to Journal of Individual Differences (2020), 41(1), 37–44

S. Heintz, Humor Appreciation: Content & Structure

the NON category because the structure is more dominant. Modeling both the structure and content properties of these stimuli simultaneously allows a more adequate representation and assessment of the sources of variance contributing to humor appreciation.

The Present Study A bimodal model of humor appreciation is needed to assess the contributions of both content and structure to individual differences in humor appreciation in each stimulus (see Ruch & Hehl, 2007; Ruch & Platt, 2012). This allows a better understanding of what makes humor funny, offensive, and boring to different people. This knowledge can then be used in future theories (e.g., models of humor appreciation), research (e.g., classifications of humor stimuli), and applications (e.g., selecting appropriate stimuli for humor interventions). The present study is the first to investigate whether such a bimodal model of humor appreciation can be empirically supported. Specifically, a set of jokes and cartoons that contain two contents (SEX and aggression, AGG) and two structures (INC-RES and NON) are rated for funniness, offensiveness, and boringness. These contents and structures were selected as the most prevalent ones in theoretical and empirical work on humor appreciation. To test the model, four different structural equation models are compared separate for each of the three ratings: Structure-only, content-only, four-factor, and bimodal models (see Figure 1). It is expected that the bimodal model will fit the data best. To test the convergent and discriminant validity of this bimodal model of humor appreciation, the two content and structure factors are related to the 3 WD as a standard test of humor appreciation (Ruch, 1992). It is expected that the largest correlations occur between the funniness ratings of the INC-RES, NON, and SEX factors and categories, respectively. The relationship between the offensiveness and boringness ratings in the bimodal model and the aversiveness ratings in the 3 WD are investigated exploratorily, though the largest correlations are again expected among the corresponding factors and categories (INC-RES, NON, and SEX). The relationship of the AGG factor with the 3 WD categories is expected to be lower, as the 3 WD contains only few aggressive contents that are spread across all three categories. Finally, gender differences in humor appreciation are explored. A recent systematic review showed that men and women appreciated INC-RES and NON humor to a similar extent, while men appreciated SEX and AGG humor more than women did (Hofmann, Platt, Lau, & TorresMarín, 2019). Gender differences are thus expected for the two content factors (with men appreciating them more than women), but not for the two structure factors. Ó 2019 Hogrefe Publishing

S. Heintz, Humor Appreciation: Content & Structure

Figure 1. Visualization of the four competing SEM models tested in the present study. (a) Bimodal model with two contents and two structures; (b) Unimodal model (content-only); (c) Unimodal model (structure-only); (d) Four-factor model.

Methods Sample A total of 232 participants completed the humorous stimuli. Seven participants were excluded from further analyses because they (a) provided implausible responses (n = 1), (b) completed the study too quickly (n = 1), and (c) indicated an age of less than 18 years (n = 5). The final sample thus consisted of 225 participants (64% women and 36% men) with a mean age of 32.37 years (SD = 12.20, range 18–87 years). Most participants were Swiss (50%) or German (43%). They were mostly well educated, with 45% having a university degree, 23% currently studying, 20% having a university-entrance diploma, and 9% having an apprenticeship.

Measures Humorous Stimuli The stimuli were selected in a two-step procedure. In the first step, two raters judged a large pool of jokes and cartoons regarding their content and structure to ensure the content validity of the items (see the Electronic Supplementary Material, ESM 1, for a detailed description of the initial Ó 2019 Hogrefe Publishing

stimuli selection). Of the final set of 50 jokes and cartoons, 13 contained sexual and nonsense humor (SEX-NON; example joke “How do you recognize a gay snowman? He has the carrot in his buttocks.”), 15 contained aggressive and nonsense humor (AGG-NON; e.g., “Three snails on a rail track. The first one says ‘Take care, a train is coming’ Crack! – ‘Where?’ Crack! – ‘There!’ Crack!”), 11 contained sexual and incongruity-resolution humor (SEX-INC-RES; e.g., “As her partner was getting sexually exhausted, the women threw a coin out of the window to the snake charmer and yelled ‘Continue playing a bit longer.’”), and 11 contained aggressive and incongruity-resolution humor (AGG-INC-RES; e.g., “The leading candidate of the party is speaking. His advertising consultant whispers: ‘Yesterday was better.’ – ‘But I didn’t give a speech yesterday!’ – ‘Exactly!’”). Each stimulus was rated for funniness, offensiveness, and boringness (see Ruch & Rath, 1993) on 5point scales (from 1 = not at all to 5 = very much). Four additional stimuli (one from each category) were employed as a warm-up and were not scored. Cronbach’s alphas, means, standard deviations, and intercorrelations among the initial 12 Content Structure Rating combinations are shown in Table S1 in ESM 1. The internal consistencies were sufficient for all combinations (.73–.94, Mdn = .85), yet the correlations among the combinations of the same ratings were high as well (all rs = .54–.89, all ps < .01, Mdn = .70). Journal of Individual Differences (2020), 41(1), 37–44

In a second step, these 50 stimuli were empirically refined by selecting the ones that were most representative for each combination and that differed from the other combinations (see ESM 1 for a detailed description of the empirical stimuli selection, Reiss & Reips, 2016). This resulted in a total of 29 stimuli for the funniness ratings (4–9 for each combination), 33 for offensiveness (4–11 each), and 24 for boringness (3–8 each). Cronbach’s alphas, means, standard deviations, and intercorrelations among the 12 refined Content Structure Rating combinations are shown in Table S2 in ESM 1. Again, the internal consistencies were sufficient for all combinations (.59–.93, Mdn = .74), while the correlations among the combinations of the same ratings were reduced (all rs = .32–.73, all ps < .01, Mdn = .55). Thus, a median 11% loss of true-score variance (which is expected due to the lower item number) was compensated by a reduction of the median overlap between the combinations of 19% (i.e., a better differentiation between the Content Structure combinations). Furthermore, confirmatory factor analyses supported that the resulting stimuli are unidimensional (see Table S3 in ESM 1). 3 WD (Ruch, 1992) The 3 WD assesses the appreciation of jokes and cartoons of the three humor categories INC-RES, NON, and SEX. Ten cartoons and jokes for each dimension plus five warm-up stimuli (not scored) are rated for funniness and aversiveness using 7-point Likert-type scales (from 1 to 7). Three funniness and three aversiveness scores were computed (for INC-RES, NON, and SEX). Cronbach’s α ranged from .79 (NON funniness) to .92 (INC-RES aversiveness), and the means ranged from 1.68 (INC-RES aversiveness, SD = 1.00) to 3.50 (INC-RES funniness, SD = 1.33).

S. Heintz, Humor Appreciation: Content & Structure

Analyses The bimodal, unimodal (content-only and structure-only), and four-factor models (one for each combination) were computed separately for the funniness, offensiveness, and boringness ratings. The two latent content (SEX and AGG) and structure factors (INC-RES and NON) were allowed to correlate with one another, respectively, but correlations between the content and structure factors were set to zero to ensure model identification. The convergent and discriminant validity with the 3 WD categories as well as the criterion validity with gender were tested in SEM regression analyses. Each of these variables were added as observed criteria in the bimodal SEM, and the four latent factors (SEX, AGG, INC-RES, and NON) served as predictors. The robust maximum likelihood estimator was used in all SEMs to accommodate deviations from the normal distribution (especially in the offensiveness ratings). The analyses were conducted in R (R Core Team, 2018), using the packages psych (Revelle, 2018) and lavaan (Rosseel, 2012). Fit indices were interpreted in line with the guidelines of Schermelleh-Engel, Moosbrugger, and Müller (2003): w2 (good: p > .05, acceptable: p .01), w2/df (good: 2, acceptable: 3), comparative fit index (CFI; good: .97, acceptable: .95), root mean square error of approximation (RMSEA; good: .05, acceptable: .08), and standardized root mean square residual (SRMR; good: .05, acceptable: .10).The BIC (sample-size adjusted Bayesian information criterion) was also computed, for which smaller values indicate a better fit to the data.

Results Model Comparisons

Procedure The study was conducted online (https://www.unipark.com/). Participants first completed the demographic items, followed by the 54 humorous stimuli. Each stimulus was shown on a separate page, followed by the three ratings (funniness, offensiveness, and boringness). Afterward, participants completed the 3 WD, also with each stimulus shown on a separate page. Other variables were assessed that are not relevant for the present study. Participants could receive a general feedback on the results of the study and course credit (for psychology students). An a priori power analysis was not feasible to determine the sample size, but at least 200 participants were recruited to conduct the structural equation modeling (SEM) analyses. The study was conducted in line with the local ethical guidelines. Journal of Individual Differences (2020), 41(1), 37–44

Table 1 shows the model fits of the unimodal (content-only or structure-only), bimodal, and four-factor models of the three ratings. For all ratings (funniness, offensiveness, and boringness), the bimodal models showed an acceptable fit. The two structures (INC-RES and NON) correlated positively (r = .42–.59, all ps < .01) and shared 18–35% of their true-score variance. The contents (r = .66–.75, all ps < .01) were also positively correlated and shared 44–56% of their true-score variance. These overlaps might represent general humor appreciation, as some people (e.g., those higher in cheerfulness; Ruch & Hofmann, 2012) tend to enjoy humorous stimuli more than others, independent of its content and structure. Alternative unimodal models that specified only two content factors or only two structure factors and a model that specified four latent factors (one for each Content Ó 2019 Hogrefe Publishing

S. Heintz, Humor Appreciation: Content & Structure

Table 1. Model fits of the bimodal, unimodal (content-only or structure-only), and four-factor models of the three ratings (funniness, offensiveness, and boringness) Rating and model

df w2/df CFI RMSEA SRMR

BIC

Funniness Bimodal

533.80*** 346 1.54 .847

.049

.068 18,495

Structure

674.98*** 376 1.80 .757

.059

.064 18,593

Content

766.48*** 376 2.04 .682

.068

.086 18,710

Four-factor

617.40*** 371 1.66 .800

.054

.076 18,531

Bimodal

564.19** 460 1.23 .920

.032

.060 15,401

Structure

755.30*** 494 1.53 .798

.048

.095 15,776

Content

676.51*** 494 1.37 .859

.041

.082 15,563

Four-factor

656.11*** 489 1.34 .871

.039

.077 15,528

Bimodal

320.99*** 226 1.42 .945

.043

.053 17,481

Structure

431.53*** 251 1.72 .895

.057

.066 17,559

Content

532.50*** 251 2.12 .837

.071

.076 17,665

Four-factor

390.64*** 246 1.59 .916

.051

.064 17,518

Offensiveness

Boringness

Notes. N = 225. CFI = comparative fit index; RMSEA = root mean square error of approximation; SRMR = standardized root mean square residual; BIC = sample-size adjusted Bayesian information criterion. **p < .01; ***p < .001.

Structure combination) showed a worse fit to the data, as indexed by all absolute and relative fit indices. These results support the hypothesis that both contents and structures should be distinguished simultaneously. A bimodal model adequately represented the variance in humor appreciation due to the content and structure properties within each humorous stimulus. Furthermore, some stimuli were mostly loaded by the content and others by the structure factor, which also varied between the three ratings (see Tables S4–S6 in ESM 1 for details). For example, Stimulus 27 (“Dad asks Max: ‘What did you do in physics today?’ – ‘Built a bomb!’ – Dad: ‘Alright, and what will you do tomorrow at school?’ – Max: ‘Which school?’”) was found funny both because of its structure (INC-RES) and its content (AGG), while people were offended by it mostly due its content, and they were bored by it mostly due its structure. To explore this pattern across the ratings, an arbitrary cut-off of .10 was set to interpret loadings as dissimilar. This showed that the funniness ratings were mostly influenced by the structure factor (16 of 29 stimuli), the offensiveness ratings were mostly influenced by the content factor (24 of 33 stimuli), and the boringness ratings were influenced by both the contents (8 of 24 stimuli) and structures (9 of 24 stimuli).

Validity of the Bimodal Model The regressions were conducted separately for the three ratings, with the four content and structure factors as Ó 2019 Hogrefe Publishing

predictors and the 3 WD scores and gender as criteria (see Table 2). As expected, convergent validity was established between the funniness ratings of the INC-RES, NON, and SEX categories of the 3 WD with the corresponding funniness ratings of the three latent factors. The SEX-factor of the boringness ratings positively predicted the SEX category of the 3 WD (aversiveness rating; p = .012), and the INC-RES-factor significantly and negatively predicted the INC-RES-category of the 3 WD (funniness rating).The offensiveness ratings showed a convergence of the SEX-factor with the aversiveness ratings of the SEX-category in the 3 WD. The AGG-factor was spread across the three 3 WD categories and did not significantly predict any category. This is to be expected as aggressive contents were not targeted in the 3 WD. In terms of gender, men found the SEX stimuli funnier and less boring than women, which is in line with the expectations. No significant prediction was found for the AGG stimuli. Finally, women found NON-stimuli more offensive than men, which was not hypothesized.

