Social Psychology 1/2018 by Hogrefe

Volume 49 / Number 1 / 2018

Social Psychology

Editor-in-Chief Kai Epstude Associate Editors Anna Baumert Marco Brambilla Adam Fetterman Ilka Gleibs Michael Häfner Hans J. IJzerman Ulrich Kühnen Toon Kuppens Ruth Mayo Kim Peters

Start using strengths today! “The GO-TO book for building character.” Martin E. P. Seligman, the founder of positive psychology

Ryan M. Niemiec

Character Strengths Interventions A Field Guide for Practitioners 2018, xx + 300 pp. US $59.00 / € 46.95 ISBN 978-0-88937-492-8 Also available as eBook This book is the epitome of positive psychology: it takes the “backbone” of positive psychology – character strengths – and builds a substantive bridge between the science and practice. Working with clients’ (and our own) character strengths boosts well-being, fosters resilience, improves relationships, and creates strong, supportive cultures in our practices, classrooms, and organizations. This unique guide brings together the vast experience of the author with the science and the practice of positive psychology in such a way that both new and experienced practitioners will benefit. New practitioners will learn about the core concepts of character and signature strengths and how to fine-tune their approach and troubleshoot. Experienced practitioners will deepen their knowledge about advanced topics such as strengths overuse and collisions, hot button issues, morality, and integrating strengths with savoring,

www.hogrefe.com

flow, and mindfulness. Hands-on practitioner tips throughout the book provide valuable hints on how to take a truly strengths-based approach. The 24 summary sheets spotlighting each of the universal character strengths are an indispensable resource for client sessions, succinctly summarizing the core features of and research on each strength. 70 evidence-based step-by-step activity handouts can be given to clients to help them develop character strengths awareness and use, increase resilience, set and meet goals, develop positive relationships, and find meaning and engagement in their daily lives. No matter what kind of practitioner you are, this one-of-a-kind field guide is a goldmine in science-based applications. You’ll be able to immediately bring the science of well-being into action!

Social Psychology

Volume 49/Number 1/2018

Editor-in-Chief

Kai Epstude, University of Groningen, Department of Psychology, Grote Kruisstraat 2/1, 9712 TS Groningen,The Netherlands, Tel. +31 50 363-7632, Fax + 31 50 363-4581, E-mail k.epstude@rug.nl

Editorial Ofﬁce

Wim Meerholz, University of Groningen, Department of Psychology, Grote Kruisstraat 2/1, 9712 TS Groningen,The Netherlands, Tel. +31 50 363-6393, Fax + 31 50 363-4581, E-mail SocialPsych.EditorialOfﬁce@gmail.com MichaelHäfner,UniversitätderKünsteBerlin,Germany Anna Baumert, Max-Planck Institute for Hans IJzerman, Université Grenoble Alpes, France Collective Goods, Germany Ulrich Kühnen, Jacobs University, Germany Marco Brambilla, University of Milano-Bicocca, Toon Kuppens, University of Groningen, Italy The Netherlands Adam Fetterman, University of Texas, USA Ruth Mayo, Hebrew University of Jerusalem, Israel Ilka Gleibs, London School of Economics, Kim Peters, University of Queensland, Australia United Kingdom

Associate Editors

Consulting Editors

Susanne Abele (Oxford, OH, USA) Andrea Abele-Brehm (Erlangen-Nürnberg, Germany) Anja Achtziger (Friedrichshafen, Germany) Herbert Bless (Mannheim, Germany) Gerd Bohner (Bielefeld, Germany) Oliver Christ (Hagen, Germany) Paul Conway (Tallahassee, FL, USA) Katja Corcoran (Graz, Austria) Olivier Corneille (Louvain, Belgium) Amanda Diekman (Oxford, OH, USA) Andrew Elliott (Rochester, NY, USA) Bertram Gawronski (Austin, TX, USA) Guido Gendolla (Geneva, Switzerland) Jessica Good (Davidson, NC, USA) Tobias Greitemeyer (Innsbruck, Austria) Bettina Hannover (Berlin, Germany) Nina Hansen (Groningen, The Netherlands) Nicole Harth (Jena, Germany) S. Alexander Haslam (Brisbane, Australia) Ying-Yi Hong (Hong Kong, ROC) Roland Imhoff (Mainz, Germany) Eva Jonas (Salzburg, Austria) Franciska Krings (Lausanne, Switzerland) Daniel Lakens (Eindhoven, The Netherlands)

Publisher

Hogrefe Publishing, Merkelstr. 3, 37085 Göttingen, Germany, Tel. +49 551 99950-0, Fax +49 551 99950-111, E-mail publishing@hogrefe.com North America: Hogrefe Publishing, 7 Bulﬁnch Place, Suite 202, Boston, MA 02114, USA, Tel. (866) 823-4726, Fax (617) 354-6875, E-mail publishing@hogrefe.com

Production

Regina Pinks-Freybott, Hogrefe Publishing, Merkelstr. 3, 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 99950-900, Fax +49 551 90050-998

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 1864-9335, ISSN-Print 1864-9335, ISSN-Online 2151-2590

Ó 2018 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 6 issues per annual volume. Social Psychology is the continuation of Zeitschrift für Sozialpsychologie (ISSN 0044-3514), the last annual volume of which (Volume 38) was published in 2007.

Subscription Prices

Calendar year subscriptions only. Rates for 2018: Institutions 374.00/US $478.00; Individuals 159.00/US $223.00 (all plus US $24.00/ 18.00 shipping & handling; Germany: 5.00).

Payment

Payment may be made by check, international money order, or credit card, to Hogrefe Publishing, Merkelstr. 3, D-37085 Göttingen, Germany. US and Canadian subscriptions can also be ordered from Hogrefe Publishing, 7 Bulﬁnch Place, Suite 202, Boston, MA, 02114, USA.

Thomas Morton (Exeter, UK) Corinne Moss-Racusin (Saratoga Springs, NY, USA) Roland Neumann (Trier, Germany) Esther Papies (Glasgow, UK) Susanne Quadﬂieg (Bristol, UK) Marc-André Reinhard (Kassel, Germany) Toni Schmader (Vancouver, BC, Canada) Manfred Schmitt (Landau, Germany) Thomas W. Schubert (Oslo, Norway) Beate Seibt (Oslo, Norway) Frank Siebler (Tromsø, Norway) Monika Sieverding (Heidelberg, Germany) Dagmar Stahlberg (Mannheim, Germany) Fritz Strack (Würzburg, Germany) Rolf van Dick (Frankfurt/Main, Germany) Harm Veling (Nijmegen, The Netherlands) Tobias Vogel (Mannheim, Germany) Ulrich Wagner (Marburg, Germany) Eva Walther (Trier, Germany) Michaela Wänke (Mannheim, Germany) Michael Wohl (Ottawa, ON, Canada) Bogdan Wojciszke (Warsaw, Poland) Vincent Yzerbyt (Louvain-la-Neuve, Belgium)

Electronic Full Text

The full text of Social Psychology is available online at http://econtent.hogrefe.com and in PsycARTICLES.

Abstracting Services

Abstracted/indexed in Current Contents/Social and Behavioral Sciences (CC/S&BS), Social Sciences Citation Index (SSCI), PsycINFO, PASCAL, PSYNDEX, ERIH, Scopus, and EM Care. Impact Factor (2016): 2.602

Social Psychology (2018), 49(1)

Ó 2018 Hogrefe Publishing

Contents Editorial

Moving Ahead Kai Epstude

Original Articles

Election Poster Persuasion: Attitude Formation in the Void Malte Schott and Jule Wolf

News and Announcements

Ó 2018 Hogrefe Publishing

Belonging Mediates Effects of Student-University Fit on Well-Being, Motivation, and Dropout Intention Michèle Suhlmann, Kai Sassenberg, Benjamin Nagengast, and Ulrich Trautwein

Taking Priming to Task: Variations in Stereotype Priming Effects Across Participant Task Katherine R. G. White, Rose H. Danek, David R. Herring, Jennifer H. Taylor, and Stephen L. Crites

Less Power, Greater Conﬂict: Low Power Increases the Experience of Conﬂict in Multiple Goal Settings Petra C. Schmid

Call for Papers: ‘‘Ego Depletion and Self-Control: Conceptual and Empirical Advances’’: A Special Issue of Social Psychology Guest Editors: Junhua Dang and Martin S. Hagger

Social Psychology (2018), 49(1)

Editorial Moving Ahead Kai Epstude Department of Psychology, University of Groningen, The Netherlands

When the former Zeitschrift für Sozialpsychologie became Social Psychology more than 10 years ago, the previous editors and the journal hoped for more international visibility of the journal. Those hopes became reality with a steadily increasing impact factor and a large number of submissions from all parts of the world. At the same time, Social Psychology has tried to incorporate the evolving scientific standards in our field into its editorial policy (e.g., Epstude, 2017; Unkelbach, 2016), and to be a forum for new developments in that regard. Continuing this endeavor, we have decided to further adjust our editorial policy when it comes to open data and materials. Those changes are outlined below. Social Psychology will remain a journal that is open for different perspectives on studying social behavior, as well as for insights from both fundamental as well as applied research projects.

Changes in Manuscript Submissions In the last few years, Social Psychology encouraged authors of published papers to share data and materials. Many authors already did that. However, it was not mandatory yet. We have decided to go one step further and will require authors to agree to make their data and materials available before papers are finally accepted. This can be done either via the journal website or via an existing open repository. Exceptions from this policy can only be made when ethical or legal concerns prevent authors form sharing this information. The submission guidelines have been adjusted accordingly.

Special Issue We selected a proposal for a special issue to be published in 2019. Junhua Dang (Lund University, Sweden) and

Ó 2018 Hogrefe Publishing

Martin S. Hagger (Curtin University, Australia) will guest edit a special issue on “Ego depletion and self-control.” In recent years, the existing empirical evidence in that area has come under scrutiny. Both guest editors have previously examined this topic using different types of approaches and with high levels of transparency (e.g., Dang, Liu, Liu, & Mao, 2017; Hagger et al., 2016). Therefore, we are delighted to have them guest editing the special issue. The call for papers can be found in the present issue of Social Psychology.

Changes in the Editorial Team Julia Becker, Christian Unkelbach, and Michaela Wänke have completed their terms as Associate Editors in 2017. I would like to express my gratitude for their many years of service for the journal. I would also like to welcome Anna Baumert (Max Planck Institute for Collective Goods, Germany), Marco Brambilla (University of Milano-Bicocca, Italy), and Kim Peters (University of Queensland, Australia) as new members of the editorial team. With an increasingly international editorial team, we hope to broaden the impact as well as the mission of the journal.

Online Presence of the Journal Social Psychology is also available on Twitter now: @SocPsy_journal. We will post regular updates once new articles are available online.

References Dang, J., Liu, Y., Liu, X., & Mao, L. (2017). The ego could be depleted, providing initial exertion is depleting. Social Psychology, 48, 242–245. https://doi.org/10.1027/1864-9335/a000308

Social Psychology (2018), 49(1), 1–2 https://doi.org/10.1027/1864-9335/a000336

Epstude, K. (2017). Towards a replicable and relevant social psychology. Social Psychology, 48, 1–2. https://doi.org/10.1027/ 1864-9335/a000303 Hagger, M. S., Chatzisarantis, N. L., Alberts, H., Anggono, C. O., Batailler, C., Birt, A. R., . . . Calvillo, D. P. (2016). A multilab preregistered replication of the ego-depletion effect. Perspectives on Psychological Science, 11, 546–573. https://doi.org/ 10.1177/1745691616652873 Unkelbach, C. (2016). Increasing replicability. Social Psychology, 47, 1–3. https://doi.org/10.1027/1864-9335/a000270

Social Psychology (2018), 49(1), 1–2

Editorial

Kai Epstude Department of Psychology University of Groningen Grote Kruisstraat 2/1 9712 TS Groningen The Netherlands k.epstude@rug.nl

Ó 2018 Hogrefe Publishing

Original Article

Election Poster Persuasion Attitude Formation in the Void Malte Schott and Jule Wolf Institute of Psychology, Heidelberg University, Germany

Abstract: We examined the effect of presenting unknown policy statements on German parties’ election posters. Study 1 showed that participants inferred the quality of a presented policy from knowledge about the respective political party. Study 2 showed that participants’ own political preferences influenced valence estimates: policy statements presented on campaign posters of liked political parties were rated significantly more positive than those presented on posters of disliked political parties. Study 3 replicated the findings of Study 2 with an additional measure of participants’ need for cognition. Need for cognition scores were unrelated to the valence transfer from political parties to policy evaluation. Study 4 replicated the findings of Studies 2 and 3 with an additional measure of participants’ voting intentions. Voting intentions were a significant predictor for valence transfer. Participants credited both their individually liked and disliked political parties for supporting the two unknown policies. However, the credit attributed to the liked party was significantly higher than to the disliked one. Study 5 replicated the findings of Studies 2, 3, and 4. Additionally, participants evaluated political clubs that were associated with the same policies previously presented on election posters. Here, a second-degree transfer emerged: from party valence to policy evaluation and from policy evaluation to club evaluation. Implications of the presented studies for policy communications and election campaigning are discussed. Keywords: political persuasion, party identification, attitude change, policy communication, election campaigning, election poster

Election posters are a major part of political campaigning in Germany. The two big German parties alone (ChristianDemocratic party [CDU] and Social-Democratic party [SPD]) placed about 17,000 election posters during the federal election campaign of 2013 (Geise & Leidecker, 2015). Election posters are subject of many case studies, both for German and European elections (Dezelan & Maksuti, 2012; Dumitrescu, 2010; Geise, 2010; Geise & Leidecker, 2015). However, with some notable exceptions (e.g., negative campaigning: Leidecker, 2010; spontaneous affective reaction: Geise & Brettschneider, 2010), election posters have rarely been used as stimulus materials for experimental psychological research. The present paper contributes to closing this gap. Dumitrescu (2010) concluded that minor political parties in France made more use of informative content on election posters than bigger parties who, in comparison, emphasized their political candidates more personally. Dezelan and Maksuti (2012) describe a similar relation of information use for Slovenian election posters. In Germany, this coherence is less pronounced (cf. Geise, 2010) and not stable over time. During the election campaign of 2005 the CDU generally did not, but the SPD did use informative content on election posters. Then, for the 2013 campaign both big parties refrained from

Ó 2018 Hogrefe Publishing

widely using informative content on election posters (Geise & Leidecker, 2015). Overall the information given on German election posters is typically rather vague. If policies are mentioned at all, buzzwords and categorical statements are often preferred over meaningful content. But why would election campaigners want to display policy content on election posters? Chances are that voters might appreciate informative content. The display of actual policies could signal a more transparent style of politics and more accountability than mere personality advertising. The display of policy content also plays a communicative role within political culture. An election campaign that communicates political issues helps citizens to understand and integrate relevant topics and strengthens political participation (Geise, 2010). Parties therefore play an important role in the structuring of public opinion (Slothuus & de Vreese, 2010). The display of new policy concepts on election posters stimulates the political debate and could potentially be used to gain public support for the respective issues. The presentation of policies may represent a claim for a political party on a political issue. This, for example, could be observed in the German federal election of 2013. Here, the party Die Linke demanded a minimum wage on their election posters. Later a minimum wage was voted into

Social Psychology (2018), 49(1), 3–15 https://doi.org/10.1027/1864-9335/a000323

law by parliament. Even though this would not have been possible without the government parties (CDU and SPD), since then this is communicated as a success by Die Linke. In view of these advantages the question arises, why campaigners refrain from depicting informative content on election posters. First of all, as Geise (2010) describes in detail, the usage of informative content has to be coherent with the contemporary situation and strategies of a political party. However, one general apprehension could be that many voters might be unable to understand and appreciate more detailed policy content on election posters. This is especially likely if common citizens are not well informed about actual legislative work, which is a plausible assumption after all. The present paper surveys the effects of presenting unknown policy statements on election posters. We thereby test a situation where voters are confronted with allegedly meaningful content that is hard to understand and examine their reactions. Predictions about this paradigm can be derived from several psychological models. Issue framing by political parties has been shown (Slothuus & de Vreese, 2010) by using newspaper vignettes. Similarly, Cohen (2003) has demonstrated that party affiliation had more effect on participants’ policy evaluations than the policies’ actual content. Smith, Ratliff, and Nosek (2012) demonstrated that party framing of political issues can lead to rapid formation of durable attitudes. Participants evaluated vignettes describing welfare policies. They were not only influenced by policy content, but also by the political party presenting a policy. Hawkins and Nosek (2012) demonstrated with newspaper vignettes, that even implicit party identity of independent voters (measured with an implicit-association test, IAT) could predict political judgment. Thus, it can be expected that content displayed on election posters will have a comparable effect: party preferences will influence the evaluation of policy statements. However, a situation where an unknown policy statement is presented on an election poster is yet to be tested. The Michigan model (Campbell, Converse, Miller, & Stokes, 1960) states that voters are influenced by political parties to accept new information. Citizens adopt their own political opinions via the orientation on partisan opinion leaders. The display of policies on election posters is therefore feasible for concept coining, policy claiming, and political persuasion, and to inform voters about the viewpoints of a party. Bartels (2002, p. 138) concluded that “. . . partisanship is not merely a running tally of political assessments, but a pervasive dynamic force shaping citizens’ perceptions of, and reactions to, the political world.” Goren (2005) demonstrated that party identification has a

Social Psychology (2018), 49(1), 3–15

M. Schott & J. Wolf, Election Poster Persuasion

more stable impact on political opinions than other political values. This supports the assumption that policies on election posters will be evaluated with respect to the presenting parties: irrespective of participants being instantly able to understand the policy statements. Predictions derived from Heider’s (1946) balance theory are also congruent with these assumptions. As many scholars have demonstrated (e.g., Granberg & Brent, 1974; Shaffer, 1981) participants’ attitudinal orientation toward a given entity (here: a political party), in combination with this entity’s attitude toward an object (here: a policy statement), will result in the participants adapting their own attitude toward the object. Therefore, we can assume that participants will react with (dis-)approval to policy statements that are supported by political parties they (dis-)like. Apart from the considered valence transfer, a transfer of quality is also subject of the present studies. Assumptions from the abovementioned theories are extendable to a transfer of quality. And other paradigms are in line with this prospect, too. The heuristic-systematic model (Chaiken & Maheswaran, 1994; Chen & Chaiken, 1999) describes the process of attitude formation as two distinct, nonexclusive processes. The presentation of a policy by a certain political party can be interpreted as an informative cue and guide policy evaluation. The pragmatic persuasion model (Waenke & Reutner, 2010) refers to participants’ languagepragmatic interpretation of persuasive information. Participants should infer a certain fit to partisan characteristics from the fact that a policy is presented on an election poster. Finally, (non-)evaluative conditioning procedures (Hofmann, De Houwer, Perugini, Baeyens, & Crombez, 2010; Förderer & Unkelbach, 2011) classically show how the paired presentation of items leads to a transfer of quality and valence from unconditioned to conditioned stimuli. The heuristic-systematic model, (non-)evaluative conditioning, and components of the pragmatic persuasion paradigm all concur in the prediction that qualities associated with a presenting political party should be transferred to the presented policy statements. The studies reported here use election posters as stimulus material. Participants are confronted with unknown policy statements presented on posters of political parties. We expect that participants will evaluate those policies in accordance with their attitudes toward the parties: quality and valence estimates associated with the presenting parties will be transferred to the participants’ evaluation of the presented policies. We expect participants to credit the presenting parties for their use of informative content. We also expect a second-degree transfer of valence. That is, the positive (negative) evaluation of policies will transfer further to another agent that is associated with

Ó 2018 Hogrefe Publishing

M. Schott & J. Wolf, Election Poster Persuasion

the respective policy. We thereby will demonstrate that the presentation of policies on election posters can be used to gain public support.

Table 1. Agreement to reasons for voting behavior Potential reason for voting behavior My parents vote for this party

1.66

0.97

My friends vote for this party

1.76

0.75

This party is going to win the election

1.79

0.84

Exploratory Study

This party represents my own opinion

3.34

0.74

This party is the lesser evil

2.42

0.95

Before experimental evidence for our hypothesis is presented, we include an exploratory study in support of our paradigm.

The candidates of this party are likeable

2.84

0.79

This party has a good reputation

2.45

0.76

This party pursues good politics

3.05

0.73

This party convinced me in the past

2.86

0.82

I hope to gain personal advantages from this partyâ&#x20AC;&#x2122;s politics This partyâ&#x20AC;&#x2122;s basic demands are important and right

2.16

1.03

3.71

0.56

Procedure Participants were students of a university course in political psychology who took part in a short questionnaire study for course credit. The study was conducted in April 2017. The participants were asked for voting intention, reasons for their voting behavior, and knowledge questions about the politics of their party. Participants also provided an actual-theoretical comparison concerning several components of election posters, like party slogans, candidate pictures, political arguments, demands, and achievements on a 5-point Likert scale (1 = is not used at all to 5 = is used all the time).

Results The dataset includes n = 38 participants (23 females, 13 males, and 2 other) between 18 and 54 years old (M = 22.24, SD = 5.53). The party vote resulted in the following distribution: CDU n = 2 (5.3%), Freie Demokratische Partei (FDP) n = 2 (5.3%), SPD n = 10 (26.3%), Die GrĂźnen n = 12 (31.6%), Die Linke n = 10 (26.3%), nonvoter n = 1 (2.6%), invalid votes n = 1 (2.6%). For the actual-theoretical comparison of election poster components, results showed that there was a significant difference between the actually perceived frequencies and the frequency participants recommended components to be used. Participants stated that party slogans, t(37) = 4.77, p < .001, d = 0.78, and candidate pictures, t(37) = 6.08, p < .001, d = 1.02, should be presented less, and political arguments, t(37) = 10.92, p < .001, d = 1.84, demands, t(37) = 9.65, p < .001, d = 1.59, and achievements, t(37) = 10.68, p < .001, d = 1.78, should be presented more often on election posters. The participantsâ&#x20AC;&#x2122; agreement to possible reasons for their voting behavior is depicted in Table 1. Ă&#x201C; 2018 Hogrefe Publishing

Notes. The actual question was â&#x20AC;&#x153;Please state if the stated reasons for voting decisions apply to you; I vote for my party because. . .â&#x20AC;? The provided answers range from 1 (= I do not agree) to 4 (= I strongly agree).

Asked to provide up to three of the basic demands of their parties, n = 21 (55.3%) participants named three, n = 6 (15.8%) participants named two, n = 4 (10.5%) participants named one, and n = 7 (18.4%) participants were unable to name any demand of the party they had voted for. Asked to provide up to three law proposals their parties had opposed in parliament, n = 3 (7.9%) participants named three, n = 3 (7.9%) participants named two, n = 3 (7.9%) participants named one, and n = 29 (76.3%) participants were unable to name any law proposal opposed by the party they had voted for.

Discussion The results demonstrate that even in highly educated university students, with an explicit interest in politics, the available knowledge about policy issues in terms of actual parliamentary work is minimal. Nevertheless, the participants agreed most to those reasons for their voting behavior that were associated with actual political content. Furthermore, participants stated that they would appreciate election poster designs that provided more political information. These results highlight the educational role election posters could have. It is reasonable to assume that political parties could profit from more informative campaigns with regard to public opinion and voting behavior.

Study 1 Study 1 examined how participants react to a political statement that is difficult to understand, and that is presented on an election campaign poster: does an assumed Social Psychology (2018), 49(1), 3â&#x20AC;&#x201C;15

M. Schott & J. Wolf, Election Poster Persuasion

feature of a political party transfer onto the evaluation of an unknown policy? In the German political culture there are not many stereotypes as prominent as the following: the Greens (Bündnis 90-Die Grünen) stand for environmentally responsible politics and the Liberals (Freie Demokratische Partei, FDP) usually advocate politics that are in the interest of the economy. This implies that Green voters are usually concerned with environmentalism and Liberal voters are commonly concerned with economic issues. These assumptions are useful to test participants’ reactions to unknown policy issues presented on election posters of known political parties. Since the German parliament has decided for an exit strategy from nuclear power generation, questions about the specifics of the exit have been discussed: environmental interests oppose the interests of energy-providers. In the year 2012, the German parliament voted in favor of a law regulating financial aspects of the extended time German nuclear reactors are allowed to produce electricity (Laufzeitverlängerung). But the law’s name1 does not clearly indicate its intention: is it in favor of an environmental approach or is it in the interest of the energy providing companies? We used the name of the law to test the hypothesis that participants would interpret an unknown statement in accordance with knowledge about the political party presenting the statement.

Material and Procedure Participants were recruited via the online recruiting site of the psychological department of the Heidelberg University (Bock, Nicklisch, & Baetge, 2012). The study was conducted in October 2015 and took approximately 3 min. It was part of a battery of other, unrelated experiments. Participants were welcomed, seated, and filled in informed-consent forms before the battery of experiments started. Participants were randomly assigned to one of the two experimental conditions (Liberal poster or Green poster). Debriefing was implemented in the experimental software.2 Payment was carried out at the end of the experimental session (8 € for 1 hr). The study procedure followed a between-subjects pre-post design. Participants were initially asked if they knew about the law. Then they estimated the degree to which the law was environmentally motivated or in favor of energy providing companies on a single dimension (scroll bar from 0 = economy friendly to 100 = environment

friendly). Subsequently, they were shown one of the two versions of the campaign poster (Liberal or Green, see Appendix, Figure A1). We used original campaign posters of the Liberal and the Green party but instead of the respective original statements we added one and the same statement on both campaign posters: “The only party in favor of the law for the skimming off of added profits from the extended nuclear reactor life time.” Finally, participants were asked again, in the light of the poster they had just seen, to indicate their estimate of the environment/ economy-orientation of the law.

Results Of the participants, n = 9 were excluded from further analysis since they knew about the law. This resulted in n = 101 participants (83 females) between 18 and 61 years old (M = 22.56, SD = 5.90). Across conditions, the participants’ dependent variable (DV) estimates were M = 36.52 (SD = 24.88) at the first measurement and M = 51.82 (SD = 32.15) at the second measurement. Descriptively, the participants who had seen the Liberal poster judged the proposal much more in favor of the energy-providers at the second measurement (M = 30.40, SD = 24.63) than those participants who had seen the Green poster (M = 70.46, SD = 25.75). A two-way repeated-measures analysis of variance (ANOVA) was conducted to analyze the effect of the manipulated campaign posters (condition) on the participants’ estimation of the law’s impact (measure). The results showed a significant main effect of the poster on the measure, F(1, 99) = 33.58, p < .001, η2 = 0.25. And there was a significant main effect of time on the measure, F(1, 99) = 18.64, p < .001, η2 = 0.16. Most importantly there was also a significant interaction effect of Time Poster, F(1, 99) = 30.70, p < .001, η2 = 0.24. Participants in the Green (Liberal) poster condition estimated the law to be more environmentally (economically) oriented at the post-manipulation measurement as compared to the pre-manipulation measurement.

Discussion All participants across conditions estimated the law proposal at the first time of measurement to be more in favor of the economy. The manipulation with the Green party poster led participants to significantly change their estimates resulting in the opinion that the law proposal

The law proposal about the “skimming off of added profits from the extended [nuclear reactor] life time” (German: Abschöpfung von Zusatzgewinnen aus der Laufzeitverlängerung) was part of a law proposal about the revision of the nuclear energy law introduced into parliament by the CDU/CSU/FDP coalition in 2010. The open-access software tool sosci survey (Leiner, 2014) was used for all of the presented studies.

Social Psychology (2018), 49(1), 3–15

Ó 2018 Hogrefe Publishing

M. Schott & J. Wolf, Election Poster Persuasion

was rather pro-environment. Participants in the Liberal condition corrected their estimates in the hypothesized direction, too. This change was not as strong as for the Green condition though. The main effects of time and condition, as well as the interaction affect, are mainly driven by the change in estimation of the participants in the Green poster condition. The main effect of time is due to the initial estimation across participants that the law was more in the interest of the economy. Comparable to a floor effect, participants who saw the Liberal poster thus had not as much room to correct their estimation in the economy-oriented direction as the other participants could correct their estimates into the environment-oriented direction. This pulled the average estimate at the second time of measurement across all participants more into the direction of the ecological (rather than the economical) interpretation. The average estimates of participants in the Green poster condition differed significantly from the estimates in the Liberal poster condition, reflecting the main effect of the poster condition. Importantly, the interaction effect of Time Poster supports our hypothesis, that participants were willing to adapt their estimation congruently to the presented campaign posters and the implied information at the post-manipulation measurement. The first study demonstrated the feasibility of quality transfer from party knowledge to a policy statement that is presented on an election poster. The expectancy that the Liberal party stands for economy friendly politics, and the Green party for environment friendly politics, respectively, transferred to the estimates about the political orientation of a formerly unknown law proposal. These results are consistent with the assumptions derived from the literature (e.g., Cohen, 2003; Slothuus & de Vreese, 2010; Shaffer, 1981). An additional explanation of the findings of Study 1 could be a demand effect. Nevertheless, this possibility does not change the rationale of our hypothesis: a situation where insufficient information about a target is available and participants therefore orient their estimates on knowledge about a political party is perfectly in line with the accounted psychological theories. For example regarding the pragmatic persuasion paradigm (Waenke & Reutner, 2010) and the balance theory (Granberg & Brent, 1974; Heider, 1946; Shaffer, 1981), the understanding of the policy statement in the context of a party’s persuasion attempt, and the subsequent adaption of an estimate, has an equivalent effect as the demand characteristic within the study design. It is reasonable to assume that voters 3

face comparable situations when they have to determine political questions formerly unknown to them. Therefore the idea of a demand effect is in this case supporting the hypothesis.

Study 2 The purpose of Study 2 was to examine if also the valence associated with a political party transfers to participants’ valence estimates of an unknown policy. For this purpose, we used political parties the participants explicitly liked or disliked.

Material and Procedure Participants were recruited via the online recruiting site of the psychological department of the Heidelberg University (Bock et al., 2012). The study was conducted in January 2016 and took approximately 6 min. It was part of a battery of other, unrelated experiments. Participants were welcomed, seated, and filled in informed-consent forms before the battery of experiments started. Debriefing was implemented in the experimental software. Payment was carried out at the end of the session (8 € for 1 hr). We used the five major German political parties3 for the study (B90-Grüne, CDU, FDP, SPD, and Die Linke) and presented two different policy statements on each party’s election poster (Appendix, Figure A2). The statements were pretested to be difficult to understand: (a) “Yes to the tax deductibility of employee reliefs.” And: (b) “Yes to a national quotation of the European Stability Mechanism.” At first participants had to rank order the parties, resulting in their most favored party situated at the top, and their most disliked party at the bottom (Appendix, Figure A3). Next the cover story was introduced, stating that the study concerned election posters displaying policy content instead of mere slogans. Participants were randomly assigned to one of the 2 2 balancing conditions (Order Pairing), resulting in their (least) favorite party’s campaign poster presented first or second, and paired with either one of the two statements (Appendix, Figure A2). After participants had seen both posters, they were asked if they had any prior knowledge about the respective law proposals4 and they answered two source-memory items (i.e., which political party advertised each of the two statements). Then participants were asked to indicate their

Note that the cover story for the Studies 2–5 related to the federal election of 2013 when four of the five abovementioned parties entered the parliament. Though the FDP did not enter the Bundestag (due to the German five-percent rule), it was still included in the study because it was present in the preceding parliament. Also, as the AFD had not been present in the preceding parliament, it was not included in the studies. This question’s purpose was to support the consistency of the cover story: as the statements and posters were designed for the study no participant could have prior knowledge about them. All but one participant indicated that they did not have any prior knowledge.

Ó 2018 Hogrefe Publishing

Social Psychology (2018), 49(1), 3–15

M. Schott & J. Wolf, Election Poster Persuasion

personal evaluation of the two law proposals on a scroll bar (1 = very negative to 100 = very positive).

Results All participants were excluded: (a) if they had not answered all questions (n = 1), (b) if they had stated their suspicion that the stimulus material was fake (n = 4), or (c) if they had correctly described the study’s purpose (n = 1). This resulted in a dataset of n = 65 participants (50 females) between 18 and 48 years old (M = 23.46, SD = 4.51). The two source-memory items were analyzed and most participants recalled both pairings correctly (n = 57, 87.7%). The participants’ topmost and lowest rankings of the political parties are depicted in Table 2. The central analysis for Study 2 compared the participants’ valence estimates for the two presented policies. For this purpose, the data was recoded into two dummy items: one for the policy statement that had been paired with a participant’s most favored party (positive) and one for the negatively paired policy. With these variables, a paired sample t-test was conducted to compare the participants’ opinions for the positive paired concept and the negative paired concept. There was a significant difference in the scores for the positive paired (M = 59.42, SD = 14.12) and the negative paired (M = 47.52, SD = 15.85) concept; t(64) = 4.43, p < .0001, d = 0.79.5 One further exploratory analysis concerned the general effectiveness of our manipulation. For this purpose, we calculated a difference score between the positively and the negatively paired statement for each participant. Participants who rated the valence of the positively paired statement not much different6 from the negatively paired statement were labeled “nonresponders.” Note that the difference cutoff of one tenth of the scale is arbitrary. We argue that participants who repeatedly try to answer by placing the scroll bar on the midpoint of the scale, for example because they want to indicate that they have no positive or negative opinion on the matter, would fall into this category. This evaluation showed that 24 of 65 participants (36.1%) had a difference score of 10 or less. Difference scores were significantly different from zero for the subgroup of responders, t(40) = 5.024, p < .0001, but not for the subgroup of nonresponders, t(23) = 0.511, p = .614.

Discussion Study 2 demonstrated that the valence transfer from parties to policies worked as predicted. The combined presentation 5

Table 2. Rank order distributions of political parties by participants in Study 2 Topmost rank (frequency)

Topmost rank (%)

Lowest rank (frequency)

Lowest rank (%)

CDU

15.4

18.5

FDP

3.1

43.1

Grüne

35.4

1.5

Linke

20.0

30.8

SPD

26.2

6.2

Note. The topmost ranking should represent the most favored political party.

of a (dis)liked political party with an unknown policy statement was utilized by participants to infer policy valence. Statements that had been paired with the campaign posters of favored political parties were more positively rated than those paired with the campaign posters of disliked political parties. However, about one third of participants did not meaningfully adapt their valence estimates in line with the manipulation. There are several possible explanations why this is the case. Some participants could simply have refrained from adapting their opinion in accordance with a statement they most probably could not understand. With regard to the pragmatic persuasion paradigm, Waenke and Reutner (2010) found that participants with high scores on skepticism (Obermiller & Spangenberg, 1998) were less willing to behave in accordance with an advertisement manipulation. On the other hand, high scores on the need for cognition scale (Haugtvedt, Petty, & Cacioppo, 1992; for a German version, see Waenke, Bohner, Fellhauer, & Schwarz, 1991) did not impair the effect of another advertisement manipulation (Waenke & Reutner, 2010). Arguably, as the skepticism scale predominantly targets the consumer domain, and the need for cognition concept more generally assesses the rational examination of content statements, it is an unresolved question if the same relationship occurs for the present experimental design. Here, participants with high scores on need for cognition could very well be the same as those who are less responsive to the manipulation. Also, we have insufficient information if necessary preconditions for nonresponders were met. We let participants order the political parties, but we did not explicitly ask for valence ratings concerning the parties. It could well be the case that nonresponders just did not care as much for the respective parties as other participants did. For those participants the experimental procedure would only have paired neutral parties with unknown policy statements, naturally resulting in neutral valence estimates.

If only those participants who had correct source memory for both concepts were included (n = 57) there still was a significant difference in the scores for the positive paired (M = 58.68, SD = 14.17) and the negative paired (M = 47.05, SD = 16.82) concept; t(56) = 3.91, p < .001, d = 0.75. Only 1/10 or less of the provided scroll-bar scale (0–100) difference between the positive paired and the negative paired statement.

Social Psychology (2018), 49(1), 3–15

Ó 2018 Hogrefe Publishing

M. Schott & J. Wolf, Election Poster Persuasion

Study 3 Study 3 was a direct replication of Study 2 with an additional measure of need for cognition (NFC; Waenke et al., 1991) in order to test if participants’ responsiveness to the manipulation can be explained by their NFC scores.

Material and Procedure The study used the same experimental software as Study 2 (see above) with an additional NFC measure at the end of the study, right before the debriefing. The rest of the procedure was identical to Study 2. Study 3 was conducted in April 2016.