Discussion The first aim on the present study was to empirically test a bimodal model of humor appreciation (Ruch & Hehl, 2007; Ruch & Platt, 2012), in which two structure and content factors were specified. As expected, this model showed a descriptively better fit to alternative models. In other words, considering the structure and content of humorous stimuli simultaneously as latent factors better represented individual differences in funniness, offensiveness, and boringness ratings of a diverse set of jokes and cartoons. This is in line with previous theoretical and empirical work that took both stimulus properties into account (e.g., Carretero-Dios et al., 2010; Freud, 1905; Godkewitsch, 1976; Ruch, 1992). Neglecting either contents or structures could lead to erroneous conclusions; for example, if humorous stimuli have different contents and structures, but only one of these properties is modeled, the results and interpretations would be biased (see also Ruch & Hehl, 2007; Ruch & Platt, 2012). Additionally, the contribution of contents and structures to the appreciation of humorous stimuli differed across the individual stimuli and across the three ratings (funniness, offensiveness, and boringness). The bimodal approach used in the present study allows, for the first time, to adequately model these different sources of humor appreciation. This finding also supports the need for a careful stimuli selection and evaluation to specify the loadings in the bimodal models in future research. This should include expert ratings on the content and structure of the Journal of Individual Differences (2020), 41(1), 37–44

S. Heintz, Humor Appreciation: Content & Structure

Table 2. Standardized regression weights with the four latent factors as predictors and gender and the 3 WD categories as criteria Rating and factor

3 WD funniness Gender INC-RES NON

3 WD aversiveness

SEX INC-RES NON

SEX

Funniness INC-RES

.06

.65**

.21** .37** .07

.09

.02

NON

.01

.06

.52** .25** .06

.10

SEX

.19**

.29**

.20

.65** .02

.10

.42**

AGG

.08

.10

.22

.25

.27

.25

INC-RES

.03

.04

.25

.01

.03

.11

.07

NON

.16**

.00

.15

.12

.08

.27

.36**

.18

(Hofmann et al., 2019). As previous studies did however not simultaneously model the content and structure properties of the humorous stimuli, it is unclear to what extent these gender differences are comparable. Future studies should extend this initial validation by relating the content and structure factors to other individual-difference variables, such as self-reports of humor (e.g., styles related to mockery and nonsense; see Ruch et al., 2018).

Offensiveness

SEX

.05

.03

.20

.24

.35** .44**

AGG

.14

.16

.18

.02

.01

.03

.02

.56**

.25

.01

.12

.03

.06

Boringness INC-RES NON

.05

.15

.12

.07

.25

.18

SEX

.21**

.02

.20

.06

.04

.49

AGG

.11

.36

.26

.18

.02

.06

.24

Note. N = 225. AGG = aggressive; INC-RES = incongruity-resolution, NON = nonsense; SEX = sexual humor. Gender coded as 1 = men, 2 = women. Fully standardized solution reported for the 3 WD, while only latent variables were standardized for gender. Convergent regression weights in bold. **p < .01.

humorous stimuli to ensure their content validity (see Carretero-Dios, Pérez, & Buela-Casal, 2009). In the present study, the stimuli were selected to have both a salient content and structure. Future studies should extend these ratings to further contents (e.g., dark, innocent/harmless, benevolent, or satirical) and structures (e.g., different kinds of incongruity-resolution or nonsense; see Raskin, 1985; Raskin, Hempelmann, & Taylor, 2009). This would allow testing how many content and structure factors can be distinguished, thus providing an empirical means of developing a comprehensive taxonomy of humor appreciation. Furthermore, if stimuli are included that mainly contain a salient content or structure, the bimodal model can also include latent correlations between the content and structure factors to better understand their mutual interplay. The second aim of the present study was to investigate the construct and criterion validity of these two content and structure factors by relating them to the 3 WD and gender. Convergent validity was established with the 3 WD categories for INC-RES, NON, and SEX for the funniness ratings. As anticipated, the convergence was lower for the offensiveness and boringness ratings, showing that the negative reactions to humorous stimuli (aversiveness vs. offensiveness and boringness) captured different aspects. Furthermore, the gender differences reported in the literature were replicated for SEX (men appreciating it more than women) and INC-RES (no gender differences), while the previous findings were not replicated for NON and AGG Journal of Individual Differences (2020), 41(1), 37–44

Limitations and Future Directions The present study only captured two structures and contents as a minimal test of the bimodal model of humor appreciation. Thus, replicating and extending the present approach to further structures and contents is an important next step in research. Comparing these latent factors shows how many active stimuli ingredients can be distinguished in humor appreciation. Also, investigating the nomological network of these contents and structures would showcase the relevance of humor appreciation in terms of traits, habits, abilities, attitudes, and interests, building on the idea that humor appreciation can serve as an implicit personality measure (e.g., Cattell & Tollefson, 1966). Next, the stimuli in the present study were selected to capture both a salient content and structure. Selecting stimuli a-priori that primarily contain a certain structure or content as well as mixed stimuli would enable testing interactions between these two stimulus properties in determining a person’s evaluation of the funniness, offensiveness, and boringness of humorous stimuli (see Godkewitsch, 1976). Also, people likely differ in how they perceive and interpret the content and structure of a stimulus (i.e., individual differences in humor comprehension and the response dimensions of humor appreciation). This would provide another direction for future studies, for example, by having open responses on the comprehension of humorous stimuli rated for the different contents and structures, by using eye-tracking methods, or by having the stimuli rated for the extent to which different contents and structures are perceived (as was done by Wicker et al., 1981). Furthermore, the large number of stimuli in the models resulted in a ratio of participants to estimated parameters of 2:1–3:1, rather than the recommended minimum of 10:1 (e.g., Kline, 2011). Thus, the obtained parameter estimates might not be stable and are in need of replication in new and larger samples. Importantly, additional analyses conducted using parcels (with a ratio of 6:1–9:1) replicated the present findings (three parcels for each of the 12 Content Structure Rating combinations using a radial parceling approach as recommended by Little, Rhemtulla, Gibson, & Schoemann, 2013), yet future studies should Ó 2019 Hogrefe Publishing

S. Heintz, Humor Appreciation: Content & Structure

increase the ratio beyond 10:1. Also, post-hoc power analyses (conducted with the WebPower package, Zhang & Yuan, 2018; see Tables S4–S6 in ESM 1) showed that the power of 82.1% (133 of 162) of the estimated parameters was higher than .80. The code for the power analyses is provided in ESM 1 to allow planning the sample sizes of future studies. Lastly, females, younger adults, and well-educated people were overrepresented in the sample. Thus, future studies should employ larger and more representative samples to replicate the present findings. Replications in different cultures (see e.g., Carretero-Dios & Ruch, 2010; Eysenck, 1944) would also be desirable.

Conclusion Overall, the present study suggests that it is viable to model both contents and structures of humorous stimuli simultaneously to adequately represent individual differences in humor appreciation (Ruch & Hehl, 2007; Ruch & Platt, 2012). Such a bimodal model was found to outperform alternative models in all three ratings (funniness, offensiveness, and boringness). Furthermore, the construct and criterion validity of the content and structure factors in the model received initial support. This study thus showcases a promising future direction for theories, research, and applications of individual differences in humor appreciation, and possibly also other humor domains such as humor production and comprehension.

Electronic Supplementary Material The electronic supplementary material is available with the online version of the article at https://doi.org/10.1027/ 1614-0001/a000301 ESM 1. This document provides detailed description of the initial stimuli selection, the empirical stimuli selection, additional analyses, and the code for the power analyses.

References Carretero-Dios, H., Pérez, C., & Buela-Casal, G. (2009). Content validity and metric properties of a pool of items developed to assess humor appreciation. The Spanish Journal of Psychology, 12, 773–787. https://doi.org/10.1017/S1138741600002146 Carretero-Dios, H., Pérez, C., & Buela-Casal, G. (2010). Assessing the appreciation of the content and structure of humor: Construction of a new scale. Humor: International Journal of Humor Research, 23, 307–325. https://doi.org/10.1515/ humr.2010.014 Carretero-Dios, H., & Ruch, W. (2010). Humor appreciation and sensation seeking: Invariance of findings across culture and assessment instrument? Humor: International Journal of

Ó 2019 Hogrefe Publishing

Humor Research, 23, 427–445. https://doi.org/10.1515/ HUMR.2010.020 Cattell, R. B., & Tollefson, D. L. (1966). IPAT Humor Test of Personality. Champaign, IL: Institute for Personality and Ability Testing. Eysenck, H. J. (1942). The appreciation of humour: An experimental and theoretical study. British Journal of Psychology, 32, 295–309. https://doi.org/10.1111/j.2044-8295.1942.tb01027.x Eysenck, H. J. (1944). National differences in “sense of humor”: Three experimental and statistical studies. Journal of Personality, 13, 37–54. https://doi.org/10.1111/j.1467-6494.1944. tb01971.x Ferguson, M. A., & Ford, T. E. (2008). Disparagement humor: A theoretical and empirical review of psychoanalytic, superiority, and social identity theories. Humor: International Journal of Humor Research, 21, 283–312. https://doi.org/10.1515/ HUMOR.2008.014 Freud, S. (1905). Der Witz und seine Beziehung zum Unbewussten [The joke and its relation to the unconscious]. Vienna, Austria: Deutike. Godkewitsch, M. (1976). Thematic and collative properties of written jokes and their contribution to funniness. Canadian Journal of Behavioural Science/Revue canadienne des sciences du comportement, 8, 88–97. https://doi.org/10.1037/h0081937 Hofmann, J., Platt, T., Lau, C., & Torres-Marín, J. (2019). Gender differences in humor traits, appreciation, production, comprehension, (neural) responses, use, and correlates: A systematic review. Manuscript in preparation Kline, R. B. (2011). Principles and practice of structural equation modeling (3rd ed.). New York, NY: Guilford Press. Little, T. D., Rhemtulla, M., Gibson, K., & Schoemann, A. M. (2013). Why the items versus parcels controversy needn’t be one. Psychological Methods, 18, 285–300. https://doi.org/10.1037/ a0033266 Martin, R. A., & Ford, T. E. (2018). The psychology of humor: An integrative approach (2nd ed.). Oxford, UK: Academic Press. Mindess, H., Miller, C., Turek, J., Bender, A., & Corbin, S. (1985). The Antioch Humor Test: Making sense of humor. New York, NY: Avon. R Core Team. (2018). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from https://www.r-project.org/ Raskin, V. (1985). Semantic mechanisms of humor. Dordrecht, The Netherlands: D. Reidel Publishing. Raskin, V., Hempelmann, C. F., & Taylor, J. M. (2009). How to understand and assess a theory: The evolution of the SSTH into the GTVH and now into the OSTH. Journal of Literary Theory, 3, 285–311. https://doi.org/10.1515/JLT.2009.016 Reiss, S., & Reips, U.-D. (2016). Online assessment. In K. Schweizer & C. DiStefano (Eds.), Principles and methods of test construction: Standards and recent advances (pp. 120– 134). Göttingen, Germany: Hogrefe. Revelle, W. (2018). Psych: Procedures for personality and psychological research (R package version 1.8.12) [Computer software]. Evanston, IL: Northwestern University. Retrieved from https://CRAN.R-project.org/package=psych Rosseel, Y. (2012). lavaan: An R Package for Structural Equation Modeling. Journal of Statistical Software, 48(2), 1–36. https:// doi.org/10.18637/jss.v048.i02 Ruch, W. (1992). Assessment of appreciation of humor: Studies with the 3 WD Humor Test. In C. D. Spielberger & J. N. Butcher (Eds.), Advances in personality (pp. 27–75). Hillsdale, NJ: Erlbaum. Ruch W. (Ed.). (2007). The sense of humor: Explorations of a personality characteristic (2nd ed.). Berlin, Germany: Mouton de Gruyter.

Journal of Individual Differences (2020), 41(1), 37–44

Ruch, W. (2012). Towards a new structural model of the sense of humor: Preliminary findings. In V. Raskin & J. M. Taylor (Chairs), Artificial intelligence of humor: Papers from the 2012 AAAI Fall Symposium (AAAI Technical Report FS-12-02) (pp. 68–75). Retrieved from http://www.ilhaire.eu/pdf/Ruch_ 2012.pdf Ruch, W., & Hehl, F.-J. (2007). A two-mode model of humor appreciation: Its relation to aesthetic appreciation and simplicity-complexity of personality. In W. Ruch (Ed.), The sense of humor: Explorations of a personality characteristic (2nd ed., pp. 109–142). Berlin, Germany: Mouton de Gruyter. Ruch, W., Heintz, S., Platt, T., Wagner, L., & Proyer, R. T. (2018). Broadening humor: Comic styles differentially tap into temperament, character, and ability. Frontiers in Psychology, 9, 6. https://doi.org/10.3389/fpsyg.2018.00006 Ruch, W., & Hofmann, J. (2012). A temperament approach to humor. In P. Gremigni (Ed.), Humor and health promotion (pp. 79–113). Hauppauge, NY: Nova Science. Ruch, W., & Platt, T. (2012). Separating content and structure in humor appreciation: The need for a bimodal model and support from research into aesthetics. In A. Nijhold (Ed.), 3rd International workshop on computational humor (pp. 23–27). Amsterdam, The Netherlands: University of Amsterdam. Ruch, W., & Rath, S. (1993). The nature of humor appreciation: Toward an integration of perception of stimulus properties and affective experience. Humor: International Journal of Humor Research, 6, 363–384. https://doi.org/10.1515/humr.1993.6. 4.363 Schermelleh-Engel, K., Moosbrugger, H., & Müller, H. (2003). Evaluating the fit of structural equation models: Tests of significance and descriptive goodness-of-fit measures. Methods of Psychological Research Online, 8, 23–74. Retrieved from http://www.mpr-online.de

Journal of Individual Differences (2020), 41(1), 37–44

S. Heintz, Humor Appreciation: Content & Structure

Suls, J. (1983). Cognitive processes in humor appreciation. In P. E. McGhee & J. H. Goldstein (Eds.), Handbook of humor research (pp. 39–57). New York, NY: Springer. Wicker, F. W., Thorelli, I. M., Barron, W. L., & Ponder, M. R. (1981). Relationships among affective and cognitive factors in humor. Journal of Research in Personality, 15, 359–370. https://doi. org/10.1016/0092-6566(81)90033-7 Zhang, Z. & Yuan, K.-H. (Eds.). (2018). Practical statistical power analysis using Webpower and R. Granger, IN: ISDSA Press. History Received December 4, 2018 Revision received May 21, 2019 Accepted May 23, 2019 Published online September 10, 2019 Acknowledgments I would like to thank Willibald Ruch for his helpful feedback on the design of study, Alex Junghans for his help in collecting and selecting the humorous stimuli, and the raters for their help in selecting the final set of humorous stimuli. ORCID Sonja Heintz https://orcid.org/0000-0002-6229-7095 Sonja Heintz Section on Personality and Assessment Department of Psychology University of Zurich Binzmühlestrasse 14/7 8050 Zurich Switzerland s.heintz@psychologie.uzh.ch

Ó 2019 Hogrefe Publishing

Original Article

Positive and Negative Wayfinding Inclinations, Choice of Navigation Aids, and How They Relate to Personality Traits Chiara Meneghetti , Francesco Grimaldi, Massimo Nucci, and Francesca Pazzaglia Department of General Psychology, University of Padova, Italy

Abstract: This study aimed to examine the relationship between people’s self-reported wayfinding inclinations, their preference for certain navigation aids (maps vs. GPS vs. verbal directions), and their personality traits. A sample of 222 undergraduates completed questionnaires on personality traits, wayfinding inclinations and preferred navigation aids, and two spatial tasks. The results showed that spatial ability, positive wayfinding inclinations and negative wayfinding inclinations are distinct factors. Only wayfinding inclinations were related to personality traits: positive inclinations correlated positively, and negative inclinations inversely with Extraversion, Agreeableness, and Openness. Negative inclinations were only associated with poor Emotional stability. Further, Conscientiousness and Openness were correlated with a preference for map use, and Agreeableness with a preference for verbal directions. Analyzing facets of these personality traits clarified the relations. These findings are discussed within the spatial cognition domain, broadening the array of individual factors (such as spatial attitudes and personality traits) and their relation to consider in defining individual spatial profiles. Keywords: personality traits, wayfinding inclinations, navigation aids

Many individual factors play a central part in environment learning and navigation. Spatial performance relates to individual spatial competence, represented both by spatial abilities, such as mental rotation and visuospatial working memory, and by spatial self-reports (Hegarty, Montello, Richardson, Ishikawa, & Lovelace, 2006). The latter include variables such as: Sense of Direction (SoD), in one’s own estimation (Kozlowski & Bryant, 1977); individuals’ preferences for representing environments (Bryant, 1997; Klatzky, 1998) based on landmarks in relation to one another (allocentric), and landmarks in relation to a person’s viewpoint (egocentric), the former of which are also called orientation (Lawton, 1994) or survey (Pazzaglia & Meneghetti, 2017) strategy, and the latter route strategy (Lawton, 1994). Spatial self-reporting can also concern spatial anxiety (Lawton, 1994), and pleasure in exploring new places as opposed to a preference for navigating in well-known environments (Meneghetti, Borella, Pastore, & De Beni, 2014). Such spatial self-reported measures, though related to one another, can be classified as coinciding with positive or negative wayfinding inclinations (Meneghetti et al., 2014): the former include functional

Ó 2019 Hogrefe Publishing

environment and navigation attitudes, such as a strong SoD, a preference for survey representations, and pleasure in exploring new places; the latter include less functional attitudes, such as spatial anxiety and a preference for navigating in well-known places (see also Muffato, Toffalini, Meneghetti, Carbone, & De Beni, 2017). Positive and negative inclinations correlate inversely with spatial cognitive skills, but when both types are considered together, the positive (but not the negative) inclinations are functionally related to environment-learning performance (as in pointing task; Meneghetti et al., 2014; Muffato et al., 2017). Hence the importance of distinguishing between these attitudes. These factors remain fairly stable over time (Münzer, Fehringer, & Kühl, 2016), and can be seen as stable individual behavioral or emotional traits (as in the case of spatial anxiety; Lawton, 1994). They can therefore plausibly be expected to correlate with habitual patterns of behavior, thinking and emotions, including personality traits. Support for this assumption derives from some studies suggesting a link between SoD and traits like Flexibility (Bryant, 1982) and, using the Big Five model (Goldberg,

Journal of Individual Differences (2020), 41(1), 45–52 https://doi.org/10.1027/1614-0001/a000302

C. Meneghetti et al., Personality Traits, Wayfinding Inclinations and Navigation

1990), with Extraversion, Intellect/Openness, Conscientiousness, and Emotional stability (correlation range .22/ .32), but not with Agreeableness (Condon et al., 2015). Pazzaglia, Meneghetti, and Ronconi (2018) found that pleasure in exploring new places (considered as a positive inclination) correlated positively with Extraversion (Dynamism, Dominance facets), Openness (Openness to experience facet), and Conscientiousness (Perseverance facet; correlation range .19/.22). On the other hand, spatial anxiety (considered as a negative inclination) correlated inversely (i.e., negatively) with Extraversion (Dynamism facet), Openness (Openness to experience facet), and Emotional stability (Emotion control facet) (correlation range .28/ .37), and positively with Conscientiousness (Scrupulousness facet, .19). No association with spatial abilities (such as mental rotation) emerged in either of the abovementioned studies, however (Condon et al., 2015; Pazzaglia et al., 2018). These results suggest that considering a variegated set of spatial inclinations can shed light on their multifaceted relationships with general personality (traits and facets), but evidence of this is still limited. Another indicator of spatial behavior is the type of navigation aid people prefer to use. One often-studied navigation aid is the GPS, which does not facilitate the construction of cognitive maps (Münzer, Zimmer, Schwalm, Baus, & Aslan, 2006), whereas maps (Münzer et al., 2016) seem to favor the mental construction of integrated spatial representations (Gardony, Brunyé, Mahoney, & Taylor, 2013). Another aid to navigation involves asking for verbal directions (Wen, Ishikawa, & Sato, 2014). Preferences concerning navigation aids remain stable, too (Münzer et al., 2016), and can be seen as spatial dispositions, but how they relate to personality traits (and facets) remains to be seen. Based on the above brief review, the goals of this study were twofold: the first aim was to elucidate how individual spatial competence is organized, assessing whether it is best represented by a one-factor (e.g., Thurstone & Thurstone, 1941), two-factor (Hegarty et al., 2006), or three-factor (spatial ability; positive and negative inclination; Meneghetti et al., 2014) model (see rationale for the models tested in Results section). After ascertaining the best model (based on three factors), our second aim was to explore to what extent these three factors relate to personality traits (and their facets), as previously suggested (Condon et al., 2015; Pazzaglia et al., 2018). This type of investigation enables a better delineation of individual spatial profiles – based on selfreports and objective spatial abilities (not actual wayfinding) – in relation to general personality traits (and facets). We also examined how preferences for different navigation aids related to people’s spatial inclinations (as in Münzer et al., 2016), and personality traits (and facets). Journal of Individual Differences (2020), 41(1), 45–52

Method Participants A total of 222 undergraduates (142 females; Mage = 20.52, SD = 1.36) at the School of Psychology at the University of Padua (Italy) voluntarily participated in the study, which was approved by the local ethics committee for psychological research (Number: 98A6FA618641EF507725E6828432EEEE).

Materials Big Five Personality Questionnaire (BFQ; Caprara, Barbaranelli, Borgogni, & Perugini, 2008) There are 134 statements referring to 5 traits (2 facets for each trait), 12 items for each facet (6 positive, 6 negative) and social desirability (14 items), scored on a Likert scale (from 1 to 5): Extraversion, that is, expansiveness (Dynamism), and assertiveness (Dominance); Agreeableness, that is, sensitivity toward others (Cooperativeness), and kindness (Politeness); Conscientiousness, that is, selfregulation in orderliness and precision (Scrupulousness), and tenacity and persistence (Perseverance); Emotional stability, that is, coping with one’s own anxiety and emotionality (Emotion control), and controlling one’s own emotional state (Impulse control); Openness, that is, breadth of cultural interests (Openness to culture), and receptivity to new experiences (Openness to experience). The internal consistency of each facet is good ( .80 in our sample; .78 in the normative sample; Caprara et al., 2008; see Table 1 for details). Scores are calculated as the sum of item ratings, reversing the negative ones, for each facet. Self-Reported Wayfinding Inclinations Sense of Direction and Spatial Representation Questionnaire (SDSR; Pazzaglia & Meneghetti, 2017) There are 11 items covering three factors (using a 5-point Likert scale): SoD – preference for survey mode (4 items); Knowledge and use of cardinal points (3 items); preference for landmark/route mode (4 items). The internal consistency is good ( .70 in our sample, see Table 1; .78 in the normative sample; Pazzaglia & Meneghetti, 2017). The score is the sum of the item ratings comprising each factor. Attitudes to Orientation Tasks Scale (AtOT; De Beni, Meneghetti, Fiore, Gava, & Borella, 2014) There are 10 items assessing two factors (using a 6-point Likert scale): 5 measure pleasure in exploring new places and 5 measure a preference for navigating in well-known Ó 2019 Hogrefe Publishing

Ó 2019 Hogrefe Publishing

.22

.21

.12

.02

.16

.35

.25

.34

.15

.18

.24

12. Perseverance

13. Emotion control

14. Impulse control

15. Openness to culture

16. Openness to experience

17. Spatial anxiety

18. Sense of directionpreference for survey mode

19. Knowledge and use of cardinal points

20. Preference for landmark/ route mode

21. Pleasure in exploring new places

22. Preference for navigating in well-known places

.03

.18

.11

.18

.12

.02

.13

.01

.10

.41

.31

.18

.12

.23

.31

.45

.23

.15

.22

.32

.02

.21

.33

.51

.08

.01

.12

.10

.02

.17

.23

.14

.20

.17

.08

.16

.25

.15

.29

.20

.17

.13

.04

.28

.01

.12

.16

.17

.15

.06

.11

.09

.33

.29

.03

.05

.14

.10

.60

.10

.27

.03

.08

.03

.13

.09

.03

.06

.10

.21

.10

.25

.08

.01

.04

.03

.05

.02

.03

.20

.13

.04

.13

.03

.04

.10

.07

.25

.08

.21

.33

.01

.02

.03

.01

.08

.09

.15

.08

.07

.11

.12

.18

.04

.15

.04

.11

.05

.04

.17

.09

.04

.11

.17

.29

.19

.11

.54

.02

.15

.06

.08

.01

.07

.06

.08

.15

.09

.19

.10

.04

.14

.09

.14

.04

.14

.11

.20

.09

.28

.20

.03

.41

.03

.02

.12

.14

.44

.31

.08

.22

.21

.27

.88

.87

.89

.87

.82

.85

8.01

.80

5.47

.80

6.18

.82

7.20

.88

7.12

Note. M = Mean; SD = Standard Deviation. *Biserial correlations for dichotomous variable; r .18 p .01.