Results All participants were excluded if (a) they had not answered all questions and/or if (b) they had stated their suspicion that the stimulus material was fake and/or if (c) they had guessed the study’s purpose correctly (n = 6). All participants who did not recall both statement-party pairings correctly were excluded (n = 4). This resulted in a dataset of n = 88 participants (71 females) between 18 and 62 years old (M = 22.40, SD = 5.03). The participants’ topmost and lowest rankings of the political parties are depicted in Table 3. The central analysis for Study 3 compared the participants’ valence estimates for the two law proposals. The data was recoded into two dummy items (see Study 2). With these variables, a paired sample t-test was conducted to compare the participants’ opinions for the positive and the negative paired policy. There was a significant difference in the valence scores for the negative paired (M = 46.90, SD = 16.03) and the positive paired statement (M = 58.70, SD = 16.70), t(87) = 5.13, p < .00001, d = 0.72. As in Study 2, participants were grouped into responders and nonresponders according to the difference score of the dependent measures: 42 out of 88 participants (47.73%) were classified as nonresponders. There were no differences between responders and nonresponders regarding formal education. The difference scores were significantly different from zero for the subgroup of nonresponders, t(41) = 2.395, p = .021. An independent sample t-test revealed that the participants’ need for cognition (NFC) scores did not differ significantly between the two subgroups of responders (M = 80.11, SD = 11.98) and nonresponders (M = 76.74, SD = 12.47), t(86) = 1.29, p = .200, d = 0.26. Similarly, a linear regression analysis to predict the difference scores for all participants based on their NFC scores did not reveal significant results. A regression equation was found, F(1, 86) = .533, p = .467, with an R2 of .006. Ó 2018 Hogrefe Publishing

Discussion The primary goal of Study 3 was to replicate the findings of Study 2. This could be achieved. Therefore, we regard the effect of the presented manipulation as a solid finding. The secondary goal of Study 3 was to examine why some participants were more responsive to the manipulation than others. Neither education levels nor NFC scores revealed significant insights into this question. Thus, why our paradigm’s manipulation works better for some participants than for others remains an open question.

Study 4 Study 4 was a replication of Studies 2 and 3 with a few additions. This should support the demonstration of the effect’s robustness and pursue the open question of participants’ responsiveness to the manipulation.

Material and Procedure The study used the same experimental software as Studies 2 and 3 (see above). In addition to the ranking task, participants had to evaluate the five political parties according to their voting intentions and general agreement. Also, one additional dependent variable was included: participants had to indicate if they credited the two parties for their support of the presented policies (scroll bar 0–100). The rest of the procedure was identical to Study 2. Study 4 was conducted in April and May 2017.

Results Participants were excluded if they were no German citizens and were thus unable to state party preferences (n = 4), or if they had guessed the study’s purpose correctly (n = 2). This resulted in a dataset of n = 70 participants (53 females) between 18 and 49 years old (M = 23.37, SD = 4.93). The participants’ topmost and lowest rankings of the political parties are depicted in Table 4. As before (Studies 2 and 3), the central analysis for Study 4 was a paired sample t-test to compare the participants’ valence estimates for the two policies. There was a significant difference in the valence scores for positive paired (M = 57.86, SD = 18.90) and negative paired policy statement (M = 50.10, SD = 17.51), t(69) = –2.84, p < .01, d = 0.34. Another paired sample t-test was conducted to compare how strongly the participants credited the respective political parties for supporting the presented policies. Participants credited their preferred party significantly Social Psychology (2018), 49(1), 3–15

M. Schott & J. Wolf, Election Poster Persuasion

Table 3. Rank order distributions of political parties by participants in Study 3 Topmost rank (frequency)

Topmost rank (%)

Lowest rank (frequency)

Lowest rank (%)

CDU

12.5

14.8

FDP

6.8

51.1

Grüne

54.5

2.3

Linke

14.8

SPD

11.4

Table 4. Rank order distributions of political parties by participants in Study 4 Topmost rank (frequency)

Topmost rank (%)

Lowest rank (frequency)

Lowest rank (%)

CDU

24.3

21.4

FDP

10.1

31.4

Grüne

32.9

5.7

31.8

Linke

11.4

40.0

0.0

SPD

21.4

1.4

Note. The topmost ranking should represent the most favored political party.

more for the supported policy (M = 55.60, SD = 20.49) than their least favored political party (M = 42.86, SD = 22.28), t(69) = 4.174, p < .0001, d = 0.50. In order to control for the necessary requirements of the manipulation the voting intentions of the participants were analyzed. By subtracting the voting intention of the participants’ least favored party from the voting intention of the most favored party, a dummy variable for the difference score of voting intentions was calculated (M = 67.87, SD = 28.09). As before (Studies 2 and 3), participants were grouped into responders and nonresponders according to the difference score of the dependent measures: 40 out of 70 participants (57.1%) were classified as nonresponders. The difference scores of the dependent measures were not significantly different from zero for the subgroup of nonresponders, t(39) = 1.025, p = .312. A linear regression analysis to predict the difference scores of the dependent measures based on the voting intentions difference scores revealed significant results. A regression equation was found, F(1, 68) = 4.751, p < .05, with an R2 of .065.

This is of interest regarding the applicability of the findings and speaks to a general appraisal of transparently communicating political work. The fact that even disliked political parties were credited for policies presented on election posters should be acknowledged as an important advantage.

Discussion The results of Study 4 demonstrate the robustness of the paradigm’s effects. However, the considerably high number of labeled nonresponders and the smaller effect size put emphasize on the question of participants’ responsiveness to the manipulation. With the difference between the stated voting intentions for the most and the least favored party one plausible answer to this question could be identified. For those participants who do not differ much in their pre-experimental ratings of the political parties, necessary requirements for a successful manipulation are barely met. The observed connection between voting intentions and the eventual effects is compatible with this assumption. The participants in Study 4 explicitly gave credit to the political parties for their support of the presented policies. While the party favored by participants was credited more strongly, the disliked party also was positively credited. Social Psychology (2018), 49(1), 3–15

Study 5 Study 5 was a replication of Studies 2, 3, and 4 with another added second part of the experiment. The second part of the experiment addressed the question if a second-degree transfer of valence can be demonstrated. That is, if a formerly unknown policy which is evaluated in accordance with a presenting party can transfer this valence even further, onto an unknown social club associated with the same policy.

Material and Procedure The study used the same experimental software as Study 4 (see above). Additionally, two different social clubs were presented to the participants. The clubs were described in general terms and the participants were told that each one of them supported one of the policies that had been presented on the election posters. The pairing of clubs with policies and their order of presentation was balanced across participants. The participants had to rate the clubs on two dimensions: if they could imagine participating in them and if they could imagine to support them with a small donation (1€) on a monthly basis (both scroll bars 0–100). The rest of the procedure was identical to Study 4. Study 5 was conducted in July 2017.

Results Participants were excluded if they guessed the study’s purpose correctly (n = 4) or if they did not answer all questions Ó 2018 Hogrefe Publishing

M. Schott & J. Wolf, Election Poster Persuasion

(n = 5). This resulted in a dataset of n = 88 participants (52 females) between 18 and 59 years old (M = 29.66, SD = 9.90). The participants’ topmost and lowest rankings of the political parties are depicted in Table 5. Again, the central analysis for Study 5 was a paired sample t-test to compare the participants’ valence estimates for the two presented policies. There was a significant difference in the valence scores for positive paired (M = 56.34, SD = 16.78) and negative paired policy statement (M = 49.00, SD = 14.56), t(87) = 3.20, p < .01, d = 0.34. Another paired sample t-test was conducted to compare how strongly the participants credited the respective political parties for supporting the policies presented on their election posters. Participants credited their preferred party significantly more for the supported policy (M = 55.18, SD = 18.73) than their disliked political party (M = 45.13, SD = 16.47), t(87) = 4.174, p < .001, d = 0.44. To control for the requirements of the experiment’s manipulation the voting intentions of the participants for two, respectively, used political parties were analyzed. By subtracting the voting intention of the participants’ least favored party from the voting intention of the most favored party a dummy variable for the difference score of their voting intention was calculated (M = 66.50, SD = 27.35). As before (Studies 2, 3, and 4), participants were grouped into responders and nonresponders: 38 out of 88 participants (43.2%) were classified as nonresponders. The difference scores of the dependent measures were not significantly different from zero for the subgroup of nonresponders, t(37) = 1.766, p = .086. A linear regression analysis to predict the difference scores of the dependent measures based on the voting intentions difference scores did not reveal significant results. A regression equation was found, F(1, 86) = 1.296, p = .258, with an R2 of .015. However, only n = 54 participants consistently rated their topmost rank ordered party best (i.e., with the highest policy agreement and the highest voting intention) and their lowest rank ordered party worst. If only those n = 54 participants were included in a similar linear regression analysis, the voting intentions difference scores were a significant predictor, F(1, 52) = 5.785, p = .020, with an R2 of .100. Finally, the participants’ club ratings were compared. Two sum scores were calculated with the clubs’ ratings for willingness to participate and willingness to donate. A paired sample t-test compared the ratings for the two clubs. The club that was associated with the positively paired policy (i.e., that had been presented by the favored political party) was rated significantly better (M = 31.70, SD = 21.73) than the club that was associated with the negatively paired policy (M = 26.30, SD = 20.01), t(87) = 2.610, p < .05, d = 0.28. Ó 2018 Hogrefe Publishing

Table 5. Rank order distributions of political parties by participants in Study 5 Topmost rank (frequency)

Topmost rank (%)

Lowest rank (frequency)

Lowest rank (%)

CDU

39.8

8.0

FDP

6.8

18.2

Grüne

19.3

8.0

Linke

5.7

64.8

SPD

28.4

1.1

Note. The topmost ranking should represent the most favored political party.

Discussion The results of Study 5 underline the robustness of the effects. The valence transfer from parties to policies reliably occurred in four consecutive studies. Also, under the condition that participants rank ordered the parties consistent with their actual party evaluations, the responsiveness to the manipulation was again predicted by differences in party evaluations. Notably, for those participants who did not consistently rank the parties, the paradigm’s preconditions are not strictly met. The fact that the basic valence transfer reaches a significant level even without excluding those participants emphasizes the strength of the effect. Most importantly, Study 5 demonstrated a second-degree transfer of valence. This effect is comparable to seconddegree conditioning effects (e.g., Walther, 2002; Walther, Nagengast, & Trasselli, 2005). The positive (or negative) valence policy statements acquired from the parties that presented them were further transferred to formerly unknown clubs that were associated with those policies. Yet our paradigm differs from classical evaluative conditioning procedures (De Houwer, Thomas, & Baeyens, 2001): the stimuli are not very briefly and repeatedly presented together with one another, but one time presented within the same object. Nevertheless, the second-degree transfer of valence demonstrated in Study 5 emphasizes the durability of the attitude formation process, animated by the election posters. And it supports the notion that election posters can be a potent medium influencing public debate and citizens’ opinion.

General Discussion We examined the effects of presenting policies on election posters on participants’ attitude formation. We tested hypotheses derived from several psychological paradigms: issue framing by political parties (e.g., Cohen, 2003; Slothuus & de Vreese, 2010), balance theory (Granberg & Brent, 1974; Social Psychology (2018), 49(1), 3–15

Heider, 1946; Shaffer, 1981), the pragmatic persuasion paradigm (Waenke & Reutner, 2010), (non-)evaluative conditioning (Förderer & Unkelbach, 2011; Hofmann, De Houwer, Perugini, Baeyens, & Crombez, 2010), the Michigan model (Campbell et al., 1960), and the heuristicsystematic model (Chaiken & Maheswaran, 1994; Chen & Chaiken, 1999). While others have already demonstrated issue framing by political parties using vignettes (e.g., Cohen, 2003; Hawkins & Nosek, 2012; Smith et al., 2012) we demonstrated the simple use of election posters: a medium commonly used by the thousands in German election campaigning. The design of election posters lies entirely in the hands of the political parties. Therefore the efficient usage of election posters is an important issue. Our results suggest that issue framing of unknown policies on election posters works nicely in the German political landscape and results in strong public support by partisan voters. Bishop, Tuchfarber, and Oldendick (1986) demonstrated participants’ willingness to express opinions on fictitious issues and interpreted their finding as caused by education, answer formats, and interviewer’s behavior. Yet we have to consider the numerous models that all allow for congruent predictions about the psychological processes relevant to our experimental design. Regarding these models we assume that the documented effects represent a more substantial attitude formation process, thus contrasting the findings of Bishop et al. (1986). This assumption is also due to the fact that we tested for any influence of education and the answer format in our studies consistently allowed for undecided answers (by placing the slider in the midpoint of the scale). Notably, many of the participants made use of this option. We descriptively categorized those participants as nonresponders: they arguably tried to give undecided estimates at the midpoint of the scale, possibly indicating the absence of an opinion. In Studies 4 and 5 we were able to show a connection between the responsiveness to the manipulation and the participants’ voting intentions, reflecting the manipulation’s necessary precondition. It remains unclear to which degree unknown political parties could profit from presenting policy statements on election posters. Arguably, unknown political parties would lack the preconditions for the demonstrated paradigm: a situation where an unknown political party would advertise unknown policies would resemble the case of those participants labeled nonresponders. If strong attitudes toward the parties do not exist and thus the necessary condition for the paradigm’s manipulation is not met no effect can be expected. Therefore, in accordance with balance theory, unknown political parties would indeed be required to present positively valent policy statements on election

Social Psychology (2018), 49(1), 3–15

M. Schott & J. Wolf, Election Poster Persuasion

posters. Interestingly, even well-established German parties often seem to exclusively present simple and irrefutable statements, rather than courageously presenting more finely nuanced policies. In contrast to such a campaign strategy where election posters only use categorical statements, or even refrain from presenting policy content altogether, our studies support a different strategy – at least for established political parties. Election posters should include nuanced policy statements and disregard that some voters may not be able to a priori understand them. Voters leaning to the presenting party will support the presented statements anyway – independent of their a priori understanding. Therefore, political parties should not refrain from detailed policy statements, for example, about already successful policy projects, or demands for future legislation. This presentation of policies will serve both an educational purpose and encourage public debate. More precisely, the expectable effects of presenting policies on election posters are promising for the respective parties, as issues presented on the election posters become associated with the parties. This association can be used to claim political efforts and subsequent successes in the political debate. And it also includes that partisans become more prone to engage in public debate situations, supporting both the concrete policies and thereby the political parties itself, because their attitude formation process is supported by the content presented on election posters. Furthermore, the results of our studies suggest that partisans strongly credit parties for their political work, presented on election posters. Voters who are more involved in the political debate are likely to influence others, and are due to consistency effects also more likely to actually vote for their party on electionday. This represents a strong support for strategic mobilization. Based on our studies it is very advisable for established political parties to more courageously present policy issues on election posters. Voters who oppose a presenting party will probably not be persuaded to support a policy presented by a political adversary. Yet, they nevertheless, to some degree, will credit a party for actual political work as the results of Studies 4 and 5 suggest. This effect deserves more empirical testing in the future. If it turns out in conceptual replications to be as robust as the basic finding of our studies it could support strategic demobilization efforts. Taken together the mobilization of partisans and the demobilization of adversary voters is an undeniably strong campaign prospect. However, the positive effect on partisan voters already proved to be a robust finding within our studies. And this prospect alone warrants the recommendation for political parties to present more policy issues on their election posters.

Ó 2018 Hogrefe Publishing

M. Schott & J. Wolf, Election Poster Persuasion

References Bartels, L. M. (2002). Beyond the running tally: Partisan bias in political perceptions. Political Behavior, 24, 117–150. https:// doi.org/10.1023/A:1021226224601 Bishop, G. F., Tuchfarber, A. J., & Oldendick, R. W. (1986). Opinions on fictitious issues: The pressure to answer survey questions. Public Opinion Quarterly, 50, 240–250. Bock, O., Baetge, I., & Nicklisch, A. (2012). hroot: Hamburg registration and organization online tool. European Economic Review, 71, 117–120. Campbell, A., Converse, P. E., Miller, W. E., & Stokes, D. E. (1960). The American voter. New York, NY: Wiley. Chaiken, S., & Maheswaran, D. (1994). Heuristic processing can bias systematic processing: Effects of source credibility, argument ambiguity, and task importance on attitude judgment. Journal of Personality and Social Psychology, 66, 460–473. https://doi.org/10.1037/0022-3514.66.3.460 Chen, S., & Chaiken, S. (1999). The heuristic-systematic model in its broader context. In S. Chaiken & Y. Trope (Eds.), Dual-process theories in social psychology (pp. 73–96). New York, NY: Guilford Press. https://doi.org/10.1207/S15327957PSPR0402_01 Cohen, G. L. (2003). Party over policy: The dominating impact of group influence on political beliefs. Journal of Personality and Social Psychology, 85, 808. De Houwer, J., Thomas, S., & Baeyens, F. (2001). Association learning of likes and dislikes: A review of 25 years of research on human evaluative conditioning. Psychological Bulletin, 127, 853–869. https://doi.org/10.1037/0033-2909.127.6.853 Dezelan, T., & Maksuti, A. (2012). Slovenian election posters as a medium of political communication: An informative or persuasive campaign tool? Communication, Politics & Culture, 45, 140–159. Dumitrescu, D. (2010). Know me, love me, fear me: The anatomy of candidate poster designs in the 2007 French legislative elections. Political Communication, 27, 20–43. https://doi.org/ 10.1080/10584600903297117 Förderer, S., & Unkelbach, C. (2011). Beyond evaluative conditioning! Evidence for transfer of non-evaluative attributes. Social Psychological and Personality Science, 2, 479–486. https://doi.org/10.1177/1948550611398413 Geise, S. (2010). Unser Land kann mehr. . . [Our nation can do more ...]. Zeitschrift für Politikberatung, 3, 151–175. https://doi. org/10.1007/s12392-010-0251-y Geise, S., & Brettschneider, F. (2010). Die Wahrnehmung und Bewertung von Wahlplakaten: Ergebnisse einer EyetrackingStudie [The perception and evaluation of election posters: Results of an eye-tracking study]. In T. Faas, K. Arzheimer, & S. Roßteutscher (Eds.), Information – Wahrnehmung – Emotion (pp. 71–95). Wiesbaden, Germany: VS Verlag für Sozialwissenschaften. https://doi.org/10.1007/978-3-531-92336-9_5 Geise, S., & Leidecker, M. (2015). Visueller Wahlkampf: Strategien in der Plakatkommunikation zur Bundestagswahl 2013 [Visual election campaign: Strategies for poster communication in the 2013 federal election]. ZPB Zeitschrift für Politikberatung, 7, 14–27. https://doi.org/10.5771/1865-4789-2015-1-2-14 Goren, P. (2005). Party identification and core political values. American Journal of Political Science, 49, 882–897. Granberg, D., & Brent, E. E. (1974). Dove-hawk placements in the 1968 election: Application of social judgment and balance theories. Journal of Personality and Social Psychology, 29, 687–695. https://doi.org/10.1037/h0036631 Haugtvedt, C. P., Petty, R. E., & Cacioppo, J. T. (1992). Need for cognition and advertising: Understanding the role of personality

Ó 2018 Hogrefe Publishing

variables in consumer behavior. Journal of Consumer Psychology, 1, 239–260. Hawkins, C. B., & Nosek, B. A. (2012). Motivated independence? Implicit party identity predicts political judgments among self-proclaimed independents. Personality and Social Psychology Bulletin, 38, 1437–1452. https://doi.org/10.1177/ 0146167212452313 Heider, F. (1946). Attitudes and cognitive organization. The Journal of Psychology, 21, 107–112. Hofmann, W., De Houwer, J., Perugini, M., Baeyens, F., & Crombez, G. (2010). Evaluative conditioning in humans: A meta-analysis. Psychological Bulletin, 136, 390–421. https://doi.org/10.1037/a0018916 Leidecker, M. (2010). Angreifende Plakatwerbung im Wahlkampf – effektiv oder riskant? Ein Experiment aus Anlass der SPDEuropawahlplakate 2009 [Negative campaign election posters – effective or risky?]. In C. Holtz-Bacha (Ed.), Die Massenmedien im Wahlkampf (pp. 117–139). Wiesbaden, Germany: VS Verlag für Sozialwissenschaften. https://doi.org/10.1007/9783-531-92509-7 Leiner, D. J. (2014). SoSci survey (Version 2.5.00-i). [Computer software]. Retrieved from http://www.soscisurvey.com Obermiller, C., & Spangenberg, E. R. (1998). Development of a scale to measure consumer skepticism toward advertising. Journal of Consumer Psychology, 7, 159–186. Shaffer, S. D. (1981). Balance theory and political cognitions. American Politics Quarterly, 9, 291–320. Slothuus, R., & de Vreese, C. H. (2010). Political parties, motivated reasoning, and issue framing effects. The Journal of Politics, 72, 630–645. Smith, C. T., Ratliff, K. A., & Nosek, B. A. (2012). Rapid assimilation: Automatically integrating new information with existing beliefs. Social Cognition, 30, 199–219. https://doi.org/10.1521/soco. 2012.30.2.199 Waenke, M., Bohner, G., Fellhauer, R. F., & Schwarz, N. (1991). Need for Cognition: Eine Skala zur Erfassung von Engagement und Freude bei Denkaufgaben [Need for cognition: A scale to measure commitment and fun with brainteasers]. Mannheim, Germany: ZUMA. Waenke, M., & Reutner, L. (2010). Pragmatic persuasion or the persuasion paradox. In J. Forgas, W. Crano, & J. Cooper (Eds.), The psychology of attitudes and attitude change (pp. 183–198). New York, NY: Psychology Press. Walther, E. (2002). Guilty by mere association: Evaluative conditioning and the spreading attitude effect. Journal of Personality and Social Psychology, 82, 919–934. https://doi.org/10.1037/ 0022-3514.82.6.919 Walther, E., Nagengast, B., & Trasselli, C. (2005). Evaluative conditioning in social psychology: Facts and speculations. Cognition and Emotion, 19, 175–196. https://doi.org/10.1080/ 02699930441000274 Received November 18, 2016 Revision received August 23, 2017 Accepted August 23, 2017 Published online February 7, 2018 Malte Schott Institute of Psychology Heidelberg University Hauptstraße 47–51 69117 Heidelberg Germany malte.schott@psychologie.uni-heidelberg.de

Social Psychology (2018), 49(1), 3–15

M. Schott & J. Wolf, Election Poster Persuasion

Appendix Study Material Figure A1. Stimulus material for Study 1; Liberal party campaign poster on the left; Green party campaign poster on the right. Note that the policy statements are identical on both posters: “Die einzige Partei für das Gesetz zur Abschöpfung von Zusatzgewinnen aus der AKW Laufzeitverlängerung!” Each participant of Study 1 saw only one of the above-pictured campaign posters in a between-subjects prepost design.

Figure A2. Stimulus materials for Studies 2–5. From left to right: Green party (Bündnis-90-Die Grünen), Christian-Democratic party (CDU), Liberal party (FDP), Left party (Die Linke), and Social-Democratic party (SPD) election posters. The upper one always has policy statement A (“Ja zur steuerlichen Abschöpfung der Arbeitnehmerentlastungen.”) on it and the lower one always policy statement B (“Für eine nationale Quotierung des Europäischen Stabilitätsmechanismus”). The basic format of the posters was standardized. The original party logos (top right corner), trademark colors, and trademark slogans (lower right corner) of the respective political parties were added. The statements were placed in the middle of the posters with constant distance to the left and right frame and the party’s logo in the upper right corner. Each participant of the Studies 2–5 saw one poster from the upper row and one poster (of a different party) from the lower row, as was determined by party preference choices (the most favorite and most disliked party of each participant, respectively) and the 2 2 balancing conditions (Order Pairing; favored party’s poster presented first or second, and paired with statement A or B).

Social Psychology (2018), 49(1), 3–15

Ó 2018 Hogrefe Publishing

M. Schott & J. Wolf, Election Poster Persuasion

Figure A3. Party preferences Item within Studies 2–5: participants’ preference and antipathy for the five relevant political parties. The instructions were: “Please order the presented political parties according to your personal opinion. The uppermost rank should be for the party whose positions you usually agree with the most; the lowest rank should be for the party whose positions you usually agree with the least.” Participants could drag-and-drop the party icons to the rank order positions. Rank 1 was used to determine the participants’ most favored party, and rank 5 was used to determine the participants’ most disliked party. This resulted in the choice of campaign posters, participants saw during the remainder of the study.

Ó 2018 Hogrefe Publishing

Social Psychology (2018), 49(1), 3–15

Original Article

Belonging Mediates Effects of Student-University Fit on Well-Being, Motivation, and Dropout Intention Michèle Suhlmann,1,2 Kai Sassenberg,1,2 Benjamin Nagengast,3 and Ulrich Trautwein3 1

LEAD Graduate School and Research Network, University of Tübingen, Germany

Leibniz-Institut für Wissensmedien (IWM), Tübingen, Germany

Hector Research Institute of Education Sciences and Psychology, University of Tübingen, Germany

Abstract: About one third of university students drop out from their undergraduate studies. The fit between students’ self-construal and university norms has been suggested to contribute to academic success. Building on this idea, we tested a student-university fit model in a cross-sectional online study among 367 German university students. Results support a P-E fit effect, showing that students with a high dignity self-construal and who perceived the university norms to be highly independent indicated the greatest sense of belonging to the university. In turn, belonging positively predicted well-being and academic motivation and reduced dropout intention. In sum, this study suggests that a person-environment fit analysis can contribute to the understanding of healthy student life and academic success. Keywords: person-environment fit, dignity self-construal, sense of belonging to the university, well-being, academic success

The student body of universities is as large and diverse as ever (OECD, 2016). Statistics reveal, for instance, a record increase of nearly 30% of students at Germany’s higher educational institutions during the last 10 years (Statistisches Bundesamt, 2012). A similar trend is visible for other universities worldwide, for example, in the US, UK, and Australia (Altbach, Reisberg, & Rumbley, 2009). This is good news not only because more students get access to higher education, but also because it implies that students come from more diverse backgrounds and most importantly bring diverse values to the university institution. However, not all students are doing well and are successful at university. According to a recent survey among German university students, every fifth student receives a diagnosis of mental illness (Grobe & Steinmann, 2015). Likewise alarming, results indicate that antidepressant intake among students rose about 43% in less than a decade (Grobe & Steinmann, 2015). Next to impaired wellbeing, students’ motivation and dropout are a rising problem (e.g., Heublein, Richter, Schmelzer, & Sommer, 2012). The problem of university students’ mental health issues and dropout has been recognized in many countries around the world (Mistler, Reetz, Krylowicz, & Barr, 2012). Given that academic success is an asset in the job market (European Centre for the Development of Vocational Training, 2013; Fabian, Rehn, Brandt, & Briedis, 2013), these Social Psychology (2018), 49(1), 16–28 https://doi.org/10.1027/1864-9335/a000325

problems deserve our attention. Therefore, the aim of the present study was to contribute to the understanding of the preconditions of a healthy student life and academic success. To this end, we employed the person-environment fit (P-E fit) framework (Caplan & Harrison, 1993; Edwards, Caplan, & Harrison, 1998). Specifically, we assumed that the fit between individuals’ self-construal and (perceived) university norms affects students’ sense of belonging to the university. Furthermore, we argue that this sense of belonging to the university is a key variable in promoting students’ well-being and motivation and preventing them from forming an intention to drop out.

Person-Environment Fit and the University Context According to the P-E fit framework, human behavior is a function of both the individual and the environment (Lewin, 1951). The fit between the characteristics of the individual and the attributes and expectations of the environment (i.e., P-E fit effect) has been studied in a number of fields, ranging from personality psychology to social psychology and organizational psychology (e.g., Endler & Magnusson, 1976; Higgins, 2005; Kristof, 1996). In all Ó 2018 Hogrefe Publishing

M. Suhlmann et al., Fitting-In to Succeed at University

fields it is assumed that a P-E fit effect will positively affect well-being and motivation and negatively affect behavioral intentions to drop out of the environment (e.g., Chatman, 1989; Kristof, 1996; Edwards & Shipp, 2007; for a metaanalysis, see Verquer, Beehr, & Wagner, 2003). Similarly, the fit between the individuals’ needs and abilities and the organizational supplies and demands leads to a number of positive outcomes (Kristof-Brown, Zimmerman, & Johnson, 2005). Research on university norms suggests that there might be person and environment characteristics that could elicit such fit effects in undergraduate students. Compared to other educational institutions, universities, in particular, demand increased initiative and self-regulation from its students to succeed (Bryde & Milburn, 1990). Universities communicate that they expect students to show high independence (i.e., independent university norms, Menges & Exum, 1983; Stephens, Fryberg, Markus, Johnson, & Covarrubias, 2012). In other words, students are expected to develop their own personal opinions and pave their own individual pathways at university. In line with these expectations of students, a survey among university administrators confirmed that independence is essential for succeeding at their universities (Stephens, Fryberg, et al., 2012). There is some evidence that independent norms are also dominant at universities in other Western societies (e.g., in Germany, Gellert, 1993; Levesque, Zuehlke, Stanek, & Ryan, 2004; Nenniger, 1991). Therefore, we assume that independence is a relatively universal key norm (or demand) that students face at universities (in the US as well as in Germany). At the same time, it seems likely that the extent to which students are or perceive to be confronted with this norm might vary considerably, depending on (among other things) the extent and type of group assignments in the curriculum and on how much competition and a Socratic approach to learning is stressed by the study program (Anyon, 1980; Kim, 2002; Tweed & Lehman, 2002). In sum, this suggests that independent norms are widespread and essential demands for academic achievement. Even though other norms have recently undergone a change (e.g., diversity), students have always been and still are expected to express a certain degree of independence when studying at university. Therefore, we will focus on this norm in the current research.

Dignity Self-Construal Fits Independent University Norms Which individual student characteristics fit with independent university norms resulting in a P-E fit effect? Past research identified that students with an independent self-construal are more likely to experience a fit with the Ó 2018 Hogrefe Publishing

independent university norms (Stephens, Townsend, et al., 2012). In the present study we argue that the same logic also applies for a specific form of independent selfconstrual, namely students’ dignity self-construal. The main idea underlying the dignity self-construal is that at birth each individual possesses an equal intrinsic value which cannot be taken away by others (Ayers, 1984; Kim & Cohen, 2010; Leung & Cohen, 2011). Thus, individuals with a high dignity self-construal believe that their worth does not depend on others’ but rather on their own judgment. Hence, the higher individuals’ dignity self-construal, the more independent from others is their self-esteem. Students who are independent of others in their selfworth (i.e., high dignity self-construal) are advantaged at university as they naturally act in accordance with the independent university norms. High independent university norms imply the expectation that the student will succeed independently at university by studying and performing independently (i.e., without much support and reinforcement). Hence, even in the face of possible threats to their self-worth, students with a high dignity self-construal probably question less whether they belong to university. Rather, as they experience the demands and possible threats to self-worth during their pursuit of finding and navigating their own individual path through college (Stephens, Fryberg, et al., 2012), they are likely to “feel at home” at university (Fulmer et al., 2010; Stephens, Brannon, Markus, & Nelson, 2015). Therefore, students with a high dignity self-construal are particularly likely to experience a fit with universities they perceive as endorsing independent norms. Like P-E fit in other contexts, this P-E fit in the university context could elicit a number of positive effects.

Consequences of P-E Fit: Belonging According to Stephens, Fryberg, et al. (2012) P-E fit (or in their terminology, cultural match) leads to the perception of lower task difficulty which in turn mediates the P-E fit effects, resulting in higher academic success. Whereas this earlier research found support for perceived task difficulty as a relevant mediator, we argue that another more socially oriented outcome of P-E fit might influence a broad range of outcomes related to academic success, namely the sense of belonging to the university. The university environment is characterized by interactions with other students and academic staff and is therefore a social environment. Hence, it seems reasonable that students who experience a fit between their self-construal and the university norms, which are communicated by the faculty and shown by their co-students, are more likely to feel that they belong to the university. This belonging to the university is the feeling that one is a valued part of the university community due Social Psychology (2018), 49(1), 16–28

to one’s positive relations with other students as well as with academic staff (Goodenow, 1993). Indeed, it seems likely that students who hold values that are in line with what is expected from them in the university environment do experience a greater fit and related greater sense of belonging to the university (cf. Walton & Cohen, 2007; for a similar argument regarding ethnic minorities and belonging uncertainty). Moreover, there is evidence from the high school context showing that belonging differs among students from diverse backgrounds. Mok and colleagues (Mok, Martiny, Gleibs, Keller, & Froehlich, 2016) found that belonging only increased among those students whose ethnic background fit the environmental norms that were operationalized as the proportion of migrant students in the classroom. Thus, when there was an increase in the proportion of Turkish students in the classroom (i.e., environmental norm), only Turkish origin students, whose personal values were likely to fit these norms, expressed an increased sense of belonging. No increase in sense of belonging among German-origin students was found. This idea is in line with the similarity attraction paradigm, which states that similar values facilitate interpersonal relations (Berscheid & Reis, 1998; Byrne, 1971) and bind people together (Kandel, 1978; McPherson, Smith-Lovin, & Cook, 2001). Moreover, there is support that interpersonal similarities within ingroups are related to a higher feeling of group belonging (Easterbrook & Vignoles, 2013). In line with this notion, research on P-E fit in organizations has repeatedly shown that P-E fit has a positive impact on concepts related to social belonging such as organizational identification and commitment, satisfaction with supervisors and colleagues, group cohesion, low employee turnover intentions, and actual employee turnover (see meta-analysis by Kristof-Brown et al., 2005). In sum, the fit between dignity self-construal and independent norms of the university discussed above should result in a higher sense of belonging to the university. Therefore, we predicted that a higher dignity self-construal is related to higher sense of belonging to the university when the university norms are perceived more as endorsing independence (Hypothesis 1).

Consequences of P-E Fit: Beyond Belonging As indicated by the results of the meta-analysis by KristofBrown et al. (2005) summarized above, the effects of P-E fit go far beyond belonging. In the present study, we considered three of these additional outcomes that previous research identified to be related to students’ academic success (Quiroga, Janosz, Bisset, & Morin, 2013; Vallerand, Fortier, & Guay, 1997). First, P-E fit effects extend to lower Social Psychology (2018), 49(1), 16–28

M. Suhlmann et al., Fitting-In to Succeed at University

stress and, thus, better well-being (e.g., Caplan, 1987; Caplan & Harrison, 1993). In organizations, for example, decreased levels of burnout occur for individuals who experienced a fit (Maslach & Leiter, 1997). In addition, experimental studies show that a misfit between group and personal standards increased negative affect and physical and mental symptoms, representing decreased well-being (Sassenberg, Matschke, & Scholl, 2011). Most relevant, in the university context increased levels of well-being were shown to be related to (a) the fit of the individual with the norms of fellow students (Sortheix & Lönnqvist, 2015) and (b) with the norms of the university major (Sagiv & Schwartz, 2000). Second, as evident from organizational and vocational research (Maslach & Leiter, 2008), as well as from educational research (Eccles et al., 1993) a fit between the individual and the environment is likely to be associated with enhanced engagement and motivation. Thus, individuals pursue and persist in tasks and contexts that they believe are congruent with their personal goals and values. This idea resonates with findings from Elliot and Harackiewicz (1994) who found that intrinsic motivation was greatest among students who experienced a fit between their achievement orientation (i.e., student value) and the specific goal-setting focus of a task (i.e., environmental norm). Hence, there exists evidence for positive effects of a P-E fit on academic motivation. Finally, a fit between the individual’s values and the environmental norms is also related to the persistence of the person in the specific environment (Caplan, 1987; Holland, 1997; Kristof, 1996). In the context of undergraduate studies, research found that Latino college students who did not experience a fit between their ethnic minority values and the university environment, which is based on ethnic majority norms of individualism and competition, also showed higher intentions to drop out of university, compared to white American students, who experienced greater fit (Gloria, Castellanos, & Orozco, 2005; Gloria & Kurpius, 1996). In sum, theory and empirical evidence suggest that higher P-E fit will not only facilitate feelings of belonging but also well-being and motivation. In addition, it should prevent students from forming dropout intentions. In what follows we argue that these additional effects are brought about by belonging.