Cronbach’s α

7.23

.88

8.56

.82

7.75

.82

7.40

.85

6.96

12.06

.10

.07

.15

.11

.17

.32

.30

.10

.29

.25

.17

.83

.85

.14

.40

90.37 40.15 34.05 47.36 45.66 41.79 43.92 31.00 36.16 44.84 45.53

.04

.15

.06

.07

.03

.14

.08

.07

.15

.28

.16

.05

.86

.89

.11

.17

.18

.01

.05

.12

13.24 10.42 11.69 14.31

.01

.04

.02

.17

.03

.11

.17

.04

.02

.01

.03

.27

.03

.81

.82

.02

.14

.30

.19

.18

.04

74.20 93.02 85.71 67.15

.08

.04

.25

.01

.11

.10

.16

.13

.05

.09

.10

.30

.21

.16

.02

.08

.03

.91

.88

.17

.30

.10

.06

.11

.01

.06

.07

.32

.31

.18

.15

.25

.27

.29

.22

.23

.21

.35

.11

.01

.10

.88

.85

.30

.01

.28

.06

.18

.14

11. Scrupulousness

27. Short Object Perspective Taking task

.19

10. Politeness

.38

.26

9. Cooperativeness

26. Short Mental Rotations Test

.29

8. Dominance

.28

.01

7. Dynamism

25. Navigation aids: verbal indications

.10

6. Openness

.15

.19

5. Emotional stability

.07

.22

24. Navigation aids: GPS

.25

4. Conscientiousness

23. Navigation aids: maps

.18

2. Extraversion

3. Agreeableness

1. Gender*

Table 1. Correlations, descriptive statistics, and internal consistency of variables

.19

.16

.25

.22

.09

.17

.07

.33

.44

.16

.85

6.84

.81

.75

5.10 2.79

23.23 18.53 6.17

.16

.02

.08 .02

.13 .13

.55

.72

.31

.48

.02

.66

.48

.19

.33

.54

.05

.01

.08

.03

.12

.20

.15

.01

.11

.20

.55

.15

.09

.01

.17

.12

.05

.08

.04

.34

.06

.01

.11

.02

.08

.43

.70

2.25

.70

4.18

.84 –

–

.79

5.21 1.53 1.26 1.27 2.69

.65

89.70

15.09 18.73 16.54 2.52 4.91 3.56 5.23 136.44

.31

C. Meneghetti et al., Personality traits, wayfinding inclinations and navigation 47

Journal of Individual Differences (2020), 41(1), 45–52

C. Meneghetti et al., Personality Traits, Wayfinding Inclinations and Navigation

places. The internal consistency is good (.70 and .84, respectively, in our sample; .78 in the normative sample; De Beni et al., 2014). The score is the sum of the item ratings comprising each factor. Spatial Anxiety Scale (SAS; De Beni et al., 2014; Adapted From Lawton, 1994) There are 8 items assessing (using a 6-point Likert scale) the degree of space-related anxiety experienced in everyday spatial tasks. The internal consistency is good (.85 in our sample; .87 in the normative sample; De Beni et al., 2014). The score is the sum of each item rating. Navigation Aids Questions (NAQ) There are 3 questions about navigation aids (adapted from Münzer et al., 2016) assessing how much individuals use maps, GPS, and verbal directions to reach a place, using 6-point Likert scale.

Spatial Tasks Short Mental Rotations Test (sMRT; De Beni et al., 2014; Adapted From Vandenberg & Kuse, 1978) There are 10 items that each involve identifying two of four abstract 3D-objects matching a target object but in a rotated position. The maximum time allowed is 5 min. The internal consistency is good (.79 in our sample; .90 in the normative sample; De Beni et al., 2014). One point is awarded if both correct options are detected from among the four (maximum score: 10). Short Object Perspective Taking Task (sOPT; De Beni et al., 2014; Adapted From Kozhevnikov & Hegarty, 2001) There are 6 items that involve imagining standing at one object in a 7-object conﬁguration, facing another, and marking the direction of a third on a circle. The time limit is 5 min. The internal consistency was moderate (.65) in our sample, and high (.90) in the normative sample (De Beni et al., 2014). Scores are calculated as sum of the degrees of difference between the angles identified and the right answers (i.e., higher scores meaning greater errors).

Procedure Participants completed the BFQ in smalls groups in a quiet room during the first session. All other measures (in four different orders) were administered at the second session.

Journal of Individual Differences (2020), 41(1), 45–52

Results The correlations between all measures, descriptive statistics, and internal consistency are shown in Table 1. All analyses were run using R software (R Core Team, 2017).

Factor Compositions of Self-Reported Wayfinding Inclinations and Spatial Tasks Using Confirmatory Factor Analyses (CFAs), three models were tested with: (1) one-factor comprising all spatial measures, based on the idea that visuo-spatial competence is a single factor (e.g., Thurstone & Thurstone, 1941); (2) two-factors, based on the assumption that objective spatial ability and spatial self-reports are distinct (as previously suggested; Hegarty et al., 2006), so the model included a spatial ability factor (sMRT and sOPT) and a wayfinding inclination factor (all self-reports); and (3) three-factors, with spatial ability as a distinct factor (as in the two-factor model), and distinguishing positive wayfinding inclinations (SoD-preference for survey mode, knowledge and use of cardinal points, preference for landmark/route mode, pleasure in exploring new places) from negative ones (preference for navigating in well-known places, and spatial anxiety), as previously suggested (Meneghetti et al., 2014). The fit indices considered for judging the goodness of the models were: the chi-square (w2) test, where a p-value > .05 indicates a good fit; the Root Mean Square Error of Approximation (RMSEA; Steiger & Lind, 1980), where small is good (< .06 is acceptable; Hu, & Bentler, 1999); the Standardized Root Mean Square Residual (SRMR; Jöreskog & Sörbom, 1981), where small is good (< .08 is acceptable, Hu, & Bentler, 1999); the Non-Normed Fit Index (NNFI; Tucker & Lewis, 1973), normally ranging from 0 to 1, though it can sometimes fall beyond this range, where large is good (> .95 is acceptable; Hu, & Bentler, 1999). The Bayesian Information Criterion (BIC; Schwarz, 1978) was used as a comparative criterion, and the model with the lowest BIC was preferred. Table 2 shows the fit indices for the three models. The single-factor model had unsatisfactory fit indices and the cumulative variance explained was .36. The two-factor model had better, but still unsatisfactory fit indices, and the cumulative variance explained was .45. The three-factor model had the best fit indices of the three, which were considered satisfactory, and the cumulative variance explained was .54. The BIC of the three-factor model was also lower ( 34.50) than that of the two-factor ( 21.65) and onefactor ( 13.76) models, with a ΔBIC = 12.85 in favor of the three-factor model.

Ó 2019 Hogrefe Publishing

C. Meneghetti et al., Personality traits, wayfinding inclinations and navigation

Table 2. Fit indices for the three models tested RMSEA

SRMR

NNFI

BIC

Small is good

Large is good

Lower is good

1. One-factor model

w2(20) = 94.29, p < .001

.13

.10

0.802

2. Two-factor model

w2(13) = 48.58, p < .001

.11

.06

0.854

21.65

w2(7) = 3.32, p = .85

.001

.02

1.03

34.50

3. Three-factor model

13.76

Note. RMSEA = Root Mean Square Error of Approximation; SRMR = Standardized Root Mean Square Residual; NNFI = Non-Normed Fit Index; BIC = Bayesian Information Criterion. In bold, the three-factor model that showed the best fit indices.

The NAQ ratings were not included in the models because doing so made the fit indices inadequate. We thus proceeded with the subsequent analyses considering the three-factor model (Figure 1).

Relations Between Personality Traits/ Facets, Wayfinding Inclination Factors, and Navigation Aids Used Table 3 shows the correlations between personality traits/ facets (standardized scores for each trait and facet calculated on the normative sample; Caprara et al., 2008) and wayfinding inclinations and preferred navigation aids, considering values associated with a p .01. The pattern of correlations between personality traits/ facets and positive/negative wayfinding inclinations showed a partial overlap (even with an inverse valence). The positive inclinations factor correlated positively with: Extraversion (facets: Dynamism r = .29; and Dominance r = .21); Agreeableness (facet: Cooperativeness r = .20); Openness (facets: Openness to culture r = .24 and to experience r = .29). The negative inclinations factor correlated inversely with the facets mainly associated with positive inclinations: Extraversion (facet: Dynamism; r = .37); Agreeableness (facet: Cooperativeness r = .18); Openness

(facet: Openness to experience r = .35). Only the negative inclinations factor correlated negatively with Emotional stability (facet: Emotion control r = .21). The positive inclinations factor also correlated with a preference for using maps as navigation aids (r = .18). As concerns navigation aids and traits/facets, map use was associated with Scrupulousness, a facet of Conscientiousness (r = .20), and Openness (r = .17, due to the involvement, however marginal, of both facets), whereas preferring to ask for verbal directions correlated with Cooperativeness, a facet of Agreeableness (r = .24). No significant correlations emerged between the selfreported variables and the spatial ability tasks.

Discussion and Conclusions This study examined how individual spatial competence related to people’s general disposition in terms of personality traits (and facets thereof). First, our CFA showed that the best model was represented by three factors in which spatial ability was distinguished by positive and negative inclinations (in line with Meneghetti et al., 2014). This distinction is also relevant because of the different role of these factors in

Figure 1. Three-factor model: positive and negative wayfinding inclinations and spatial ability factors. The estimated standardized coefficients are reported. Ó 2019 Hogrefe Publishing

Journal of Individual Differences (2020), 41(1), 45–52

C. Meneghetti et al., Personality Traits, Wayfinding Inclinations and Navigation

Table 3. Correlations between positive/negative wayfinding inclinations and spatial ability factors with personality trait/facet factors Positive wayfinding inclinations factor Extraversion

.29***

Negative wayfinding inclinations factor .29***

Spatial ability factor

Navigation aids: maps

Navigation aids: GPS

Navigation aids: verbal directions

.04

.08

.05

.03

Dynamism

.29***

.37***

.06

.10

.01

.12

Dominance

.21**

.15

.03

.09

.08

Agreeableness

.18**

.16

.07

.09

.01

.21**

Cooperativeness

.20**

.18**

.01

.10

.03

.24***

Politeness

.12

.11

.08

.02

.15

.08

.01

.05

.16

.02

Conscientiousness Scrupulousness

.02

.11

.04

.20**

.03

.01

Perseverance

.13

.11

.05

.08

.01

.03

Emotional stability

.09

.21**

.09

.01

.09

.02

Emotion control

.08

.21**

.05

.01

.06

.01

Impulse control

.07

.15

.10

.01

.09

.04

.31***

.24***

.09

.17**

.10

.15

.24***

.06

.13

.15

.04

.15

.14

.11

Openness Openness to culture

.29***

.35***

.01

Navigation aids: maps

Openness to experience

.18**

.05

.01

Navigation aids: GPS

.12

.16

.04

Navigation aids: verbal directions

.01

.12

Note. **p .01–.002; ***p .001.

relation to personality traits (and facets) and the use of navigation aids. Indeed, as for the relationship between wayfinding inclinations and personality traits/facets, individuals positively inclined to wayfinding (strong SoD, preference for survey mode, knowledge and use of cardinal points, preference for landmark/route mode, and pleasure in exploring new places) showed a propensity for social relations, that is, they were more expansive, assertive (Extraversion), collaborative with others (Agreeableness facet of Cooperativeness), and receptive to new experiences and cultures (Openness). Conversely, individuals negatively inclined to wayfinding (high spatial anxiety and a preference for navigating in well-known places) were more introverted (less expansive), and less cooperative and open to experience. Only negative wayfinding inclinations correlated with difficulty coping with anxiety and emotions (low Emotion control, a facet of Emotional stability). The distinction between positive and negative wayfinding inclinations therefore not only showed inverse relations with some personality traits (and facets) – Extraversion, Agreeableness and Openness – but also enabled the relation between negative inclinations and low Emotional stability to emerge (that a two-factor model would have overlooked). This latter result is at least partly consistent with studies showing a specific negative impact of neuroticism (or poor emotional stability) on environment learning accuracy (Walkowiak, Foulsham, & Eardley, 2015). Journal of Individual Differences (2020), 41(1), 45–52

Thus, being more sociable, dynamic, collaborative, and curious about the world seems to be related to more functional wayfinding inclinations (as suggested by previous studies that separately examined SoD and pleasure in exploring; Condon et al., 2015; Pazzaglia et al., 2018), while being more prone to emotional instability seems to be related to dysfunctional wayfinding inclinations. Conscientiousness showed no link with wayfinding inclinations in our study (although such a relation was found in Condon et al., 2015; Pazzaglia et al., 2018). Its role emerged, however, when preferred navigation aids were examined. As regards navigation aids, individuals who preferred to use maps seemed more inclined to precision and organization (Scrupulousness, a facet of Conscientiousness), and more open (Openness). Map use also related to positive wayfinding inclinations. These results indicate that a selfreported use of maps to navigate, associated with a refined mental map (Gardony et al., 2013), relates to more functional spatial inclinations (as in Münzer et al., 2016), and newly show that this may express a general propensity to be scrupulous and open-minded. Conversely, individuals who preferred to ask for directions tended to be sensitive toward others (Cooperativeness, a facet of Agreeableness): this is intriguing as the present study pointed to it having a role vis-a-vis wayfinding inclinations that contrasts with previous reports (Condon et al., 2015; Pazzaglia et al., 2018). Agreeableness might be related to spatial tasks involving a role for others, as when imagining taking Ó 2019 Hogrefe Publishing