Belonging Mediates Effects of P-E Fit on Well-Being, Motivation, and Dropout Intention A wide range of studies suggests that P-E fit is likely to affect well-being, motivation, and dropout intentions Ó 2018 Hogrefe Publishing

M. Suhlmann et al., Fitting-In to Succeed at University

through belonging to the university. We present arguments for the three outcomes below. First, sense of belonging is known to assert a positive influence on well-being and health (for review, see Baumeister & Leary, 1995). According to the buffer hypothesis, a positive relationship with one’s social group and social support can act as a buffer to stress, positively affecting individuals’ well-being (for evidence, see Cohen & Wills, 1985; Tay & Diener, 2011; Walton & Cohen, 2011). Hence, belonging might result in particularly positive effects in the context of studying at university, because it is experienced as a very stressful period. For instance, levels of distress of enrolled students were found to be higher compared to levels of distress at preregistration (Bewick, Koutsopoulou, Miles, Slaa, & Barkham, 2010). We therefore expected that students who experience a higher sense of belonging to the university will also report higher well-being (Hypothesis 2a). Moreover, we expected a moderated mediation effect of the P-E fit interaction of students’ dignity self-construal and perceived independent university norms on well-being via sense of belonging to the university (Hypothesis 3a). Second, it seems likely that P-E fit also affects motivation via increased sense of belonging to the university. For example, self-determination theory argues that more autonomous forms of motivation (e.g., intrinsic motivation) are contingent on the fulfillment of the need to belong (cf. relatedness), because only in a socially supportive learning environment does the student feel safe and free to pursue personal interests (Deci, Vallerand, Pelletier, & Ryan, 1991; Niemiec & Ryan, 2009). Worrying about whether one belongs to a social group and whether one is a valued member represents a chronic stressor that undermines students’ motivation and engagement (Yeager, Walton, & Cohen, 2013). In line with this notion, even minimal belonging cues might increase the motivation to engage in the context one is part of (Walton, Cohen, Cwir, & Spencer, 2012). Moreover, evidence among school children supports that belonging influences academic motivation (e.g., Furrer & Skinner, 2003; Hamre & Pianta, 2005). Therefore, we predict that a higher sense of belonging to the university will be associated with higher academic motivation (Hypothesis 2b). In addition, we expected a moderated mediation effect of the P-E fit interaction of students’ dignity self-construal and perceived independent university norms on academic motivation via sense of belonging to the university (Hypothesis 3b). Finally, there is evidence that belonging is negatively related to dropout intentions and behavior. For instance, in the work context, a sense of belonging to an organization was associated with lower employee turnover intentions (Ng, 2015). In addition, research on individual mobility in an intergroup context has shown that social identification (i.e., a specific form of belonging) reduces the likelihood Ó 2018 Hogrefe Publishing

of leaving negatively evaluated groups (Ellemers, Spears, & Doosje, 1997). Likewise, students who felt part of the university community were less likely to consider dropping out of their study (Morrow & Ackermann, 2012). Therefore, we expected that a higher sense of belonging to the university would be related to a lower dropout intention (Hypothesis 2c). Accordingly, we also expected a moderated mediation effect of the P-E fit interaction of students’ dignity selfconstrual and perceived independent university norms on dropout intention via sense of belonging to the university (Hypothesis 3c).

The Present Study The aim of the present study was twofold. First, we aimed to investigate the applicability of the P-E fit framework to the German university context by examining how students’ dignity self-construal, as a specific student value variable, interacts with perceived independent university norms in relating to students’ sense of belonging to the university. Second, we aimed to investigate whether sense of belonging to the university transmits the P-E fit effects of student values and university norms onto relevant variables of students’ health and academic success. In sum, we expected that a higher dignity self-construal is related to higher sense of belonging to the university when the university norms are more perceived to endorse independence (Hypothesis 1). Sense of belonging to the university in turn was expected to be associated with higher well-being (Hypothesis 2a), higher academic motivation (Hypothesis 2b), and lower dropout intention (Hypothesis 2c). Together, these hypotheses suggest that there should be indirect effects of the interaction between dignity self-construal and perceived independent university norms via sense of belonging to the university on well-being, academic motivation, and dropout intentions (Hypotheses 3a–3c; see also Figure 1 for an overview of the theoretical model). These predictions were tested in a cross-sectional online questionnaire study. By testing the seven hypotheses derived above, the present research adds to existing research in three ways. First, it extends previous research that has primarily focused on how membership in specific underrepresented groups fits to the environment by taking a more intrapersonal approach focusing on belonging as an underlying psychological process variable. Second, by specifically focusing on the dignity self-construal, a new concept is introduced that has the potential to interact with the perceived environmental norms. Third, by linking the research on student academic success at university to work on P-E fit, the current research integrates these two fields to test a theoretical process model that underscores the key role of students’ sense of university belonging, which to Social Psychology (2018), 49(1), 16–28

Figure 1. Overview of the proposed theoretical model.

the best of our knowledge has not yet received adequate attention in that combination.

Method Sample and Design Four hundred five undergraduate students of different majors from a German university participated in a crosssectional online questionnaire study in exchange for a €5 online store voucher. Participants were registered in six1 different faculties: philosophy (41.3%), mathematical and natural sciences (24.5%), economics and social sciences (14.5%), medicine (12.8%), law (4.3%), and theology (combining Evangelic, Catholic, and Islamic faculties, 2.6%). No significant differences between faculties were found in perceived independent university norms. An additional 287 participants started the survey, but did not complete some or any of the main measures (i.e., 87% of them completed neither the two independent variables nor any of the dependent variables). Therefore, they could not be included in the analyses reported below. Among the 405 participants who completed the whole survey, there were very few who were 30 years or older – a factor that might undermine the sense of belonging directly or indirectly, because these students do not fit the prototype of the student body and are thus likely to drop out for different reasons (Bean & Metzner, 1985). Ideally, we would have liked to control for this variable. However, given that this variable was heavily skewed, it would have been problematic to include it in the analysis. Thus, we excluded these participants from the analyses reported below. The final sample consisted of 367 students (74% female, with 27 participants who did not indicate their sex, Mage = 23.11, range: 18–29 years).

M. Suhlmann et al., Fitting-In to Succeed at University

The university at which the study was conducted represents one of the oldest universities in Germany and is located in a student town in Southwest Germany. The university is relatively diverse and large as it consists of a total of eight faculties, offering nearly 300 different degree programs and has more than 27,000 registered students at the time the study was conducted. Moreover, the university has a good reputation as research university and belongs to the top 11 “excellence universities” of Germany. This is also communicated explicitly via the university website. Therefore, it is likely that independence norms are relatively prevalent at this university.

Procedure Participants were recruited for a study about “German university culture” via an e-mail with a link to the online questionnaire which was sent to all registered undergraduate students. After providing informed consent, which assured the anonymity of responses and other rights, participants completed the measures presented in detail below in the listed order. Socio-demographics were assessed at the end. After participants had completed the questionnaire they had the opportunity to provide an e-mail address where the voucher should be sent. The questionnaire contained scales that assessed concepts beyond the ones listed below; those were not relevant for the current hypotheses. It was presented in German and English language, as we originally intended to also collect data from international students and students with a migration background. However, we failed to recruit a sufficient number of them.

Measures Dignity self-construal was measured with nine items (α = .88) taken from the Inalienable Worth Scale (Leung & Cohen, 2011; e.g., “How others treat me is irrelevant to my worth as a person.”) and the Others’ Approval subscale of the Contingencies of Self-Worth Scale (Crocker, Luhtanen, Cooper, & Bouvrette, 2003; e.g., “I don’t care what other people think of me.”). A combination of these scales was used as both of them captured relevant aspects of the concept of independence of others in one’s own self-worth and no single dignity self-construal scale is available. Participants provided responses on a 7-point scale ranging from 1 (= highly disagree) to 7 (= highly agree). In this and all other scales, reversed items were recoded before items were averaged. Higher values indicate a higher dignity selfconstrual.

The three theological faculties (Evangelic, Catholic, and Islamic) were combined into one category due to small number of participants.

Social Psychology (2018), 49(1), 16–28

Ó 2018 Hogrefe Publishing

M. Suhlmann et al., Fitting-In to Succeed at University

Perceived independent university norms were measured with six items adapted from Stephens, Fryberg, et al. (2012) college culture measure (e.g., “The University of XXX expects me to work independently.”) and one selfcreated item (“I feel that the University of XXX values the autonomy and independence of their students.”; α = .68). To the best of our knowledge this is the only existing scale that directly assessed the perception of independence university norms. We adapted this scale for students as it was originally used to measure perception of independence norms among university administrators. Again, items were rated on a 7-point scale ranging from 1 (= highly disagree) to 7 (= highly agree). Higher values indicate a higher perception of independence norms. Sense of belonging to the university was measured by 12 items (α = .85) adapted from the Psychological Sense of School Membership Scale (Goodenow, 1993; e.g., “Other students of my university like me the way I am.”). We employed this scale as it represents a reliable and valid scale that assesses students’ sense of belonging to the educational institution (Goodenow, 1993). We adapted this scale to the university context as this scale was originally developed for the middle and high school context. Responses were assessed on a 7-point scale ranging from 1 (= highly disagree) to 7 (= highly agree). Higher values indicate a higher sense of belonging. Well-being was measured by the 25-item (α = .93) Hopkins Symptoms Checklist (HSCL-25; Hesbacher, Rickels, Morris, Newman, & Rosenfeld, 1980; Nettelbladt, Hansson, Stefansson, Borgquist, & Nordström, 1993). This scale is a well-established measure of mental and physical symptoms associated with depression and anxiety, those aspects of well-being that are impaired among many university students (Mistler, Reetz, Krylowicz, & Barr, 2012). The participants indicated how often during the last week they had experienced specific depressive (e.g., “feeling blue”) and anxiety symptoms (e.g., “heart pounding or racing”). Response options ranged from 1 (= not at all) to 4 (= extremely). For interpretation purposes responses were recoded so that higher values indicate higher well-being. Academic motivation was measured with 12 items (α = .79) including the intrinsic motivation to know (three items), to experience stimulation (three items), and toward accomplishment (two items) as well as the external identified subscale of the Academic Motivation Scale (AMS; Vallerand et al., 1992) – a very well-established measure of academic motivation. We decided to focus on these subscales as they capture autonomous forms of motivation either directly in the present (e.g., intrinsic subscale) or in the future (e.g., external identified subscale). Participants were asked to indicate how much different answers to the question “Why do you go to university?” apply to them (e.g. “Because I experience pleasure and satisfaction while Ó 2018 Hogrefe Publishing

learning new things.”, “Because eventually it will enable me to enter the job market in a field that I like.”), on a scale ranging from 1 (= does not correspond at all) to 7 (= corresponds exactly). Dropout intention was measured directly with one item by asking whether the student considered dropping out from the present study program, with the binary response options “no” (coded as 1) and “yes” (2; see also Vallerand, Fortier, & Guay, 1997). We chose this one-item measure as we were interested in a behavioral intention which is not a multifaceted construct.

Results For bivariate correlations and descriptive statistics, see Table 1. First, we tested the hypothesized model with a path analysis conducted in Mplus version 7.3 (Muthén & Muthén, 1998–2015) with the weighted least squares means and variance adjusted (WLSMV) estimator as dropout intention was measured on a categorical scale. Thus, similar to the PROCESS approach used by Hayes (2013) we calculated regression analyses, but as we had several dependent variables we analyzed the data in a structural equation framework. We allowed for correlations of the dependent variables. The P-E fit interaction was computed by multiplying the z-standardized independent variables: dignity self-construal and perceived independent university norms. The results confirmed the expected relationships (see Table 2 and Figure 2). However, the overall fit of the model was not optimal w2 = 45.58, df = 9, CFI = 0.86, RMSEA = .11. Therefore, we decided to adjust the model based on the modification indices and included an additional path between dignity self-construal and well-being as well as between perceived independent university norms and academic motivation (i.e., direct main effects). As we made no assumption about the direct main effects of dignity self-construal and perceived independent university norms on the outcome variables, but rather focused on interactions and their indirect effects, these paths neither support nor contradict our hypothesis. The resulting final model (see Table 3 and Figure 3) had an excellent fit, w2 = 7.82, df = 7, CFI = 1.00, RMSEA = .02. The final model indicates that sense of belonging to the university is significantly predicted by the perceived independent university norms, B = .30, SE = .04, p < .001, but not by dignity self-construal B = .03, SE = .04, p = .53. More importantly, the predicted Dignity SelfConstrual Perceived Independent University Norms interaction on sense of belonging to the university was significant (B = .12, SE = .03, p < .001). Simple slope analyses of the final model revealed that only for students with a Social Psychology (2018), 49(1), 16–28

M. Suhlmann et al., Fitting-In to Succeed at University

Table 1. Pearson’s correlations and descriptive statistics for all main variables of the present study Variable

(1)

(2)

(3)

(4)

(5)

(6)

–

(1) Dignity Self-construal

–

(2) Perceived Independent University Norms

.05

(3) Sense of Belonging to the University

.04

.33**

(4) Well-being

.17**

.06

.36**

(5) Academic Motivation

.02

.39**

.41**

.19**

(6) Dropout Intention (1 = no; 2 = yes)

.02

.06

.20**

.28**

3.90

4.91

5.10

3.31

5.02

1.07

1.18

0.84

0.93

0.53

0.84

0.26

Range/percentage distribution

1.00–6.89

1.57–6.71

1.42–7.00

1.00–4.04

2.33–6.92

– – – –

.26**

1 (93%)–2 (7%)

Notes. N = 367 for all variables besides for academic motivation (N = 366) and dropout intention (N = 364) due to missing values; **p < .01.

Table 2. Decomposition of effects of the hypothesized path analysis model Regression of . . . on

Standardized coefficient (β)

Unstandardized coefficient

p-value

Sense of Belonging to the University Dignity Self-construal (DSC)

.07

.06

.04

.160

Perceived Independent University Norms (PIUN)

.39

.35

.04

< .001

PE-Fit Interaction (DSC PIUN)

.13

.11

.03

< .001

.35

.20

.02

< .001

.49

.45

.04

< .001

.33

.37

.09

< .001

Well-being Sense of Belonging to the University Motivation Sense of Belonging to the University Dropout intentiona Sense of Belonging to the University Note. N = 367; aDropout intention was coded with: 1 = no, 2 = yes.

Figure 2. Hypothesized path analysis model with unstandardized estimates. ***p < .001.

higher dignity self-construal (1 SD above the mean), the data supports that the higher they perceived the university norms to stress independence (1 SD above the mean) the higher was their sense of belonging to the university (B = .15, SE = .05, p = .003). In contrast, when the university norms were not perceived to endorse independence (1 SD below the mean) dignity self-construal was not related to sense of belonging to the university (B = .10, SE = .06,

Social Psychology (2018), 49(1), 16–28

p = .09; see also Figure 4). Thus, the results support the predicted student-university fit effect of Hypothesis 1. Supporting Hypothesis 2a, sense of belonging to the university positively predicted well-being (B = .19, SE = .02, p < .001). Additionally, well-being was also significantly predicted by dignity self-construal (B = .08, SE = .02, p < .001) in the final model. Moreover, in line with Hypothesis 2b, higher sense of belonging to the university was positively related to academic motivation (B = .28, SE = .04, p < .001). The final model also showed a significant positive direct effect of independent university norms on motivation (B = .26, SE = .04, p < .001). Finally, confirming Hypothesis 2c, sense of belonging to the university significantly negatively predicted dropout intention (B = .36, SE = .08, p < .001).

Indirect Effect Analyses Bootstrapping (N = 20,000) was used to calculate 95% confidence intervals (CIs) of the indirect effects. The results show a significant moderated mediation index for

Ó 2018 Hogrefe Publishing

M. Suhlmann et al., Fitting-In to Succeed at University

Table 3. Decomposition of effects of the final path analysis model Regression of . . . on. . .

Standardized coefficient (β)

Unstandardized coefficient

p-value

Sense of Belonging to the University Dignity Self-construal (DSC)

.03

.04

.530

Perceived Independent University Norms (PIUN)

.32

.30

.04

< .001

PE-Fit Interaction (DSC PIUN)

.14

.12

.03

< .001

Dignity Self-construal

.16

.19

.02

< .001

Sense of Belonging to the University

.33

.08

.02

< .001

Perceived Independent University Norms (PIUN)

.31

.26

.04

< .001

Sense of Belonging to the University

.31

.28

.04

< .001

.33

.36

.08

< .001

Well-being

Motivation

Dropout Intentiona Sense of Belonging to the University Note. N = 367; aDropout intention was coded with: 1 = no, 2 = yes.

Figure 3. Final path analysis model with unstandardized estimates. ***p < .001.

students’ well-being (B = .02, SE = .01, 95% CI [.01, .05]). When the university norms were perceived to endorse independence (1 SD above the mean) students with a higher dignity self-construal experienced greater well-being, transmitted via sense of belonging to the university (B = .03, SE = .01, 95% CI [.01, .06]). In contrast, if anything, the opposite trend was present when university norms were not perceived to endorse independence (1 SD below the mean: B = .02, SE = .01, 95% CI [ .05, .01]). Thus, the results yield support of Hypothesis 3a. For students’ academic motivation, we likewise found a significant moderated mediation effect (B = .03, SE = .02, 95% CI [.01, .07]). Again, students with a higher dignity self-construal experienced greater academic motivation, transmitted via sense of belonging to the university, when the university norms were perceived to endorse independence (1 SD above the mean: B = .04, SE = .02, 95% CI [.01, .09]). If anything, the opposite trend was present when the university norms were not perceived to endorse independence (1 SD below the mean: B = .03, SE = .02, 95% CI [ .07, .01]). These findings are consistent with Hypothesis 3b. Ó 2018 Hogrefe Publishing

Figure 4. Interaction graph of Dignity Self-construal Perceived Independent University Norms interaction.

Finally, for students’ intention to drop out there was likewise evidence for a moderated mediation effect (B = .04, SE = .03, 95% CI [ .11, .01]). Students with a higher dignity self-construal expressed fewer intentions to drop out, transmitted via sense of belonging to the university, when university norms were perceived to endorse independence (1 SD above the mean: B = .05, SE = .03, 95% CI [ 13, .01]). In contrast, the effect pointed in the opposite direction when university norms were not perceived to endorse independence (1 SD below the mean: B = .03, SE = .03, 95% CI [ .01, .11]). In sum, these findings are consistent with Hypothesis 3c.

Discussion The present study was designed to investigate whether a PE fit between students’ values and university norms positively predicts students’ sense of belonging to the university and thereby also students’ health and academic outcomes. Social Psychology (2018), 49(1), 16–28

Consistent with Hypothesis 1 we found that a higher dignity self-construal was related to higher sense of belonging to the university when the university norms were more perceived to endorse independence. Whereas previous research had found a relation between P-E fit effect and factors related to academic success (e.g., Harackiewicz, Barron, Tauer, & Elliot, 2002; Harms, Roberts, & Winter, 2006), our results support that the fit between students’ self-construal and university norms also has implications for the sense of belonging to the university. Thereby this study identifies dignity self-construal as relevant student value variable. In addition, we found support for a relation between students’ sense of belonging to the university and their increased well-being (Hypothesis 2a), increased academic motivation (Hypothesis 2b), and students’ lower intention to drop out (Hypothesis 2c). These findings together with the supported indirect effects of Hypotheses 3a–3c stress the importance of a fit between student values and university norms, because this fit relates to belonging and, resulting from belonging, also to important aspects of academic life and success. They are in line with previous research indicating that intragroup similarity and belonging are positively related (Easterbrook & Vignoles, 2013). It therefore seems plausible that the fit that students experience between their self-construal (and the implied values) and the university norms increases the psychological closeness between the student and the university, preventing the students from experiencing uncertainty about belonging (Walton & Cohen, 2007). This study extends previous research in three important ways. First, the study uses belonging as an underlying psychological process variable, which translates a P-E fit effect into students’ increased well-being and academic motivation as well as decreased dropout intention, whereas earlier research in the higher education realm focused on task appraisals (Stephens, Fryberg, et al., 2012). Second, dignity self-construal is introduced as a relevant student variable that fits with high independent university norms, creating a P-E fit effect. Finally, this study integrates research on student academic success at university and work on P-E fit to test a theoretical process model, which has not been done by past research. Overall, the current research shows that the basis of person-environment fit framework (see also Stephens, Townsend, et al., 2012) can and should be applied beyond (a) specific underrepresented student groups and (b) stress and performance outcomes. Hence, future research should take a broader approach and address other aspects that

M. Suhlmann et al., Fitting-In to Succeed at University

might vary in their fit between student values and university norms or expectations. For instance, it would be interesting to investigate whether a self-construal fit would increase trust in the university institution and academic staff in situations involving students’ feedback or whether a value fit would increase even more basic learning-related factors such as cognitive activation. Future research should also investigate whether, in addition to the subjective norms assessed in the present research, results also hold when university norms are assessed more objectively. This can be done, for example, by analyzing how universities communicate the expectations they have of their students via university websites. The prime limitation of the present study is that due to the correlational study design, we cannot draw any conclusions about the causal relations between the variables. Although the path analysis model tested a certain causal direction, which is reasonable based on theory and previous experimental research, we cannot rule out other causal directions, for example that the dependent variable wellbeing actually influences students’ sense of belonging to the university and not the other way around. Hence, experimental and longitudinal field research is required to address this deficit. One might also consider as a limitation that the hypothesized model had to be adapted to include two additional paths, which we initially did not consider. First, we found that dignity self-construal had a direct positive effect on well-being. Thus, students who are independent of others in their self-worth indicated fewer depressive and anxiety symptoms. This is in line with the finding by Crocker and Wolfe (2001) that there are interindividual differences regarding the extent to which self-esteem (i.e., a specific aspect of well-being) is contingent on others’ approval (i.e., the opposite of dignity self-construal). Second and likewise unexpectedly, we found that perceived independent university norms were related to more autonomous academic motivation. This finding is in line with one of the key arguments of self-determination theory, namely that autonomy enhances students’ intrinsic motivation (e.g., Niemiec & Ryan, 2009).2 In other words, both paths we have added to the modified model in our analysis could have been predicted based on the literature we did not primarily focus on and thus do not point to an actual limitation of the study itself. Moreover, both paths are independent of the main prediction. In addition, there are two smaller methodological limitations of the current study that we should acknowledge. First, the internal consistency of the perceived independent

Even though there exists a conceptual differentiation between independence and autonomy (Chirkov, Ryan, Kim, & Kaplan, 2003), we argue here for the relevance of the overlapping concept, comprising that the university norms are perceived to promote the students’ freedom and selfguidance and thus do not exert a controlling and structured influence.

Social Psychology (2018), 49(1), 16–28

Ó 2018 Hogrefe Publishing

M. Suhlmann et al., Fitting-In to Succeed at University

university scale was not satisfying. One possible explanation is the rather global and abstract nature of measuring (perceived) norms in general with an acceptable number of items. Therefore, future research should aim at replicating the current findings with an improved scale. Second, the current study only recruited participants from one institution and thus relied on the concept of a specific university norm. However, the independent university norm rating varied substantially (range: 1.57–6.71) and this limitation is thus not very likely to have a great impact. Nonetheless, future research should take samples from several university institutions to test whether the perceptions of independence norms universities emit differ across institutions or whether these norms are universal. This approach would also allow investigating whether our model can be replicated and the results can be generalized across several university institutions. The results of the current study suggest that students with a higher dignity self-construal are more advantaged at university the more that they perceive the university norms to stress independence, thus experiencing a P-E fit effect. This advantage plays out in a higher sense of belonging to the university and in turn in higher well-being, motivation, and lower intention to drop out. Therefore, the study has three main implications. First, dignity selfconstrual is an aspect of the self, which – together with the perceived independent university norms but potentially also with other characteristics of the university – creates the basis of a sense of belonging to the university community. Second, to accommodate the increasingly heterogeneous student body (i.e., to keep them motivated, healthy and to prevent them from dropping out), universities need to become more diverse, which might require creating programs that resonate with different self-construals and values via stressing different norms in line with the P-E fit framework. Otherwise, students with certain values and perceptions of university norms are more likely to drop out. Finally, by showing that sense of belonging to the university is related to all three indicators of student life and success assessed by us, the study highlights that belonging is a key variable in the context of study persistence. Future research should test how universities are able to simultaneously emit high and low independence norms so that students with diverse self-construals are likely to experience a fit. In this communication, the importance of independent and interdependent learning and deliverables could be mentioned, providing room for the experience of fit for individuals with different values (see also, Cross & Vick, 2001). Another path would be to stress individual and communal utility values and reasons to receive a degree. The former is likely to resonate better with individuals with a dignity or individualistic self-construal, the latter Ó 2018 Hogrefe Publishing

with individuals with a collectivistic or interdependent selfconstrual (Brown, Smith, Thoman, Allen, & Muragishi, 2015; Diekman, Clark, Johnston, Brown, & Steinberg, 2011). Possible channels for the communication of these norms can range from pictures to text on the university website or social media page or in welcome letters to newly registered students. Hence, if students feel that they do not belong to the university community this might not only negatively influence the students’ well-being and motivation, but also increase the possibility that the student will form intentions to drop out, thus decreasing the chance that the students successfully receive a university degree. Thus, the challenge for universities as institutions is to create a learning environment and a community that is not restricted to academic topics and testing, but also allows diverse students holding different values to flourish. Acknowledgment This research was funded by the LEAD Graduate School [GSC1028], a project by the Excellence Initiative of the German Federal and State governments.

References Altbach, P. G., Reisberg, L., & Rumbley, L. E. (2009). Tracking an Academic Revolution. A Report Prepared for the UNESCO 2009 World Conference on Higher Education. Paris, France: UNESCO Education. Anyon, J. (1980). Social class and the hidden curriculum of work. Journal of Education, 162, 67–92. https://doi.org/10.1177/ 002205748016200106 Ayers, E. (1984). Vengeance and justice. New York, NY: Oxford University Press. Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117, 497–529. https://doi. org/10.1037/0033-2909.117.3.497 Bean, J. P., & Metzner, B. S. (1985). A conceptual model of nontraditional undergraduate student attrition. Review of Educational Research, 55, 485–540. https://doi.org/10.2307/ 1170245 Berscheid, E., & Reis, H. T. (1998). Attraction and close relationships. In D. Gilbert, S. Fiske, & G. Lindzey (Eds.), Handbook of Social Psychology (4th ed., pp. 193–281). New York, NY: Oxford University Press. Bewick, B., Koutsopoulou, G., Miles, J., Slaa, E., & Barkham, M. (2010). Changes in undergraduate students’ psychological well-being as they progress through university. Studies in Higher Education, 35, 633–645. https://doi.org/10.1080/ 03075070903216643 Brown, E. R., Smith, J. L., Thoman, D. B., Allen, J. M., & Muragishi, G. (2015). From bench to bedside: A communal utility value intervention to enhance students’ biomedical science motivation. Journal of Educational Psychology, 107, 1116–1135. https://doi.org/10.1037/edu0000033 Bryde, J. F., & Milburn, C. M. (1990). Help to make the transition from high school to college. In R. L. Emans (Ed.), Understanding undergraduate education (pp. 203–213). Vermillion, SD: University of South Dakota Press.

Social Psychology (2018), 49(1), 16–28

Byrne, D. (1971). The attraction paradigm. New York, NY: Academic Press. Caplan, R. D. (1987). Person-environment fit theory and organizations: Commensurate dimensions, time perspectives, and mechanisms. Journal of Vocational Behavior, 31, 248–267. https://doi.org/10.1016/0001-8791(87)90042-X Caplan, R. D., & Harrison, R. V. (1993). Person-environment fit theory: Some history, recent developments, and future directions. Journal of Social Issues, 49, 253–275. https://doi.org/ 10.1111/j.1540-4560.1993.tb01192.x Chatman, J. A. (1989). Improving interactional organizational research: A model of person-organization fit. Academy of Management Review, 14, 333–349. https://doi.org/10.5465/AMR. 1989.4279063 Chirkov, V., Ryan, R. M., Kim, Y., & Kaplan, U. (2003). Differentiating autonomy from individualism and independence: A selfdetermination theory perspective on internalization of cultural orientations and well-being. Journal of Personality and Social Psychology, 84, 97–110. https://doi.org/10.1037/0022-3514. 84.1.97 Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310–357. https://doi.org/10.1037/0033-2909.98.2.310 Crocker, J., & Wolfe, C. T. (2001). Contingencies of self-worth. Psychological Review, 108, 593–623. https://doi.org/10.1037/ 0033-295X.108.3.593 Crocker, J., Luhtanen, R. K., Cooper, M. L., & Bouvrette, A. (2003). Contingencies of self-worth in college students: Theory and measurement. Journal of Personality and Social Psychology, 85, 894–908. https://doi.org/10.1037/0022-3514.85.5.894 Cross, S. E., & Vick, N. V. (2001). The interdependent selfconstrual and social support: The case of persistence in engineering . Personality and Social Psychology Bulletin, 27, 820–832. https://doi.org/10.1177/0146167201277005 Deci, E., Vallerand, R., Pelletier, L., & Ryan, R. (1991). Motivation and education: The self-determination perspective. Educational Psychologist, 26, 325–346. https://doi.org/10.1207/ s15326985ep2603&4_6 Diekman, A. B., Clark, E. K., Johnston, A. M., Brown, E. R., & Steinberg, M. (2011). Malleability in communal goals and beliefs influences attraction to stem careers: Evidence for a goal congruity perspective. Journal of Personality and Social Psychology, 101, 902–918. https://doi.org/10.1037/a0025199 Easterbrook, M., & Vignoles, V. L. (2013). What does it mean to belong? Interpersonal bonds and intragroup similarities as predictors of felt belonging in different types of groups. European Journal of Social Psychology, 43, 455–462. https:// doi.org/10.1002/ejsp.1972 Eccles, J. S., Wigfield, A., Midgley, C., Reuman, D., Mac Iver, D., & Feldlaufer, H. (1993). Negative effects of traditional middle schools on students’ motivation. Elementary School Journal, 93, 553–574. https://doi.org/10.1002/tea.20398 Edwards, J. R., Caplan, R. D., & Harrison, R. V. (1998). Personenvironment fit theory: Conceptual foundations, empirical evidence, and directions for future research. In C. L. Cooper (Ed.), Theories of organizational stress (pp. 28–67). Oxford, UK: Oxford University Press. Edwards, I. R., & Shipp, A. I. (2007). The relationship between person-environment fit and outcomes: An integrative theoretical framework. In C. Ostroff & T. A. Judge (Eds.), Perspectives on organizational fit (pp. 209–238). New York, NY: Erlbaum. Ellemers, N., Spears, R., & Doosje, B. (1997). Sticking together or falling apart: In-group identification as a psychological determinant of group commitment versus individual mobility. Journal of Personality and Social Psychology, 72, 617–626. https://doi.org/10.1037/0022-3514.72.3.617

Social Psychology (2018), 49(1), 16–28

M. Suhlmann et al., Fitting-In to Succeed at University

Elliot, A. J., & Harackiewicz, J. M. (1994). Goal setting, achievement orientation, and intrinsic motivation: A mediational analysis. Journal of Personality and Social Psychology, 66, 968–980. https://doi.org/10.1037/0022-3514.66.5.968 Endler, N. S., & Magnusson, D. (1976). Toward an interactional psychology of personality. Psychological Bulletin, 83, 956–974. https://doi.org/10.1037/0033-2909.83.5.956 European Centre for the Development of Vocational Training. (2013). Briefing note - Roads to recovery: Three skill and labour market scenarios for 2025. Retrieved from http://www.cedefop. europa.eu/files/9081_en.pdf Fabian, G., Rehn, T., Brandt, G., & Briedis, K. (2013). Karriere mit Hochschulabschluss? – Hochschulabsolventinnen und -absolventen des Prüfungsjahrgangs 2001 zehn Jahre nach dem Studienabschluss [Career with a university degree? Graduates of the 2001 examination year, ten years after graduation]. Retrieved from http://www.dzhw.eu/pdf/pub_fh/fh-201310.pdf Fulmer, C. A., Gelfand, M. J., Kruglanski, A. W., Kim-Prieto, C., Diener, E., Pierro, A., & Higgins, E. T. (2010). On “feeling right” in cultural contexts: How person-culture match affects selfesteem and subjective well-being. Psychological Science, 21, 1563–1569. https://doi.org/10.1177/0956797610384742 Furrer, C., & Skinner, E. (2003). Sense of relatedness as a factor in children’s academic engagement and performance. Journal of Educational Psychology, 95, 148–162. https://doi.org/10.1037/ 0022-0663.95.1.148 Gellert, C. (1993). Wettbewerb und leistungsorientierung im amerikanischen Universitatssystem. Beiträge zur vergleichenden Bildungsforschung, Vol. 1 [Competition and achievement orientation in the American university system: Contributions to comparative educational research, Vol. 1]. Frankfurt am Main, Germany: Peter Lang. Gloria, A. M., Castellanos, J., & Orozco, V. (2005). Perceived educational barriers, cultural fit, coping responses, and psychological well-being of Latina undergraduates. Hispanic Journal of Behavioral Sciences, 27, 161–183. https://doi.org/ 10.1177/0739986305275097 Gloria, A. M., & Kurpius, S. E. R. (1996). The validation of the Cultural Congruity Scale and the University Environment Scale with Chicano/a students. Hispanic Journal of Behavioral Sciences, 18, 533–549. https://doi.org/10.1177/07399863960184007 Goodenow, C. (1993). The psychological sense of school membership among adolescents: Scale development and educational correlates. Psychology in the Schools, 30, 79–90. https://doi.org/ 10.1002/1520-6807(199301)30:1<79::AID-PITS2310300113>3.0. CO;2-X Grobe, T., & Steinmann, S. (2015). Gesundheitsreport 2015 Techniker Krankenkasse: Gesundheit von Studierenden. Hamburg [Health Report 2015 by Techniker Krankenkasse: Students’ Health]. Retrieved from www.tk.de/centaurus/servlet/contentblob/718612/Datei/143833/Gesundheitsreport-2015.pdf Hamre, B. K., & Pianta, R. C. (2005). Can instructional and emotional support in the first-grade classroom make a difference for children at risk of school failure? Child Development, 76, 949–967. https://doi.org/10.1111/j.1467-8624.2005.00889.x Harackiewicz, J. M., Barron, K. E., Tauer, J. M., & Elliot, A. J. (2002). Predicting success in college: A longitudinal study of achievement goals and ability measures as predictors of interest and performance from freshman year through graduation. Journal of Educational Psychology, 94, 562–575. https:// doi.org/10.1037/0022-0663.94.3.562 Harms, P. D., Roberts, B. W., & Winter, D. (2006). Becoming the Harvard man: Person-environment fit, personality development, and academic success. Personality and Social Psychology Bulletin, 32, 851–865. https://doi.org/10.1177/ 0146167206287720

Ó 2018 Hogrefe Publishing

M. Suhlmann et al., Fitting-In to Succeed at University

Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: The Guilford Press. Hesbacher, P. T., Rickels, K., Morris, R. J., Newman, H., & Rosenfeld, H. (1980). Psychiatric illness in family practice. Journal of Clinical Psychiatry, 41, 6–10. Heublein, U., Richter, J., Schmelzer, R., & Sommer, D. (2012). Die Entwicklung der Schwund- und Studienabbruchquoten an den deutschen Hochschulen. Statistische Berechnungen auf der Basis des Absolventenjahrgangs 2010 [The development of shrinkage and drop-out rates at German universities. Statistical calculations based on the graduation year 2010]. Forum Hochschule 3/2012, Hannover, Germany: HIS. Retrieved from www.dzhw.eu/pdf/pub_fh/fh-201203.pdf Higgins, E. T. (2005). Value from regulatory fit. Current Directions in Psychological Science, 14, 209–213. https://doi.org/10.1111/ j.0963-7214.2005.00366.x Holland, J. L. (1997). Making vocational choices: A theory of vocational personalities and work environments (3rd ed.). Odessa, FL: Psychological Assessment Resources. Kandel, D. B. (1978). Similarity in real-life adolescent friendship pairs. Journal of Personality and Social Psychology, 36, 306–312. https://doi.org/10.1037/0022-3514.36.3.306 Kim, H. S. (2002). We talk, therefore we think? A cultural analysis of the effect of talking on thinking. Journal of Personality and Social Psychology, 83, 828–842. https://doi.org/10.1037/00223514.83.4.828 Kim, Y.-H., & Cohen, D. (2010). Information, perspective, and judgments about the self in face and dignity cultures. Personality & Social Psychology Bulletin, 36, 537–550. https://doi.org/ 10.1177/0146167210362398 Kristof, A. L. (1996). Person-organization fit: An integrative review of its conceptualizations, measurement, and implications. Personnel Psychology, 49, 1–49. https://doi.org/10.1111/ j.1744-6570.1996.tb01790.x Kristof-Brown, A. L., Zimmerman, R. D., & Johnson, E. C. (2005). Consequences of individuals’ fit at work: A meta-analysis of person-job, person-organization, person-group, and personsupervisor fit. Personnel Psychology, 58, 281–342. https://doi. org/10.1111/j.1744-6570.2005.00672.x Leung, A. K.-Y., & Cohen, D. (2011). Within- and between-culture variation: Individual differences and the cultural logics of honor, face, and dignity cultures. Journal of Personality and Social Psychology, 100, 507–526. https://doi.org/10.1037/a0022151 Levesque, C., Zuehlke, A. N., Stanek, L. R., & Ryan, R. M. (2004). Autonomy and competence in German and American university students: A comparative study based on Self-Determination Theory. Journal of Educational Psychology, 96, 68–84. https:// doi.org/10.1037/0022-0663.96.1.68 Lewin, K. (1951). Field Theory in Social Science. New York, NY: Harper and Row. Maslach, C., & Leiter, M. P. (1997). The Truth about Burnout. San Francisco, CA: Jossey Bass. Maslach, C., & Leiter, M. P. (2008). Early predictors of job burnout and engagement. Journal of Applied Psychology, 93, 498–512. https://doi.org/10.1037/0021-9010.93.3.498 McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of a feather: Homophily in social networks. Annual Review of Sociology, 27, 415–444. https://doi.org/10.1146/annurev.soc. 27.1.415 Menges, R. J., & Exum, W. H. (1983). Barriers to the progress of women and minority faculty. The Journal of Higher Education, 123–144. https://doi.org/10.1080/00221546.1983.11778167 Mistler, B. J., Reetz, D. R., Krylowicz, B, & Barr, V. (2012). The Association for University and College Counseling Center Directors Annual Survey and Report Overview. Retrieved from http://www.apa.org/monitor/2013/06/college-students.aspx