C. Meneghetti et al., Personality traits, wayfinding inclinations and navigation

another person’s perspective (as suggested by Crescentini, Fabbro, & Urgesi, 2014), but further investigations are needed on this issue. Overall, our results offer insight for expanding our theoretical knowledge of multiple individual factors to consider in spatial cognition studies. Previous evidence had indeed suggested the need to take multiple individual spatial competence into account (Hegarty et al., 2006), and this study confirms the importance of distinguishing between different types of spatial factor, and how they relate to other individual factors, such as general thinking, and emotional and behavioral dispositions (expressed by traits). Spatial factors can, to some extent, be dispositions in spatially related contexts. This certainly increases the complexity of the matter, but offers a new key to interpreting the factors involved in defining individual profiles. Examining such relations can offer insight for future applications as well. One, among several possible, would be to obtain a deeper picture of individual profiles for some occupations and motor activities that require wayfinding or space-related problem solving. For instance, military professions (aviators, field-based, naval) have been considered in terms of their spatial and wayfinding skills (e.g., Barron & Rose, 2013), and personality traits (e.g., Bakker, Hetland, Olsen, & Espevik, 2019), but not at the same time, whereas considering both of these individual factors together could better clarify their individual profiles. It should be noted that, although actual wayfinding and environment performance relate to spatial self-reports (Hegarty et al., 2006), specific strategy use (Karimpur & Hamburger, 2016), and personality traits (Pazzaglia et al., 2018), our study only directly measured our respondents’ personal perceptions of their wayfinding and spatial habits (navigation aids). At the same time, the relationship between individual factors and spatial learning is not always confirmed (e.g., Nys, Gyselinck, Orriols, & Hickmann, 2014), and differences in spatial learning may depend on the spatial information encoding modality and the demands of the task (e.g., Hamburger & Röser, 2014). To better validate the role of individual (spatial and personality) differences in wayfinding ability, future research should also include actual wayfinding tasks, using several measures to assess performance (e.g., Walkowiak et al., 2015), and examining the variability of the relations within a given study population (e.g., Weisberg, Schinazi, Newcombe, Shipley, & Epstein, 2014). For now, our interpretation of the present results must be limited to individual spatial profiles, and inferences on wayfinding ability – though plausible – should be considered with caution, as further studies are needed. To conclude, this study offers a new key to interpreting how different individual variables – in terms of

Ó 2019 Hogrefe Publishing

self-perceived space-related inclinations and general personality traits – work together to generate different individual profiles.

References Barron, L. G., & Rose, M. R. (2013). Relative validity of distinct spatial abilities: An example with implications for diversity. International Journal of Selection and Assessment, 21, 400– 406. https://doi.org/10.1111/ijsa.12049 Bakker, A. B., Hetland, J., Olsen, O. K., & Espevik, R. (2019). Daily strengths use and employee well-being: The moderating role of personality. Journal of Occupational and Organizational Psychology, 92, 144–168. https://doi.org/10.1111/joop.12243 Bryant, K. J. (1982). Personality correlates of sense of direction and geographical orientation. Journal of Personality and Social Psychology, 43, 1318–1324. https://doi.org/10.1037/00223514.43.6.1318 Bryant, D. J. (1997). Representing space in language and perception. Mind & Language, 12, 236–264. https://doi.org/10.1111/ 1468-0017.00047 Caprara, G. V., Barbaranelli, C., Borgogni, L., & Perugini, M. (2008). Big Five Questionnaire. Florence, Italy: Organizzazioni Speciali. Crescentini, C., Fabbro, F., & Urgesi, C. (2014). Mental spatial transformations of objects and bodies: Different developmental trajectories in children from 7 to 11 years of age. Developmental Psychology, 50, 370–383. https://doi.org/10.1037/ a0033627 Condon, D. M., Wilt, J., Cohen, C. A., Revelle, W., Hegarty, M., & Uttal, D. H. (2015). Sense of direction: General factor saturation and associations with the Big-Five traits. Personality and Individual Differences, 86, 38–43. https://doi.org/10.1016/ j.paid.2015.05.023 De Beni, R., Meneghetti, C., Fiore, F., Gava, L., & Borella, E. (2014). Batteria VS. Abilità visuo-spaziali nell’arco di vita adulta [VS Battery. Visuo-spatial abilities in the adult life span]. Florence, Italy: Hogrefe. Gardony, A. L., Brunyé, T. T., Mahoney, C. R., & Taylor, H. A. (2013). How navigational aids impair spatial memory: Evidence for divided attention. Spatial Cognition and Computation, 13, 319– 350. https://doi.org/10.1080/13875868.2013.792821 Goldberg, L. R. (1990). An alternative “description of personality”: The Big-Five factor structure. Journal of Personality and Social Psychology, 59, 1216–1229. https://doi.org/10.1037/00223514.59.6.1216 Hamburger, K., & Röser, F. (2014). The role of landmark modality and familiarity in human wayfinding. Swiss Journal of Psychology, 73, 205–213. https://doi.org/10.1024/1421-0185/a000139 Hegarty, M., Montello, D. R., Richardson, A. E., Ishikawa, T., & Lovelace, T. (2006). Spatial abilities at different scales: Individual differences in aptitude-test performance and spatiallayout learning. Intelligence, 34, 151–176. https://doi.org/ 10.1016/j.intell.2005.09.005 Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55. https://doi. org/10.1080/10705519909540118 Karimpur, H., & Hamburger, K. (2016). Multimodal integration of spatial information: The influence of object-related factors and self-reported strategies. Frontiers of Psychology, 7, 1443. https://doi.org/10.3389/fpsyg.2016.01443

Journal of Individual Differences (2020), 41(1), 45–52

C. Meneghetti et al., Personality Traits, Wayfinding Inclinations and Navigation

Klatzky, R. L. (1998). Allocentric and egocentric spatial representations: Definitions, distinctions, and interconnections. In C. Freksa, C. Habel, & K. F. Wender (Eds.), Spatial cognition. An interdisciplinary approach to representing and processing spatial knowledge (pp. 1–17). Berlin, Germany: Springer. https://doi.org/10.1007/3-540-69342-4_1 Kozlowski, L. T., & Bryant, K. J. (1977). Sense of direction, spatial orientation, and cognitive maps. Journal of Experimental Psychology: Human Perception and Performance, 3, 590–598. https://doi.org/10.1037/0096-1523.3.4.590 Kozhevnikov, M., & Hegarty, M. (2001). A dissociation between object manipulation, spatial ability and spatial orientation ability. Memory & Cognition, 29, 745–756. https://doi.org/ 10.3758/BF03200477 Jöreskog, K. G., & Sörbom, D. (1981). LISERL V: Analysis of linear structural relationship by maximum likelihood and least squares methods. Chicago, IL: National Educational Resources. Lawton, C. A. (1994). Gender differences in way-finding strategies: Relationship to spatial ability and spatial anxiety. Sex Roles, 30, 765–779. https://doi.org/10.1007/BF01544230 Meneghetti, C., Borella, E., Pastore, M., & De Beni, R. (2014). The role of spatial abilities and self-assessments in cardinal point orientation across the lifespan. Learning and Individual Differences, 35, 113–121. https://doi.org/10.1016/j.lindif.2014.07. 006 Muffato, V., Toffalini, E., Meneghetti, C., Carbone, E., & De Beni, R. (2017). Individual visuo-spatial factors and familiar environment knowledge: A structural equation modeling analysis. Personality and Individual Differences, 113, 96–102. https://doi. org/10.1016/j.paid.2017.03.023 Münzer, S., Fehringer, B. C., & Kühl, T. (2016). Validation of a 3factor structure of spatial strategies and relations to possession and usage of navigational aids. Journal of Environmental Psychology, 47, 66–78. https://doi.org/10.1016/j.jenvp.2016.04. 017 Münzer, S., Zimmer, H. D., Schwalm, M., Baus, J., & Aslan, I. (2006). Computer-assisted navigation and the acquisition of route and survey knowledge. Journal of Environmental Psychology, 26, 300–308. https://doi.org/10.1016/j.jenvp.2006.08. 001 Nys, M., Gyselinck, V., Orriols, E., & Hickmann, M. (2014). Landmark and route knowledge in children’s spatial representation of a virtual environment. Frontiers in Psychology, 5, 1522. https://doi.org/10.3389/fpsyg.2014.01522 Pazzaglia, F., & Meneghetti, C. (2017). Acquiring spatial knowledge from different sources and perspectives: Abilities, strategies and representations. In J. M. Zacks & H. A. Taylor (Eds.), Representations in mind and world. Essays inspired by Barbara Tversky (pp. 120–134). New York, NY: Routledge. Pazzaglia, F., Meneghetti, C., & Ronconi, L. (2018). Tracing a route and finding a shortcut: The working memory, motivational, and personality factors involved. Frontiers in Human Neuroscience, 12, 225. https://doi.org/10.3389/fnhum.2018.00225

Journal of Individual Differences (2020), 41(1), 45–52

R Core Team. (2017). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from https://www.R-project.org Schwarz, G. E. (1978). Estimating the dimension of a model. Annals of Statistics, 6, 461–464. https://doi.org/10.1214/aos/ 1176344136 Steiger, J. H., & Lind, J. C. (1980, May). Statistically-based tests for the number of common factors. Paper presented at the Annual Spring Meeting of the Psychometric Society, Iowa City, IA. Tucker, L. R., & Lewis, C. (1973). A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 38, 1–10. https://doi.org/10.1007/BF02291170 Thurstone, L. L., & Thurstone, T. G. (1941). Factorial studies of intelligence. Psychometric Monographs, 2, 94. Vandenberg, S. G., & Kuse, A. R. (1978). Mental rotations, a group test of three-dimensional spatial visualization. Perceptual and Motor Skills, 47, 599–604. https://doi.org/10.2466/pms.1978. 47.2.599 Walkowiak, S., Foulsham, T., & Eardley, A. F. (2015). Individual differences and personality correlates of navigational performance in the virtual route learning task. Computers in Human Behavior, 45, 402–410. https://doi.org/10.1016/j.chb.2014.12. 041 Weisberg, S. M., Schinazi, V. R., Newcombe, N. S., Shipley, T. F., & Epstein, R. A. (2014). Variations in cognitive maps: Understanding individual differences in navigation. Journal of Experimental Psychology: Learning, Memory, and Cognition, 40, 669–682. https://doi.org/10.1037/a0035261 Wen, W., Ishikawa, T., & Sato, T. (2014). Instruction of verbal and spatial strategies for the learning about large-scale spaces. Learning and Individual Differences, 35, 15–21. https://doi.org/ 10.1016/j.lindif.2014.06.005

History Received December 4, 2018 Revision received May 26, 2019 Accepted June 3, 2019 Published online September 10, 2019 ORCID Chiara Meneghetti https://orcid.org/0000-0002-1838-7958 Chiara Meneghetti Department of General Psychology University of Padova via Venezia, 8 35135 Padova Italy chiara.meneghetti@unipd.it

Ó 2019 Hogrefe Publishing

Original Article

Bite the Stress Away? Nail Biting and Smoking Predict Maladaptive Stress Coping Strategies Magdalena Siegel , Eva-Maria Adlmann, Georg Gittler, and Jakob Pietschnig Department of Applied Psychology: Health, Development, Enhancement and Intervention, Faculty of Psychology, University of Vienna, Austria

Abstract: Psychological correlates of adult nail biting have received little empirical attention so far, despite its high prevalence and negative associations with physical and mental health. One possible correlate of nail biting is arousal modulation, which has also been linked to smoking (i.e., another oral behavior). Here, we link nail biting to an adaptive (i.e., Task-oriented) and two maladaptive stress-coping strategies (i.e., Emotion-oriented, Avoidance-oriented) as well as smoking while controlling for personality traits and socio-demographic characteristics. In all, 838 German-speaking adults (Mage = 32.02, SD = 13.48; 431 women) completed measures of stress coping (CISS-SF), the Big Five personality traits (Mini-IPIP), and indicated nail biting and smoking behavior. In three theory-guided, hierarchical linear regressions we predicted each stress coping strategy by nail biting and smoking while controlling for personality and socio-demographic characteristics. Oral behaviors had differential effects on maladaptive stress coping strategies: The interaction between nail biting and smoking predicted Emotionoriented coping, while smoking predicted Avoidance-oriented coping. Both behaviors were unrelated to adaptive, Task-oriented coping. In sum, our results show that nail biting and smoking are important predictors of maladaptive but not adaptive coping strategies, even when controlling for confounders. Keywords: nail biting, smoking, stress coping, Big Five

Nail biting is a common habit in the adult population that has received comparatively little attention in current psychological research. Prevalence estimates indicate that 10–30% of adults habitually engage in subclinical nail biting (Heaton & Mountford, 1992; Pacan, Reich, Grzesiak, & Szepietowski, 2014; Snyder & Friman, 2012). Typically, nail biting prevalence peaks in adolescence and young adulthood (25–60%; Houghton, Alexander, Bauer, & Woods, 2018; Woods, Miltenberger, & Flach, 1996) and falls below 10% in persons over 35 (Snyder & Friman, 2012). Nail biting has been shown to be negatively associated with both aspects of physical health and mental well-being. For instance, nail biting can lead to intraoral and nail abnormalities as well as bacterial infections (Halteh, Scher, & Lipner, 2017). Deformations of the nails or the inability to break the habit can cause shame in those affected (Pacan et al., 2014), leading to psychological distress (Houghton et al., 2018), and an impaired perceived quality of life (Pacan et al., 2014). Currently, there is no consensus about the etiology and functions of nail biting (Halteh et al., 2017). Nail biting urges have been linked to general arousal modulation for decades (e.g., Coleman & McCalley, 1948), but recent evidence proposes both overstimulation (e.g., stress) and

Ó 2019 Hogrefe Publishing

understimulation (e.g., boredom; Roberts, O’Connor, & Bélanger, 2013; Tanaka, Vitral, Tanaka, Guerrero, & Camargo, 2008; Williams, Rose, & Chisholm, 2007) as possible causes. These discrepancies may arise because motives for nail biting vary both between persons (e.g., overcoming stress vs. boredom), as well as within persons (increasing arousal when the person is understimulated vs. lowering arousal when the person is overstimulated; Williams et al., 2007). Therefore, it seems likely that other oral behaviors and psychological traits that are related to arousal modulation may also be related to nail biting. However, only little is known about psychological and behavioral correlates in healthy adults. Therefore, the aim of our study is to investigate associations of subclinical nail biting with other oral behaviors (i.e., smoking) as well as psychological mechanisms related to arousal modulation (i.e., stress coping), while controlling for confounding influences (i.e., personality, sex, age).