Ó 2018 Hogrefe Publishing

Mok, S. Y., Martiny, S. E., Gleibs, I. H., Keller, M. M., & Froehlich, L. (2016). The relationship between ethnic classroom composition and Turkish-origin and German students’ reading performance and sense of belonging. Frontiers in Psychology, 7, 1071. https://doi.org/10.3389/fpsyg.2016.01071 Morrow, J. A., & Ackermann, M. E. (2012). Intention to persist and retention of first-year students: The importance of motivation and sense of belonging. College Student Journal, 46, 483–491. Muthén, L. K., & Muthén, B. O. (1998–2015). Mplus User’s Guide (7th ed.). Los Angeles, CA: Muthén & Muthén. Nenniger, P. (1991). Motivierung studentischen Lernens im Kulturvergleich [Motivating students’ learning strategies]. Zeitschrift fuer Psychologie, 199, 145–165. Nettelbladt, P., Hansson, L., Stefansson, C. G., Borgquist, L., & Nordström, G. (1993). Test characteristics of the Hopkins Symptom Check List-25 (HSCL-25) in Sweden, using the Present State Examination (PSE-9) as a caseness criterion. Social Psychiatry and Psychiatric Epidemiology, 28, 130–133. https://doi.org/10.1007/BF00801743 Ng, T. W. H. (2015). The incremental validity of organizational commitment, organizational trust, and organizational identification. Journal of Vocational Behavior, 88, 154–163. https://doi. org/10.1016/j.jvb.2015.03.003 Niemiec, C. P., & Ryan, R. M. (2009). Autonomy, competence, and relatedness in the classroom: Applying self-determination theory to educational practice. Theory and Research in Education, 7, 133–144. https://doi.org/10.1177/1477878509104318 OECD. (2016). Education at a glance 2016: OECD indicators. Paris, France: OECD Publishing. https://doi.org/10.187/ eag-2016-en, Retrieved from https://stats.oecd.org/Index. aspx?DataSetCode=RFOREIGN Quiroga, C. V., Janosz, M., Bisset, S., & Morin, A. J. S. (2013). Early adolescent depression symptoms and school dropout: Mediating processes involving self-reported academic competence and achievement. Journal of Educational Psychology, 105, 552–560. https://doi.org/10.1037/a0031524 Sagiv, L., & Schwartz, S. H. (2000). Value priorities and subjective well-being: Direct relations and congruity effects. European Journal of Social Psychology, 30, 177–198. https://doi.org/ 10.1002/(SICI)1099-0992(200003/04)30:2<177::AID-EJSP982>3.0. CO;2-Z Sassenberg, K., Matschke, C., & Scholl, A. (2011). The impact of discrepancies from ingroup norms on group members’ wellbeing and motivation. European Journal of Social Psychology, 41, 886–897. https://doi.org/10.1002/ejsp.833 Sortheix, F. M., & Lönnqvist, J. E. (2015). Person-group value congruence and subjective well-being in students from Argentina, Bulgaria and Finland: The role of interpersonal relationships. Journal of Community & Applied Social Psychology, 25, 34–48. https://doi.org/10.1002/casp.2193 Statistisches Bundesamt. (2012). Zahl der Studierenden in Deutschland auf Rekordniveau [Record number of students in Germany]. Wiesbaden, Germany. Retrieved from https://www. destatis.de/DE/PresseService/Presse/Pressekonferenzen/ 2012/hochschulen/pm_hochschule_PDF.pdf? __blob=publicationFile Stephens, N. M., Brannon, T. N., Markus, H. R., & Nelson, J. E. (2015). Feeling at home in college: Fortifying school-relevant selves to reduce social class disparities in higher education. Social Issues and Policy Review, 9(1), 1–24. https://do.org/ 10.1111/sipr.12008 Stephens, N. M., Fryberg, S. A., Markus, H. R., Johnson, C. S., & Covarrubias, R. (2012). Unseen disadvantage: How American universities’ focus on independence undermines the academic performance of first-generation college students. Journal of

Social Psychology (2018), 49(1), 16–28

Personality and Social Psychology, 102, 1178–1197. https://doi. org/10.1037/a0027143 Stephens, N. M., Townsend, S. S., Markus, H. R., & Phillips, L. T. (2012). A cultural mismatch: Independent cultural norms produce greater increases in cortisol and more negative emotions among first-generation college students. Journal of Experimental Social Psychology, 48, 1389–1393. https://doi. org/10.1016/j.jesp.2012.07.008 Tay, L., & Diener, E. (2011). Needs and subjective well-being around the world. Journal of Personality and Social Psychology, 101, 354–365. https://doi.org/10.1037/a0023779 Tweed, R. G., & Lehman, D. R. (2002). Learning considered within a cultural context: Confucian and Socratic approaches. American Psychologist, 57, 89. https://doi.org/10.1037/0003066x.57.2.89 Vallerand, R. J., Fortier, M. S., & Guay, F. (1997). Self-determination and persistence in a real-life setting: Toward a motivational model of high school dropout. Journal of Personality and Social Psychology, 72, 1161–1176. https://doi.org/10.1037/0022-3514. 72.5.1161 Vallerand, R. J., Pelletier, L. G., Blais, M. R., Briere, N. M., Senecal, C., & Vallieres, E. F. (1992). The Academic Motivation Scale: A measure of intrinsic, extrinsic, and amotivation in education. Educational and Psychological Measurement, 52, 1003–1017. https://doi.org/10.1177/0013164492052004025 Verquer, M. L., Beehr, T. A., & Wagner, S. H. (2003). A metaanalysis of relations between person-organization fit and work attitudes. Journal of Vocational Behavior, 63, 473–489. https:// doi.org/10.1016/S0001-8791(02)00036-2

Social Psychology (2018), 49(1), 16–28

M. Suhlmann et al., Fitting-In to Succeed at University

Walton, G. M., & Cohen, G. L. (2007). A question of belonging: Race, social fit, and achievement. Journal of Personality and Social Psychology, 92, 82–96. https://doi.org/10.1037/00223514.92.1.82 Walton, G. M., & Cohen, G. L. (2011). A brief social-belonging intervention improves academic and health outcomes of minority students. Science, 331, 1447–1451. https://doi.org/ 10.1126/science.1198364 Walton, G. M., Cohen, G. L., Cwir, D., & Spencer, S. J. (2012). Mere belonging: The power of social connections. Journal of Personality and Social Psychology, 102, 513–532. https://doi.org/ 10.1037/a0025731 Yeager, D., Walton, G., & Cohen, G. L. (2013). Addressing achievement gaps with psychological interventions. Phi Delta Kappa, 94, 62–65. https://doi.org/10.1177/003172171309400514 Received December 21, 2017 Revision received August 23, 2017 Accepted August 30, 2017 Published online February 7, 2018 Michèle Suhlmann University of Tübingen LEAD Graduate School Gartenstraße 29a 72074 Tübingen Germany michele.suhlmann@uni-tuebingen.de

Ó 2018 Hogrefe Publishing

Cultural diversity – challenge and opportunity “It’s a book that we were all waiting for, and will be useful not only to psychologist practitioners and students, but also to stakeholders and policy makers in education.” Bruna Zani, Professor of Social and Community Psychology, Department of Psychology, Alma Mater Studiorum-University of Bologna, Bologna, Italy; EFPA Executive Council Member

Alexander Thomas (Editor)

Cultural and Ethnic Diversity How European Psychologists Can Meet the Challenges 2018, x + 222 pp. US $56.00 / € 44.95 ISBN 978-0-88937-490-4 Also available as eBook Culture and diversity are both challenge and opportunity. This volume looks at what psychologists are and can be doing to help society meet the challenges and grasp the opportunities in education, at work, and in clinical practice. The increasingly international and globalized nature of modern societies means that psychologists in particular face new challenges and have new opportunities in all areas of practice and research. The contributions from leading European experts cover relevant intercultural issues and topics in areas as di-

www.hogrefe.com

verse as personality, education and training, work and organizational psychology, clinical and counselling psychlogy, migration and international youth exchanges. As well as looking at the new challenges and opportunities that psychologists face in dealing with people from increasingly varied cultural backgrounds, perhaps more importantly they also explain and discuss how psychologists can deepen and acquire the intercultural competencies that are now needed in our professional lives.

Use movies to learn about positive psychology “This is the most important book about movies of our times.” Frank Farley, PhD, L. H. Carnell Professor, Temple University, Philadelphia, Former President of the American Psychological Association (APA)

Ryan M. Niemiec / Danny Wedding

Positive Psychology at the Movies

Using Films to Build Character Strengths and Well-Being 2nd edition 2014, xvi + 486 pp. US $59.00 / € 41.95 ISBN 978-0-88937-443-0 Also available as eBook Positive psychology is regarded as one of the most important developments in the field of psychology over the past century. This inspiring book uses movies as a medium for learning about the latest research and concepts, such as mindfulness, resilience, meaning, positive relationships, achievement, well-being, as well as the 24 character strengths laid out by the VIA Institute of Character. Films offer myriad examples of character strengths and other positive psychology concepts and are uniquely suited to learning about them and inspiring new ways of thinking. This book systematically discusses each of the 24 character strengths, balancing film discussion, related psychological research, and practical applications. Each chapter outlines Key Concepts, Relevant Research, an Exemplar from a key movie, Overuse/Underuse, Key

www.hogrefe.com

Enablers and Inhibitors, Practical Applications, International Cinema, and a Summary. Watching the films recommended in this book will help the reader to practice the skill of strengths-spotting in themselves and others, inspiring self-improvement. Practical resources include a suggested syllabus for a complete positive psychology course based on movies, a list of suitable movies for children, adolescents, and families as well as a list of questions for classroom and therapy discussions. Positive Psychology at the Movies is conceived for educators, students, practitioners, and researchers, but anyone who loves movies and wants to change their lives for the better will find it inspiring and relevant. Read this book to learn more about positive psychology – and watch these films to become a stronger person!

Original Article

Taking Priming to Task Variations in Stereotype Priming Effects Across Participant Task Katherine R. G. White,1 Rose H. Danek,2 David R. Herring,3 Jennifer H. Taylor,4 and Stephen L. Crites5 1

Department of Psychology, Kennesaw State University, GA, USA

Department of Psychology, Lyon College, Batesville, AR, USA

Department of Psychology, Pennsylvania State University at Erie, PA, USA

Department of Psychology, Lock Haven University, PA, USA

Department of Psychology, University of Texas at El Paso, TX, USA

Abstract: The current research examined potential moderators of gender and racial stereotype priming in sequential priming paradigms. Results from five experiments suggest that stereotype priming effects are more consistent in tasks that elicit both semantic priming and response competition (i.e., response priming paradigms) rather than tasks that evoke semantic priming alone (i.e., semantic priming paradigms). Recommendations for future stereotype priming research and the implication of these results for the proper interpretation of stereotype priming effects are discussed. Keywords: stereotypes, priming, participant task, gender, race

Priming occurs when a stimulus or context influences subsequent cognitions and behavior. Priming has expanded from initial research on cognition and memory (e.g., Schacter, Dobbins, & Schnyer, 2004) to topics such as stereotyping and attitudes (see Wittenbrink, 2007). A popular paradigm for studying stereotype priming is the sequential priming paradigm. In a typical sequential priming paradigm (see McNamara, 2005; Neely, 1977), participants are presented trials containing two stimuli – a prime and target – and prime-target congruency is varied across trials. For example, a stereotypically congruent trial might be the prime “WOMEN” followed by the target “NURSE.” A stereotypically incongruent trial might be the prime “MEN” followed by the target “NURSE.” A stereotype priming effect occurs when participants’ responses to targets are faster and more accurate on stereotypically congruent than incongruent trials. While researchers have identified numerous variables that moderate this effect (for review, see Blair, 2002), some inconsistencies remain unexplained. This paper outlines a set of studies (conducted over an eight-year period) which attempted to identify variables that might explain inconsistent findings in sequential stereotype priming. Our investigation began with research that explored whether the N400 event-related potential (ERP) component Ó 2018 Hogrefe Publishing

could measure stereotype priming (White, Crites, Taylor, & Corral, 2009). In this study, gender categories “MEN” and “WOMEN” served as primes and participants were asked to indicate whether each target word (e.g., PURSE, MUSCLES) matched or did not match the prime according to common gender stereotypes. As expected, response times were slower and N400 amplitude was larger when targets did NOT match the preceding prime. In this initial study, we intentionally used an explicit decision task to enhance the N400 effect (Chwilla & Kolk, 2005). However, people do not typically make explicit stereotype judgments when they encounter other individuals, which makes a matching task less theoretically interesting than tasks that more directly assess “automatic” stereotype activation. We therefore performed a follow-up study in which participants indicated whether each target was a word or nonword (i.e., lexical decision task or LDT). Evidence of priming with this type of task would suggest that stereotype priming occurs even when people are engaged in a task that is irrelevant to the stereotype dimension under investigation. However, stereotype priming failed to manifest for both response times and the N400 (White & Crites, 2009). We then performed an additional small pilot study (just focused on response times), and this study again found no evidence of stereotype priming. Neither of these LDT studies are reported in detail here Social Psychology (2018), 49(1), 29–46 https://doi.org/10.1027/1864-9335/a000326

because they were potentially underpowered, but their results prompted us to a focus on participant task as a potential moderator. Participants can perform many different judgment tasks in a sequential priming paradigm. Some tasks used in previous stereotype priming research include the LDT, Stroop-pronunciation task, semantic classification, and stereotype classification task (SCT). The LDT requires participants to categorize each target as a word (e.g., NURSE) or nonword (e.g., NIRSE). In the pronunciation task, participants must simply read the target word out loud. Semantic classification tasks require participants to categorize targets into one of two semantic categories (e.g., person/place, noun/pronoun) that are presumably irrelevant to the stereotype dimension being examined. The SCT requires participants to categorize target stimuli along a stereotype-relevant dimension (e.g., categorize targets as male or female when gender stereotypes are under investigation, or black-white for race stereotypes). We broadly divided these various tasks into two categories – those that require a stereotype-relevant judgment about the target under investigation (SCT) and those that require a stereotype-irrelevant judgment about the target (LDT, pronunciation, semantic classification). An interesting pattern seems to emerge when examining the judgment tasks used in previous stereotype priming studies (see Table 1). Significant stereotype priming effects are consistently observed when a stereotype-relevant classification task is used, but are less consistent when a stereotype-irrelevant task is used. Notably, no study has closely examined the effect of participant task on the stereotype priming effect. One exception is the article by Banaji and Hardin (1996), which observed stronger gender stereotype priming when participants categorized targets as male/female (i.e., an SCT) versus pronoun/not pronoun (i.e., a semantic classification). Banaji and Hardin (1996) state that this finding requires further exploration in future research, but to our knowledge, this has not been done. In this article, we report results from five experiments which systematically examine the potential impact of participant task on the stereotype priming effect. These studies took place over 8 years, during which time theoretical explanations that closely align with our pattern of findings were also being articulated (Wentura & Degner, 2010; Wentura & Rothermund, 2014). We explore these theoretical explanations in the discussion and here present the five 1

Katherine R. G. White et al., Task & Stereotype Priming

experiments we conducted in an effort to better understand the inconsistencies observed in our own and others’ research.

Study 1 This study examines the impact of participant task on gender stereotype priming.1 Participants were randomly assigned to either a stereotype classification task (SCT condition) or a lexical decision task (LDT condition).2 Participants in the SCT condition categorized targets as more associated with men or women. Participants in the LDT condition categorized targets as words or nonwords. We also randomly assigned participants to a third task condition – the pre-primed LDT condition. In this condition, participants completed an LDT after engaging in a short task to make gender more salient/accessible (see Wittenbrink et al., 1997 for similar procedure). We included this pre-primed LDT condition to explore the possibility that stereotype priming can be boosted by making stereotypes temporarily more accessible (Higgins & King, 1981). We predicted that gender stereotype priming would be significant in both the SCT and pre-primed LDT conditions, but stronger in the SCT condition.

Method Participants The final sample included 171 participants: 52 in gender categorization, 59 in classic LDT, and 60 in pre-primed LDT. A total of 199 (146 female) undergraduates participated for partial course credit; however, data from 28 participants were not available for analyses or excluded from the primary analyses.3 In the entire sample, 169 participants self-identified as Hispanic/Latino. Ages ranged from 18 to 46 years (M = 19.0). Stimuli Trials consisted of a prime followed by a target. Trials were either a stereotype congruent pair (e.g., STEPHANIE: GOSSIPY), a stereotype incongruent pair (e.g., JOSEPH: GOSSIPY), or a pair with a nonword target (e.g., MARIA: EEBS). Nonword targets only appeared in the LDT and pre-primed LDT conditions.

As discussed above, this research began in 2009. To help place these studies in context, we report here when each was initiated and completed. Study 1 was initiated and completed in the fall of 2009. Study 2 took place in the spring of 2011 and Study 3 in the spring of 2013. Finally, Studies 4 and 5 were initiated in the spring of 2015 and completed in the spring of 2016. We used an LDT as an exemplar of a stereotype-irrelevant task because it is more commonly used in stereotype priming research and to maintain consistency with our previous LDT studies which failed to produce significant stereotype priming. See Appendix B for discussion of how (1) data were processed prior to analyses, (2) this led to data from participants being excluded, and (3) additional analyses that verify those reported in the text.