Smoking Besides nail biting, smoking is another common oral behavior in adults. About 20% of the global adult population and

Journal of Individual Differences (2020), 41(1), 53–60 https://doi.org/10.1027/1614-0001/a000303

30% of the European population are current smokers (World Health Organization, 2018). In Austria (30%) and Germany (24%), the adult smoking prevalence is similar, peaking in young adulthood at about 30–35% (Statistik Austria, 2015; Zeiher, Kuntz, & Lange, 2017). Smoking has been proposed as an adult transference of nail biting (Tanaka et al., 2008) and research into selfreported motives suggests further similarities, because both stress and boredom relief have been reported as common smoking motives (Fidler & West, 2009; McEwen, West, & McRobbie, 2008). Smoking appears to be associated with a higher preference for maladaptive coping strategies (McGee, Williams, Nada-Raja, & Olsson, 2013; Pietras, Witusik, Panek, Szemraj, & Górski, 2011) and less engagement in Task-oriented coping (McGee et al., 2013).

Stress Coping Stress coping refers to cognitive and behavioral responses to taxing external or internal demands (Endler & Parker, 1990). One well-established conceptualization (Endler & Parker, 1990) differentiates between adaptive Taskoriented and maladaptive Emotion-oriented and Avoidance-oriented responses to stressful situations (henceforth: stress coping strategies). Task-oriented coping refers to taking an active problem-solving approach that aims at removing or alleviating the stressor (Cohan, Jang, & Stein, 2006). Emotion-oriented coping means modulating the emotions associated with a stressor (e.g., ruminating, self-blaming, engaging in emotional responses; Cohan et al., 2006). Avoidance-oriented coping describes the avoidance of exposure to a stressor (e.g., engaging in another task; Endler & Parker, 1990). The use of different stress coping strategies changes across the lifespan: Young adulthood is marked by increases in Task-oriented coping and decreases in Emotion-oriented coping (Wingo, Baldessarini, & Windle, 2015). Generally, older people use less maladaptive strategies than younger people (Woodhead, Cronkite, Moos, & Timko, 2014) and are better able to align coping strategies with the stressor at hand (Aldwin, 2011).

M. Siegel et al., Nail Biting, Smoking, & Stress Coping

et al., 2006; Cosway, Endler, Sadler, & Deary, 2000; McWilliams, Cox, & Enns, 2003; Rafnsson, Smari, Windle, Mears, & Endler, 2006). Evidence on associations between Avoidance-oriented coping and personality is somewhat inconsistent. Findings suggest positive associations of facets of Avoidanceoriented coping with Extraversion, but both positive and negative associations with Agreeableness, Conscientiousness, and Openness (Cohan et al., 2006; Cosway et al., 2000; McWilliams et al., 2003). With regard to oral behaviors, smoking was found to be positively related to Extraversion, Neuroticism, and Openness, but negatively related to Agreeableness and Conscientiousness (Cheng & Furnham, 2016; Munafò, Zetteler, & Clark, 2007; Zvolensky, Taha, Bono, & Goodwin, 2015). To the best of our knowledge, there is currently no evidence on associations between nail biting and the Big Five personality traits. Similar to changes in stress coping, changes in personality traits occur across the life span (see McAdams & Olson, 2010, for a review). People exhibit higher levels of Conscientiousness and Agreeableness as they age, whereas their levels of Neuroticism decrease. Changes in Openness exhibit curvilinear trends and changes in Extraversion vary by subfacets (see Roberts, Walton, & Viechtbauer, 2006).

Study Aims In the present study, we investigate effects of two oral behaviors (i.e., nail biting and smoking) on Task-oriented, Emotion-oriented, and Avoidance-oriented coping strategies. Past evidence (e.g., Cheng & Furnham, 2016; Cohan et al., 2006; Roberts et al., 2006) suggests associations of both stress coping and smoking with personality traits and age. Accordingly, we controlled for possible moderating influences of these variables. We hypothesized that nail biting and smoking will predict all three stress coping strategies when controlling for confounding influences.

Methods Associations Between Stress Coping, Oral Behaviors, and the Big Five Personality Traits Task-oriented coping typically shows positive associations with Extraversion and Conscientiousness, but negative associations with Neuroticism. Conversely, Emotionoriented coping is strongly and positively associated with Neuroticism, but negatively with Extraversion (Cohan Journal of Individual Differences (2020), 41(1), 53–60

Participants The sample comprised 838 German-speaking participants (431 women) from Austria (85.3%) and Germany (9.5%, 5.2% other), ranging in age from 17 to 81 years (M = 32.02, SD = 13.48). In all, 54.06% had completed A-levels, 19.81% held a university degree, 10.38% had completed an apprenticeship, 7.28% had completed secondary schooling, 4.42% had completed compulsory schooling, and 4.06% reported another qualification. Out of 836 participants Ó 2019 Hogrefe Publishing

M. Siegel et al., Nail Biting, Smoking, & Stress Coping

(two were excluded due to missing data), 297 (35.53%) indicated that they were current smokers and 147 (17.58%) indicated they were habitual nail biters. Fifty-four participants (6.46%) reported engaging in both habits.

Measures Stress Coping As a measure of stress coping strategies, we used the 21item form of the Coping Inventory for Stressful Situations (CISS-SF; Cohan et al., 2006; Endler & Parker, 1999). Using the parallel blind technique (Behling & Law, 2000), the second and senior authors independently translated the CISS-SF from English to German and compared the resulting phrases. Discrepancies were resolved through discussion. The CISS-SF consists of three subscales assessing Task-oriented, Emotion-oriented, and Avoidanceoriented coping strategies with seven, seven, and six items respectively (one warming-up item is not used in scalecalculations; Cohan et al., 2006) on 5-point Likert-typed scales ranging from 1 = not at all to 5 = very much. The CISS-SF has been shown to possess good psychometric properties and internal consistencies (α range: Taskoriented coping .78–.87; Emotion-oriented coping .78–.87; Avoidance-oriented coping .61–.87; Cohan et al., 2006). Cronbach αs obtained in this study were similar, yielding α = .75 for Task-oriented, α = .77 for Emotion-oriented, and α = .77 for Avoidance-oriented coping. Big Five Personality Traits We used the Mini-International Personality Item Pool (Mini-IPIP; Donnellan, Oswald, Baird, & Lucas, 2006; German translation: Swami, Nader, Pietschnig, Stieger, Tran, & Voracek, 2012) to assess the Big Five personality traits. The 20-item Mini-IPIP consists of four items per subscale measuring Agreeableness, Conscientiousness, Extraversion, Neuroticism, and Openness on 5-point Likert-typed scales ranging from 1 = not at all to 5 = very much. The Mini-IPIP has been shown to possess acceptable internal consistencies, good test-retest-reliabilities, and convergent validity with other Big Five measures (Donnellan et al., 2006). Here, Cronbach αs were α = .74 for Agreeableness, α = .64 for Conscientiousness, α = .77 for Extraversion, α = .69 for Neuroticism, and α = .64 for Openness, which correspond well to estimates obtained from both the original and the German version (Donnellan et al., 2006; Swami et al., 2012).

Procedure Undergraduate data-collectors, who received course credit for data collection, recruited participants through personal contacts and word-of-mouth. Participation in the study was voluntary and confidentiality was ensured. Participants provided demographic details, including age, sex, nationality, and highest educational qualification. After completing the questionnaire, participants were thanked and verbally debriefed.

Statistical Analysis First, we examined sex differences in all variables included in our subsequent analyses. This was warranted to assess the necessity of including sex as a control variable in our regression models. Second, we conducted a confirmatory factor analysis (CFA) using maximum-likelihood estimation to test the factor structure of the German version of the CISS-SF. We specified a three factor solution (Task-oriented, Emotionoriented, Avoidance-oriented coping) with two three-item parcels reflecting subfacets of the Avoidance-oriented scale (i.e., treating oneself and social interaction, Cohan et al., 2006). Factor variances were set to one and factor correlations between the three factors were estimated freely. We assumed RMSEA and SRMR values .08 and CFI and TLI values .90 to reflect acceptable model fit (Brown, 2015). Third, we conducted three theory-guided, hierarchical linear regressions with Task-oriented, Emotion-oriented, and Avoidance-oriented stress coping as the respective outcome variables. In all three regression models, we entered the predictor variables in four consecutive blocks: Main effects of smoking and nail biting were added in block 1, followed by their interaction term in block 2 and the Big Five personality traits and the participants’ sex in block 3. We added age and education (smoking behavior has been shown to be differentiated regarding educational qualification, e.g., Zeiher et al., 2017) in our final block. Cases with missing data on any of the variables in the respective regression model were excluded from analyses. We assumed an alpha-level of .05 for all statistical significance tests.

Results Nail Biting and Smoking Status Participants indicated whether they currently habitually engaged in nail biting and smoking on two single-item measures with a yes-or-no response format.

Ó 2019 Hogrefe Publishing

Bivariate correlations and sex differences between all variables used in our models are reported in Supplementary Table S1 in Electronic Supplementary Material 1 (ESM 1).

Journal of Individual Differences (2020), 41(1), 53–60

Because sex differences emerged for a number of variables, we included sex in our regression model.

CFA of the CISS-SF Full model specification, parameter estimates, and the input covariance matrix are provided in ESMs 2 and 3. Our specified solution showed acceptable model fit (RMSEA = .049, SRMR = .051, CFI = .919, TLI = .907, w2(165) = 486.804, p < .001; all indicator loadings p < .001) and justified using the CISS-SF scales in subsequent analyses.

M. Siegel et al., Nail Biting, Smoking, & Stress Coping

significant in block 3 (Adj. R2 = .12). Including age and educational qualification in block 4 did not lead to improved model fit (ΔAdj. R2 < .01), thus indicating block 3 as the best-fitting model. Extraversion and Neuroticism showed significant positive and Openness significant negative associations with this coping style. Presently, we treated dichotomous variables as factors. When recalculating all our analyses with dichotomous variables treated as numerical values, we obtained virtually identical results, excepting non-significant effects of smoking for emotion-oriented coping in blocks 2 and 3. Because of this inconsistency and the trivial effect sizes when the predictor was significant, we do not interpret this effect as a meaningful predictor here, following a conservative approach.

Regression Models Numerical results for all final regression models are detailed in Table 1 (full models in Supplementary Table S5 in ESM 4). For Task-oriented coping, significant negative effects of nail biting and smoking were found in block 1 (Adj. R2 = .01). Adding the interaction in block 2 did not significantly improve our model. In block 3, personality traits and sex improved the model fit (Adj. R2 = .21). Openness, Conscientiousness, and Agreeableness were significantly positively and Neuroticism and sex (higher scores for men) were significantly negatively related to Taskoriented coping. Neither nail biting, smoking, nor their interaction showed any meaningful effects. Adding age and educational qualification in block 4 significantly improved model fit (Adj. R2 = .23), with age emerging as a positive significant predictor. For Emotion-oriented coping, nail biting (but not smoking) emerged as a significant predictor in block 1 (Adj. R2 = .01). Their interaction term (added in block 2) significantly improved our model (Adj. R2 = .02) and remained significant in block 3 (Adj. R2 = .35). Including age and educational qualification in block 4 did not improve model fit (ΔAdj. R2 < .01), thus indicating block 3 as the best-fitting model. The significant interaction between nail biting and smoking suggests that nail biting moderates associations between smoking status and Emotion-oriented coping: As can be seen in Figure 1, nail biting smokers reported the highest scores in Emotion-oriented coping, while non-nail biting smokers reported the lowest scores. In contrast, for non-smokers both groups exhibited virtually identical scores. Neuroticism and sex (higher scores for women) were significantly positively and Openness, Conscientiousness, and Extraversion significantly negatively related to Emotion-oriented coping. For Avoidance-oriented coping, smoking (but not nail biting) emerged as a significant predictor in block 1 (Adj. R2 = .02). Adding the interaction term in block 2 did not significantly explain additional variance. Smoking remained Journal of Individual Differences (2020), 41(1), 53–60

Discussion Task-Oriented Coping Correlational analyses showed small but significant negative associations between Task-oriented coping and both nail biting and smoking. Contrary to our hypotheses, these associations did not remain significant (in the case of nail biting) or only marginally significant (in the case of smoking) when controlling for personality and sociodemographic characteristics. Associations between Task-oriented coping, personality traits, sex, and age were broadly in line with the literature (e.g., Aldwin, 2011; Cohan et al., 2006; Cosway et al., 2000; McWilliams et al., 2003), excepting the nonsignificant association between Task-oriented coping and Extraversion in our final model. We attribute our contrasting findings to our differing analytical strategy: By simultaneously controlling for other personality traits, more complex relationships between our variables of interest emerged than in previous studies that used only bivariate associations (e.g., Cohan et al., 2006).

Emotion-Oriented Coping Correlational analyses indicated significant positive associations between Emotion-oriented coping and nail biting, but not smoking. When controlling for personality and sex, we found a significant interaction between nail biting and smoking, supporting our hypothesis that both oral behaviors are related to Emotion-oriented coping. Nail biting smokers showed the highest levels of Emotion-oriented coping, whereas non-nail biting smokers showed the lowest levels. In contrast, in non-smoking adults, nail biters and non-nail biters showed comparable levels of Emotionoriented coping. Ó 2019 Hogrefe Publishing

Ó 2019 Hogrefe Publishing 0.03 0.11 0.15 0.14 0.01

Extraversion

Agreeableness

Neuroticism

Men (0) vs. women (1)

Age 0.05

0.05

< 0.01

0.04

0.02

0.03

0.10

.01

.16

.12

.21

.13

.04

.23

.17

.04

.03

.07

.23/.02

.893

.741

< .001

.193

< .001

.334

.444

.052

< .001

.023

.016

.048

.017

.002

.056

.028

.001

< .001

.004

0.17

0.43

0.03

0.09

0.19

0.10

0.23

0.05

0.11

0.05

0.03

0.12

0.07

0.05

.12

.45

.03

.09

.20

.10

.08

.03

.07

.35/.33

< .001

.345

.003

< .001

.002

.049

.465

.034

F(6, 776) = 67.50***

F(9, 776) = 47.76***

.018

.217

.001

.012

.051

.012

.005

< .001

.003

ηp2

0.34

0.10

0.06

0.21

0.07

0.15

0.03

0.08

0.21

0.06

0.04

0.05

0.04

0.16

0.10

0.07

.19

.09

.05

.20

.06

.13

.01

.03

.11

.12/.10

< .001

.017

.206

< .001

.069

< .001

.851

.414

.002

F(6, 780) = 16.16***

F(9, 780) = 12.73***

Avoidance-oriented (n = 790)

.034

.007

.002

.036

.004

.015

< .001

.002

.015

ηp2

Notes. b = unstandardized regression coefficient; SE = standard error; β = standardized regression coefficient; ηp2 = partial eta squared. All VIFs < 1.86. aNumber of final block in the respective model. For Emotion-oriented and Avoidance-oriented coping, block 3 is presented (age and education did not significantly improve model fit). *p < .05; **p < .01; ***p < .001.

0.01

0.17

Conscientiousness

0.02

0.13

Openness

Below A-level (0) vs. University (1)

0.02

0.09

Smoking Nail Biting

Below A-level (0) vs. A-level (1)

0.03

0.05

Non-nail biters (0) vs. nail biters (1)

0.06

0.08

Non-smokers (0) vs. smokers (1)

0.04

F(3, 776) = 7.52***

ΔF

Adj. R2/ΔAdj. R2

F(12, 776) = 20.81***

F(df)

3/4a

ηp2

Emotion-oriented (n = 786)

Task-oriented (n = 789)

Predictor

Block

Table 1. Results of hierarchical multiple regressions predicting coping strategies from smoking status, nail biting, personality traits, sex, age, and education.

M. Siegel et al., Nail Biting, Smoking, & Stress Coping 57

Journal of Individual Differences (2020), 41(1), 53–60

M. Siegel et al., Nail Biting, Smoking, & Stress Coping

Figure 1. Predicted values of emotionoriented coping as a function of nail biting and smoking, both adjusted (left) and unadjusted (right) for covariates.

We suggest that nail biting may be a relatively benign habit in itself (see also Houghton et al., 2018), but may indicate psychological maladjustment in the presence of other (oral) behaviors. Previous research on nail biting in children found a similar pattern: Elevated levels of psychological maladjustment were only present in nail biters, who also engaged in other maladaptive behaviors (e.g., skin picking, hair pulling; Selles et al., 2015). Our study extends these findings to subclinical nail biting in healthy adults. Because the effect of this interaction explained only a small amount of variance in Emotion-oriented coping, we caution against over-interpreting this effect until further replication. Non-nail biting smokers exhibited the lowest levels of Emotion-oriented coping. We suggest that this group represents a different subgroup of smokers, who smoke occasionally and for social (but not for stress relieving) purposes (Rosa, Aloise-Young & Henry, 2014; Schane, Glantz, & Ling, 2009). Occasional smoking has been linked to better psychological adjustment and higher levels of selfesteem than both non-smoking and frequent smoking (O’Callaghan & Doyle, 2001; Shedler & Block, 1990). We suggest that these individuals display low levels of Emotionoriented coping, because their high self-esteem protects them from self-blame and other behaviors related to Emotion-oriented stress coping (Cohan et al., 2006). Future research can contribute to clarifying these speculations. Our findings regarding personality traits and sex were broadly consistent with previous literature. For instance, we found a strong link between Emotion-oriented coping and Neuroticism and higher engagement in Emotionoriented coping for women than for men (e.g., Cohan et al., 2006).

Avoidance-Oriented Coping Correlational analyses indicated significant and positive associations between Avoidance-oriented coping and Journal of Individual Differences (2020), 41(1), 53–60

smoking that remained significant after including personality traits and sex. These results support our hypothesis and conforms to previous findings indicating a relationship between smoking and maladaptive coping in general, and Avoidance-oriented coping in particular (Bricker, Schiff, & Comstock, 2011; McGee et al., 2013). However, our hypothesis that nail biting is associated with Avoidance-oriented coping was not substantiated. The positive associations between Avoidance-oriented coping with Extraversion and sex are consistent with previous literature (e.g., Cohan et al., 2006; McWilliams et al., 2003; Rafnsson et al., 2006). Moreover, we found positive associations between Neuroticism and Avoidance-oriented coping that have received partial empirical support so far (Cohan et al., 2006). The negative association between Openness and Avoidance-oriented coping in our study contrasts the small but positive or trivial associations that have been reported in previous studies (Cohan et al., 2006; Cosway et al., 2000; McWilliams et al., 2003). As with task-oriented coping, we attribute the discrepancies to our different analytical strategy using higher-order models.

Limitations First, we used single-item, dichotomous, self-report measures to assess oral behaviors. In the present study, this approach was appropriate because we did not measure latent constructs but behaviors. However, the dichotomous realization of our variables prevented us from collecting information on the extent of engagement in these behaviors (i.e., occasional vs. habitual smoking and nail biting). In addition, same-source bias may have influenced our results, as all our measures relied on participants’ self-reports. Therefore, social desirability may explain some of the associations between personality traits, oral behaviors, and socio-demographic characteristics. Future researchers may wish to distinguish between different functional smoking Ó 2019 Hogrefe Publishing

M. Siegel et al., Nail Biting, Smoking, & Stress Coping

and nail biting types (occasional vs. habitual nail biters and smokers) and include objective indicators (e.g., nail length) in their investigations. Second, we did not assess other body-focused repetitive behaviors linked to nail biting (e.g., skin picking; Roberts et al., 2013). Because smoking and nail biting showed different associations with stress coping strategies, it seems plausible that other behaviors may also yield different stress coping patterns. Third, we did not collect information about lifetime engagement in oral behaviors. This prevented us from distinguishing between participants who never engaged and those who had engaged in one or both of these behaviors in the past. Therefore, the role of group-specific (i.e., current vs. former smokers) (i) changes over time (e.g., reductions in perceived stress levels after cessation; Taylor, McNeill, Girling, Farley, Lindson-Hawley, & Aveyard, 2014) or (ii) differences in stress coping strategies remain to be clarified in future studies. Finally, participants were younger and showed higher educational levels than expectable in the general population, which limits the generalizability of our findings.

Conclusions In all, we showed substantial relationships between two oral behaviors (i.e., nail biting and smoking) and maladaptive stress coping strategies, even when controlling for personality and socio-demographic characteristics. Nail biting smokers reported higher preference for Emotion-oriented coping than non-smoking others. However, non-nail biting smokers reported less engagement in Emotion-oriented coping than non-smoking others. We argue that nail biting may only be predictive of psychological maladjustment in the presence of other maladaptive behaviors. Future research is necessary to investigate this idea further. We also replicated positive associations between smoking and Avoidance-oriented coping that remained stable when controlling for confounders. In all, our findings illustrate the necessity of investigating oral behaviors and their psychological correlates in adults.

Electronic Supplementary Material The electronic supplementary materials are available with the online version of the article at https://doi.org/ 10.1027/1614-0001/a000303 ESM 1. Table S1 shows descriptive statistics and sex differences. ESM 2. The file contains standardized parameter estimates of the CISS-SF (Table S2) and the full model specification of the CFA (Figure S3). Ó 2019 Hogrefe Publishing

ESM 3. The file contains the input covariance matrix S4 for the CFA. ESM 4. Table S5 shows full regression models.

References Aldwin, C. M. (2011). Stress and coping across the lifespan. In S. Folkman (Ed.), Oxford handbook of stress, health, and coping (pp. 15–34). New York, NY: Oxford University Press. Behling, O., & Law, K. S. (2000). Translating questionnaires and other research instruments: Problems and solutions. Thousand Oaks, CA: Sage Publications. Bricker, J. B., Schiff, L., & Comstock, B. A. (2011). Does avoidant coping influence young adults’ smoking? A ten-year longitudinal study. Nicotine & Tobacco Research, 13, 998–1002. https:// doi.org/10.1093/ntr/ntr074 Brown, T. (2015). Confirmatory factor analysis for applied research (2nd ed.). New York, NY: Guilford Press. Cheng, H., & Furnham, A. (2016). The Big-Five personality traits, maternal smoking during pregnancy, and educational qualifications as predictors of tobacco use in a nationally representative sample. PLoS One, 11, e0145552. https://doi.org/ 10.1371/journal.pone.0145552 Cohan, S. L., Jang, K. L., & Stein, M. B. (2006). Confirmatory factor analysis of a short form of the Coping Inventory for Stressful Situations. Journal of Clinical Psychology, 62, 273–283. https:// doi.org/10.1002/jclp.20211 Coleman, J. C., & McCalley, J. E. (1948). Nail-biting among college students. Journal of Abnormal Psychology, 43, 517–525. https://doi.org/10.1037/h0062313 Cosway, R., Endler, N. S., Sadler, A. J., & Deary, I. J. (2000). The Coping Inventory for Stressful Situations: Factorial structure and associations with personality traits and psychological health. Journal of Applied Biobehavioral Research, 5, 121–143. https://doi.org/10.1111/j.1751-9861.2000.tb00069.x Donnellan, M. B., Oswald, F. L., Baird, B. M., & Lucas, R. E. (2006). The Mini-IPIP scales: Tiny-yet-effective measures of the big five factors of personality. Psychological Assessment, 18, 192–203. https://doi.org/10.1037/1040-3590.18.2.192 Endler, N. S., & Parker, J. D. A. (1990). Multidimensional assessment of coping: A critical evaluation. Journal of Personality and Social Psychology, 58, 844–854. https://doi.org/10.1037/00223514.58.5.844 Endler, N. S., & Parker, J. D. A. (1999). Coping Inventory for Stressful Situations (CISS) manual (2nd ed.). Toronto, Canada: Multi-Health Systems. Fidler, J., & West, R. (2009). Self-perceived smoking motives and their correlates in a general population sample. Nicotine & Tobacco Research, 11, 1182–1188. https://doi.org/10.1093/ntr/ntp120 Halteh, P., Scher, R. K., & Lipner, S. R. (2017). Onychophagia: A nail-biting conundrum for physicians. Journal of Dermatological Treatment, 28, 166–172. https://doi.org/10.1080/09546634. 2016.1200711 Heaton, K. W., & Mountford, R. A. (1992). Nail-biting in the population and its relationship to irritable bowel syndrome. Journal of the Royal Society of Medicine, 85, 457. Houghton, D. C., Alexander, J. R., Bauer, C. C., & Woods, D. W. (2018). Body-focused repetitive behaviors: More prevalent than once thought? Psychiatry Research, 270, 389–393. https://doi. org/10.1016/j.psychres.2018.10.002 McAdams, D. P., & Olson, B. D. (2010). Personality development: Continuity and change over the life course. Annual Review of Psychology, 61, 517–542. https://doi.org/10.1146/annurev. psych.093008.100507

Journal of Individual Differences (2020), 41(1), 53–60

McEwen, A., West, R., & McRobbie, H. (2008). Motives for smoking and their correlates in clients attending stop smoking treatment services. Nicotine & Tobacco Research, 10, 843–850. https://doi.org/10.1080/14622200802027248 McGee, R., Williams, S., Nada-Raja, S., & Olsson, C. A. (2013). Tobacco smoking in adolescence predicts maladaptive coping styles in adulthood. Nicotine & Tobacco Research, 15, 1971– 1977. https://doi.org/10.1093/ntr/ntt081 McWilliams, L. A., Cox, B. J., & Enns, M. W. (2003). Use of the Coping Inventory for Stressful Situations in a clinically depressed sample: Factor structure, personality correlates, and prediction of distress. Journal of Clinical Psychology, 59, 423–437. https://doi.org/10.1002/jclp.10080 Munafò, M. R., Zetteler, J. I., & Clark, T. G. (2007). Personality and smoking status: A meta-analysis. Nicotine & Tobacco Research, 9, 405–413. https://doi.org/10.1080/14622200701188851 O’Callaghan, F., & Doyle, J. (2001). What is the role of impression management in adolescent cigarette smoking? Journal of Substance Abuse, 13, 459–470. https://doi.org/10.1016/ S0899-3289(01)00089-X Pacan, P., Reich, A., Grzesiak, M., & Szepietowski, J. C. (2014). Onychophagia is associated with impairment of quality of life. Acta Dermato-Venereologica, 94, 703–706. https://doi.org/ 10.2340/00015555-1817 Pietras, T., Witusik, A., Panek, M., Szemraj, J., & Górski, P. (2011). Anxiety, depression and methods of stress coping in patients with nicotine dependence syndrome. Medical Science Monitor, 17, CR272–CR276. https://doi.org/10.12659/MSM.881767 Rafnsson, F. D., Smari, J., Windle, M., Mears, S. A., & Endler, N. S. (2006). Factor structure and psychometric characteristics of the Icelandic version of the Coping Inventory for Stressful Situations (CISS). Personality and Individual Differences, 40, 1247–1258. https://doi.org/10.1016/j.paid.2005.11.011 Roberts, B. W., Walton, K. E., & Viechtbauer, W. (2006). Patterns of mean-level change in personality traits across the life course: A meta-analysis of longitudinal studies. Psychological Bulletin, 132, 1–25. https://doi.org/10.1037/0033-2909.132.1.1 Roberts, S., O’Connor, K., & Bélanger, C. (2013). Emotion regulation and other psychological models for body-focused repetitive behaviors. Clinical Psychology Review, 33, 745–762. https://doi.org/10.1016/j.cpr.2013.05.004 Rosa, J. D. R., Aloise-Young, P. A., & Henry, K. L. (2014). Using motives for smoking to distinguish between different college student smoker typologies. Psychology of Addictive Behaviors, 28, 1297–1304. https://doi.org/10.1037/a0038355 Schane, R. E., Glantz, S. A., & Ling, P. M. (2009). Nondaily and social smoking: An increasingly prevalent pattern. Archives of Internal Medicine, 169, 1742–1744. https://doi.org/10.1001/ archinternmed.2009.315 Selles, R. R., Nelson, R., Zepeda, R., Dane, B. F., Wu, M. S., Novoa, J. C., . . . Storch, E. A. (2015). Body focused repetitive behaviors among Salvadorian youth: Incidence and clinical correlates. Journal of Obsessive-Compulsive and Related Disorders, 5, 49– 54. https://doi.org/10.1016/j.jocrd.2015.01.008 Shedler, J., & Block, J. (1990). Adolescent drug use and psychological health: A longitudinal inquiry. American Psychologist, 45, 612–630. https://doi.org/10.1037/0003-066X.45.5.612 Snyder, G., & Friman, P. C. (2012). Habitual stereotypic movements. A descriptive analysis of four common types. In J. E. Grant, D. J. Stein, D. W. Woods, & N. J. Keuthen (Eds.), Trichotillomania, skin picking, and other body-focuses repetitive behaviors (pp. 43–64). Arlington, VA: American Psychiatric Publishing Inc. Statistik Austria. (2015). Österreichische Gesundheitsbefragung 2014. Hauptergebnisse des Austrian Health Interview Survey (ATHIS) und methodische Dokumentation [Austrian Health

Journal of Individual Differences (2020), 41(1), 53–60

M. Siegel et al., Nail Biting, Smoking, & Stress Coping

Survey 2014. Main results of the Austrian Health Interview Survey (ATHIS) and methodological documentation]. Vienna, Austria: Author. Swami, V., Nader, I. W., Pietschnig, J., Stieger, S., Tran, U. S., & Voracek, M. (2012). Personality and individual difference correlates of attitudes toward human rights and civil liberties. Personality and Individual Differences, 53, 443–447. https://doi. org/10.1016/j.paid.2012.04.015 Tanaka, O. M., Vitral, R. W. F., Tanaka, G. Y., Guerrero, A. P., & Camargo, E. S. (2008). Nailbiting, or onychophagia: A special habit. American Journal of Orthodontics and Dentofacial Orthopedics, 134, 305–308. https://doi.org/10.1016/j.ajodo.2006.06. 023 Taylor, G., McNeill, A., Girling, A., Farley, A., Lindson-Hawley, N., & Aveyard, P. (2014). Change in mental health after smoking cessation: Systematic review and meta-analysis. British Medical Journal, 348, g1151. https://doi.org/10.1136/bmj.g1151 Williams, T. I., Rose, R., & Chisholm, S. (2007). What is the function of nail biting: An analog assessment study. Behavior Research and Therapy, 45, 989–995. https://doi.org/10.1016/j.brat.2006. 07.013 Wingo, A. P., Baldessarini, R. J., & Windle, M. (2015). Coping styles: Longitudinal development from ages 17 to 33 and associations with psychiatric disorders. Psychiatry Research, 225, 299–304. https://doi.org/10.1016/j.psychres.2014.12.021 Woodhead, E. L., Cronkite, R. C., Moos, R. H., & Timko, C. (2014). Coping strategies predictive of adverse outcomes among community adults. Journal of Clinical Psychology, 70, 1183– 1195. https://doi.org/10.1002/jclp.21924 Woods, D. W., Miltenberger, R. G., & Flach, A. D. (1996). Habits, tics, and stuttering: Prevalence and relation to anxiety and somatic awareness. Behavior Modification, 20, 216–225. https://doi.org/10.1177/01454455960202005 World Health Organization. (2018). WHO global report on trends in prevalence of tobacco smoking 2000–2025 (2nd ed.). Geneva, Switzerland: Author. Zeiher, J., Kuntz, B., & Lange, C. (2017). Rauchen bei Erwachsenen in Deutschland [Smoking behavior in adults in Germany]. Journal of Health Monitoring, 2, 59–65. https://doi.org/ 10.17886/RKI-GBE-2017-030 Zvolensky, M. J., Taha, F., Bono, A., & Goodwin, R. D. (2015). Big five personality factors and cigarette smoking: A 10-year study among US adults. Journal of Psychiatric Research, 63, 91–96. https://doi.org/10.1016/j.jpsychires.2015.02.008

History Received November 21, 2017 Revision received May 14, 2019 Accepted June 6, 2019 Published online September 10, 2019 ORCID Magdalena Siegel https://orcid.org/0000-0002-3100-0236 Magdalena Siegel Department of Applied Psychology: Health, Development, Enhancement and Intervention Faculty of Psychology University of Vienna Liebiggasse 5 1010 Vienna Austria magdalena.siegel@univie.ac.at

Ó 2019 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 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/j/jid. 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 independent publications, with no limitations on the number

Ó 2020 Hogrefe Publishing

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/j/jid. August 2016

Journal of Individual Differences (2020), 41 (1)

The best ways to support the healthy development of children and adolescents and their families Kristin S. Mathiesen / Ann V. Sanson / Evalill B. Karevold (Editors)

Tracking Opportunities and Problems from Infancy to Adulthood 20 Years With the TOPP Study 2018, x + 272 pp. US $49.80 / € 39.95 ISBN 978-0-88937-543-7 Also available as eBook The unique longitudinal study “Tracking Opportunities and Problems (TOPP)” began following nearly 1,000 children and their families in Norway in 1993. Few studies have ever accumulated such extensive information from such a large number of families. Eight waves of data on many aspects of child and family life have been collected from children aged 18 months to 18 years. The TOPP Study has provided new knowledge about and insight into the precursors, developmental paths and predictors of both good adaptation and mental health problems of children, as well as into parenting and family relationships. The editors have collated the key findings in three parts. Part 1 addressesthe mental health and development of children and adolescents. Part 2 focuses on parents, looking at

www.hogrefe.com

individual parental and family-related factors, including parental couple relationships. Part 3 looks at methodological issues, including the sample, response rate and measurement and analytical approaches. Each chapter reviews the existing knowledge in these areas in relation to the TOPP findings and provides extensive reference lists for those who want to dig deeper. This unique book provides thoughtprovoking insights into the TOPP findings to help guide therapeutic practice, to suggest new avenues of research, to inform teaching, and to shape policy planning and preventive actions. It is thus an invaluable resource for all professionals, researchers, educators, policy makers, and students working with children and adolescents and their families.

Detailed therapeutic strategies for personality disorders “This is a very well-structured, informative, and readily accessible book that provides unique and valuable guidelines for therapists treating those with personality disorders, with a clarification- and schema-focused approach.” Elsa Ronningstam, PhD, Harvard Medical School, Harvard University, Cambridge, MA

Rainer Sachse

Personality Disorders A Clarification-Oriented Psychotherapy Treatment Model 2020, x / 254 pp. US $49.80 / € 39.95 ISBN 978-0-88937-552-9 Also available as eBook This practice-oriented guide presents a model of personality disorders (PDs) based on the latest research showing that “pure” PDs are due to relationship disturbances. The reader gains concise and clear information about the dual-action regulation model and the framework for clarification-oriented psychotherapy, which relates the relationship dysfunction to central relationship motives and games. Practical information is given on how to behave with clients and clear therapeutic strategies based on a five-phase model are outlined to help therapists manage interactional problems in therapy and to assist clients in achieving effective change.

www.hogrefe.com

The eight pure personality disorders (narcissistic, histrionic, dependent, avoidant, schizoid, passive-aggressive, obsessive-compulsive, and paranoid) are each explored in detail so the reader learns about the specific features of each disorder and the associated interactional motives, dysfunctional schemas, and relationship games and tests, as well as which therapeutic approaches are appropriate for a particular PD. As the development of a trusting therapeutic relationship is difficult with this client group, detailed strategies and tips are given throughout.

Applying psychoanalytical theory to projective methods: The French School “This compendium is a remarkable synthesis by leading figures of the French School of psychoanalytic projective methods in personality assessment. This skillfully edited and magnificently translated book provides the English-speaking world with access to the rich and vibrant tradition of the French School. I literally could not put this book down!” Howard D. Lerner, PhD, Assistant Clinical Professor of Psychology, Department of Psychiatry, University of Michigan Faculty, Michigan Psychoanalytic Institute, Ann Arbor, MI, USA

Benoît Verdon / Catherine Azoulay (Editors)

Psychoanalysis and Projective Methods in Personality Assessment The French School 2020, xiv / 214 pp. US $56.00 / € 44.95 ISBN 978-0-88937-557-4

This unique book synthesizes the work of leading thinkers of the French School of psychoanalytical projective methods in personality assessment, exploring its theories and methods and its clinical applications. Detailed case studies from different stages of lifeexamine the psychopathology of everyday life with its severe and disabling states of suffering. Contemporary advances in research and clinical work are presented, and the groundbreaking early work of Nina Rausch de Traubenberg, Vica Shentoub, and Rosine Debray are also critically reread and discussed.

New Clinical tools adapted for clinicians and researchers in the appendices include a useful schema to facilitate the interpretation of the Rorschach and TAT together, a list of latent solicitations for the TAT, and the current version of the TAT Scoring Grid. This book is essential reading for clinical psychologists, psychiatrists, psychotherapists, researchers, and students interested in applying psychoanalytical theory to projective methods.