Social Psychology (2018), 49(1), 29–46

Ó 2018 Hogrefe Publishing

Katherine R. G. White et al., Task & Stereotype Priming

Table 1. Stereotype-relevant and stereotype-irrelevant tasks and results, in chronological order Study

Task

Stereotype

Significant

Yes

Stereotype-relevant tasks Banaji & Hardin, 1996, Study 1

SCT

Gender

Blair & Banaji, 1996

SCT

Gender

Yes

Kawakami & Dovidio, 2001, Study 1

SCT

Gender

Yes

Macrae, Mitchell, & Pendry, 2002, Study 2

SCT

Gender

Yes

Castelli, Macrae, Zogmaister, & Arcuri, 2004, Studies 1 and 2

SCT

Gender

Yes

Cacciari & Padovani, 2007

SCT

Gender

Yes

Macrae & Martin, 2007

SCT

Gender

Yes

Kimura et al., 2009

SCT

Gender

Yes

Macrae & Cloutier, 2009

SCT

Gender

Yes

Müller & Rothermund, 2014

SCT

Gender

Yes

Plaza, Boiché, Brunel, & Ruchaud, 2017

SCT

Gender

Yes

Kawakami, Dovidio, Moll, Hermsen, & Russin, 2000, Study 3

SCT

Race

Yes

Kawakami & Dovidio, 2001, Study 2

SCT

Race

Yes

Stewart, Latu, Kawakami, & Myers, 2010

SCT

Race

Yes

Semantic

Gender

Karylowski and colleagues, 2001

Semantic

Gender

Marginal

Macrae, Hood, Milne, Rowe, & Mason, 2002, Study 2

LDT

Gender

Yes

White & Crites, 2009

LDT

Gender

Habibi & Khurana, 2012

LDT

Gender

Yes

Müller & Rothermund, 2014

Semantic

Gender

Wittenbrink, Judd, & Park, 1997

LDT

Race

Yes

Kawakami, Dion, & Dovidio, 1999

Pronunciation

Race

Stereotype-irrelevant tasks Banaji & Hardin, 1996, Study 2

Kawakami et al., 2000, Studies 1 and 2

Pronunciation

Race

Yes & No

Kawakami & Dovidio, 2001, Study 3

Semantic

Race

Yes

Sassenberg & Moskowitz, 2005

LDT

Race

Yes

Bartholow, Dickter, & Sestir, 2006, Studies 1 and 2

Semantic

Race

Yes

Clow & Esses, 2007

LDT

Race

Yes

Freng & Willis-Esqueda, 2011

LDT

Race

Yes

Moskowitz & Li, 2011

LDT

Race

Yes

Bean, Stone, Moskowitz, Badger, & Focella, 2013

LDT

Race

Yes

Hehman, Volpert, & Simons, 2014

Evaluation

Race

Bessenoff & Sherman, 2000

LDT

Weight

Casper, Rothermund, & Wentura, 2010

LDT

Multiple

Yes

Verhaeghen, Aikman, & Van Gulick, 2011

LDT, Evaluation, Relatedness

Multiple

Yes

Rohmer & Louvet, 2012

LDT

Disability

Yes

Notes. Studies were excluded from our review if they, (1) did not use a sequential priming paradigm (e.g., IAT), (2) did not report the results of the priming effect, (3) did not include a true control condition if examining a moderator variable other than task (e.g., priming reduction), and/or (4) used a weapons identification or shoot-don’t shoot task with a strict response time window.

Primes Primes were 54 male names, 54 female names, 54 pictures of males, and 54 pictures of females. An additional eight names and eight pictures were used in practice trials. Targets Targets were either words associated with gender or nonwords. The gender words consisted of 16 female and

Ó 2018 Hogrefe Publishing

16 male stereotypical traits or nontraits. Stereotype words were matched for valence and length and normed in the local population (see Appendix A for a complete list of stimuli and norming results). Twenty-four additional words (12 male and 12 female) associated with gender were used in practice trials. Nonwords were 56 pronounceable nonwords from the ARC Nonword Database (Rastle, Harrington, & Coltheart, 2002).

Social Psychology (2018), 49(1), 29–46

Procedure Participants were run in small groups, with each group assigned to one of the three conditions. After informed consent, participants completed a short demographics questionnaire and were told that they would see a series of stimulus pairs on the computer. They were asked to attend to the first stimulus (prime) and respond to the second (target) as quickly and accurately as possible, using designated keys on a computer keyboard. In the SCT condition, targets were always words and participants indicated whether each was associated with men or women – half were stereotypically congruent with the prime and half were stereotypically incongruent. In the LDT and pre-primed LDT conditions, half of the targets were words and half were nonwords. Participants indicated whether each was a word or nonword. For word targets, half were stereotypically congruent with the prime and half were stereotypically incongruent. The pre-priming session for the pre-primed LDT condition consisted of 24 names (12 female and 12 male) and 24 pictures (12 female and 12 male) that participants categorized as male or female. The order of names/pictures was counterbalanced across participants. The 48 stimuli in this pre-priming session were not used in the experimental session. Following the pre-priming session, participants initiated the LDT. For all conditions, trials consisted of a 200 ms focus “+”, a 150 ms prime, a 100 ms blank screen (stimulus-onset asynchrony, SOA = 250 ms), and a target that lasted until the participant responded but no more than 1,050 ms. There was a 1,500 ms intertrial interval (ITI) before the onset of the focus for the next trial. All stimuli were presented on a black background. Verbal stimuli were white capital letters. Trials were organized into seven blocks. The first consisted of 48 practice trials, after which participants could ask the experimenter if they had questions. The six experimental blocks each began with eight practice trials (i.e., warm-up trials; Wentura & Degner, 2010) followed by 96 experimental trials.4 Three experimental blocks used picture primes and three used name primes.5 The picture and name blocks alternated and were counterbalanced across participants. Following the priming procedure, all participants rated the masculinity/femininity (1 = very 4

Katherine R. G. White et al., Task & Stereotype Priming

masculine, 7 = very feminine) of all targets. Male target stimuli were perceived by participants as significantly more masculine (MM = 1.97) than female target stimuli (MF = 1.91), t = 20.92, p < .001.

Results and Discussion Response times from correct responses were analyzed with a 3 (Condition: gender categorization, LDT, or pre-primed LDT) 2 (Congruency: stereotype congruent vs. incongruent) mixed factorial analysis of variance (ANOVA) with Condition between-subjects.6 There were main effects for both Congruency, F(1, 168) = 66.75, p < .001, η2 = .284, 90% CI [.192, .369], and Condition, F(2, 168) = 10.18, p < .001, η2 = .108, 90% CI [.040, .179]. These were qualified by a significant Condition by Congruency interaction, F(2, 168) = 37.20, p < .001, η2 = .307, 90% CI [.209, .387]. As expected, stereotype priming was significant in the SCT condition as people responded faster to congruent (M = 603 ms; SD = 55.3) than incongruent (M = 623 ms; SD = 54.0) targets, F (1, 51) = 76.99, p < .001, η2 = .602, 90% CI [.449, .692]. Stereotype priming was nonsignificant in the LDT (Ms = 563 and 565 ms; SDs = 60.9 and 62.5 for congruent and incongruent, respectively), F (1, 58) = 2.33, p = .132, η2 = .039, 90% CI [.00, .144], or pre-primed LDT conditions (Ms = 570 and 571 ms; SDs = 68.6 and 68.4 for congruent and incongruent, respectively), F(1, 59) = 1.45, p = .233, η2 = .024, 90% CI [.00, .118]. Results provide initial evidence that participant task is a significant moderator of the stereotype priming effect and that increasing the salience of the stereotype dimension (via pre-priming) does not lead to greater stereotype priming in stereotype-irrelevant tasks. Specifically, gender stereotype priming was significant when participants categorized targets as male/female (i.e., SCT) but not when participants categorized targets as word/nonword (i.e., LDT) in either the traditional or pre-primed LDTs. We speculated that past inconsistencies in stereotype-irrelevant tasks may have been due to extraneous factors that lead to stereotypes being more salient in some studies (which found priming) and less salient in others (which did not find priming). Thus, we attempted to increase stereotype salience in the pre-primed LDT to test this idea

The eight practice trials proceeding each experimental block were not separated from the experimental trials and thus appeared to be experimental trials from the participants’ perspective. They were used to get participants into the flow of the experiment after a break. There were eight trials in order to present all four trial types (e.g., male-male; male-female; female-female, female-male) and an equivalent number of non-word trials for the LDT tasks. The SCT task used two presentations of the four trial types to maintain an equivalent number of practice trials. Picture versus name primes were used to explore whether this impacted magnitude of priming. Initial analyses suggested that it did not; so it will not be discussed further. Because the SCT does not have non-word trials, it has twice the number of available trials for analyses as an LDT. Thus, we examined the SCT using all trials and half of the trials (interspersed through the entire range of the task to be comparable with the word trials found in an LDT). There was no difference in the pattern of significance in this Study or Studies 3 and 5, which also compared SCT to LDT. Thus, in all three studies, we report analyses that include all SCT trials.

Social Psychology (2018), 49(1), 29–46

Ó 2018 Hogrefe Publishing

Katherine R. G. White et al., Task & Stereotype Priming

(i.e., by comparing priming in traditional vs. pre-primed LDTs). It is possible that the pre-priming procedure used in Study 1 was not strong enough to boost the salience of gender across the entire LDT. We therefore explored a different method of increasing stereotype salience in Study 2.

Study 2 An advantage of stereotype classification tasks is that they focus participants on the stereotype dimension throughout the entire task. One limitation of Study 1’s pre-priming procedure is that any increase in the salience of gender may have quickly dissipated. Recent research suggests that evaluative priming may only be observed in conditions where attention is directed to the evaluative dimension, referred to as the attention allocation perspective (Spruyt, De Houwer, Everaert, & Hermans, 2012; Spruyt, De Houwer, & Hermans, 2009). This type of attention moderation may also occur for stereotype priming. In Study 2, we adapted procedures previously used to test the attention allocation perspective to design a modified LDT that would make gender salient throughout the LDT (see Spruyt, De Houwer, Hermans, & Eelen, 2007, Study 2 for similar procedure). Participants were randomly assigned to either count the number of male and female primes throughout the LDT (gender tally condition) or the number of Hispanic and white primes throughout the LDT (ethnicity tally condition). We predicted that the increased salience of gender in the gender tally condition would be sufficient to create gender stereotype priming in the LDT. This would provide one potential explanation for prior inconsistencies in stereotype priming results while simultaneously adding support to the attention allocation perspective.

Method Participants The final sample included data from 121 participants: 53 in the ethnicity tally and 68 in the gender tally. A total of 149 (90 female) undergraduates participated for partial course credit; however, data from 28 participants were not available for analyses or excluded from the primary analyses (see Appendix B). One hundred thirty-two participants self-identified as Hispanic/Latino. Ages ranged from 18 to 45 years (M = 21.0). Stimuli Trials consisted of a prime followed by a target and could form a stereotype congruent pair, stereotype incongruent Ó 2018 Hogrefe Publishing

pair, or a pair with a nonword target. The prime names (54 male and 54 female), target words (16 male and 16 female), target nonwords, and practice stimuli were the same as those used in Study 1. Unlike Study 1, picture primes were not used. Procedure Participants were run in small groups, with each group assigned to one of the two tally conditions. Following a demographic questionnaire, participants were told that the experiment consisted of a series of trials in which they would see a name followed quickly by a word or nonword. The experimenter explained that they would have slightly different tasks across three phases of the experiment. Phase 1 Phase 1 was a practice phase to introduce participants to the LDT. All participants categorized each target as a word or nonword as quickly and accurately as possible, using two keys on a computer keyboard. The nature and timing of this LDT was identical to that in Study 1 except that there were two blocks of trials that each contained 16 trials. Phase 2 Phase 2 was also a practice phase. Participants’ primary task was to categorize each target as a word or nonword as quickly and accurately as possible. Their secondary task was to mentally tally either the number of male and female names that appeared (gender tally) or the number of Hispanic and white names that appeared (ethnicity tally) and report this number at the end of each block. The experimenter stressed that the lexical decision task was the primary task and the mental tally was secondary. The timing of each trial was identical to those in Phase 1 except that after a response to a target, a screen reminded participants to update their mental tally of male and female (or Hispanic and white) names. Phase 2 consisted of six blocks of 16 trials, and participants entered their tally at the end of each block. Phase 3 Phase 3 contained the critical experimental trials and was similar to Phase 2 except there was no reminder to update the mental tally after each trial. A 3,000 ms ITI was used between trials. Phase 3 consisted of 18 experimental blocks of 16 trials each. At the end of each block, participants reported their tally from the previous block. Experimental blocks were created with varying numbers of male/female (or Hispanic/white) primes. Of the 18 blocks, 6 contained 8 male and 8 female names, 2 contained 9 male and 7 female names, 2 contained 6 male and 10 female names, and so forth. Across all 18 blocks, the number of trials with male primes was the same as the number of trials with female primes (144 trials each). Social Psychology (2018), 49(1), 29–46

Results and Discussion Response times from Phase 3 correct responses were analyzed with a 2 (Tally: gender vs. ethnicity) 2 (Congruency: stereotype congruent vs. incongruent) mixed factorial ANOVA with Condition between-subjects. Neither the main effect of Congruency (Ms = 687 and 690 ms; SDs = 83.9 and 83.6 for congruent and incongruent, respectively), F(1, 119) = 2.02, p = .158, η2 = .017, 90% CI [.00, .072], nor the interaction between Tally and Congruency, F(1, 119) = 0.01, p = .933, η2 < .001, 90% CI [.00, .004], was significant. Because we predicted priming in the gender tally condition, we performed simple effect analyses to examine congruency in each condition. The congruency effect was nonsignificant in both the ethnicity tally condition, F(1, 52) = 0.78, p = 0.38, η2 = .015, 90% CI [.00, .106], and the gender tally condition, F (1, 67) = 1.31, p = .256, η2 = .019, 90% CI [.00, .102] (see Appendix B for supplementary analyses). Gender stereotype priming was absent in both the gender and ethnicity tally conditions. Thus, increased attention to the stereotype dimension of interest throughout the LDT was not sufficient to elicit gender stereotype priming. When combined with Study 1’s results, this suggests that gender stereotype priming effects are difficult to detect when participants perform an LDT. Study 3 examined whether this is true for another commonly researched stereotype dimension – race.

Study 3 Results from Studies 1 and 2 suggest that stereotype priming effects are difficult to capture when participants perform an LDT and also suggest that attention allocation may not adequately explain inconsistencies in previous stereotype priming research. However, both studies examined gender stereotypes. Study 3 examined whether these results generalize to race stereotype priming. As seen in Table 1, half (50%) of the published studies on gender stereotypes that used a stereotype-irrelevant task (i.e., LDT, pronunciation, or semantic classification) report nonsignificant results. This is greater than the proportion of null results reported for race stereotypes ( 21%). This suggests that race, relative to gender, stereotype

7 8

Katherine R. G. White et al., Task & Stereotype Priming

priming may be more easily observed in stereotype-irrelevant tasks. Study 3 examined race stereotypes and randomly assigned participants to complete either a lexical decision (LDT)7 or stereotype classification (SCT) condition. Given the previous pattern of results (see Table 1), we expected to observe significant race stereotype priming in the SCT condition as well as significant, albeit weaker, race stereotype priming in the LDT condition.

Method Participants The final sample included 119 participants: 76 in the LDT and 43 in the SCT. A total of 150 undergraduates participated; however, data from 31 participants were not available for analyses or excluded from the primary analyses (see Appendix B).8 One hundred thirty-two participants self-identified as Hispanic/Latino. Ages ranged from 18 to 45 years (M = 21.0). Stimuli Trials consisted of a prime stimulus followed by a target stimulus and could form a stereotype congruent pair (e.g., HISPANIC PICTURE: SOCCER), stereotype incongruent pair (e.g., BLACK PICTURE: SOCCER), or a pair with a nonword target (e.g., HISPANIC PICTURE: SPOMPH). Primes Prime stimuli consisted of 80 pictures – 40 of Hispanics (20 male and 20 female) and 40 of Blacks (20 male and 20 female). An additional 20 pictures (10 Hispanic and 10 Black) were used for practice trials. Targets Targets were 20 words stereotypically associated with blacks (e.g., basketball, trendy), 20 words stereotypically associated with Hispanics (e.g., agriculture, immigration), and 80 pronounceable nonwords obtained from the ARC Nonword Database generator (Rastle et al., 2002; see Appendix A for complete list). An additional 12 words of each type were used in practice trials. Procedure The procedure was identical to Study 1, which also had both LDT and SCT conditions, except for the following

We continued to use an LDT for consistency with our previous studies. In the initial computer program, all target words were in “CAPS” but non-words were in “lowercase.” This problem, which only affected the LDT condition, was discovered after data from 17 participants were collected in the LDT condition. We collected additional participants in the LDT condition because we anticipated having to remove these 17 participants. That is, this CAP versus lowercase confound meant that once participants realized this difference, they would not have to read the target stimuli and could respond word/non-word based solely on the form (i.e., CAP = “word” vs. lowercase = “non-word”). The pattern of results and significance were identical when including or not including these participants. So, we report analyses that include these 17 participants.

Social Psychology (2018), 49(1), 29–46

Ó 2018 Hogrefe Publishing

Katherine R. G. White et al., Task & Stereotype Priming

differences. First, there was no pre-primed LDT condition. Second, only picture primes were used. Third, in the SCT, participants indicated whether each target was more associated with blacks (e.g., BASKETBALL) or Hispanics (e.g., SOCCER), rather than men or women. Fourth, each trial consisted of a 650 ms focus “+”, a 50 ms blank screen, a 200 ms prime, another 50 ms blank screen, and a target that lasted until the participant responded but no more than 1,050 ms. Fifth, trials were organized into eight blocks. The first consisted of 24 practice trials. The following seven experimental blocks each began with four practice trials followed by 80 experimental trials.9

Study 4 Studies 1 and 3 provide evidence of robust stereotype priming when participants perform an SCT. Evidence for stereotype priming when participants perform an LDT, however, was observed for racial stereotypes but not for gender stereotypes. Study 4 directly tested stereotype dimension as a moderator of stereotype priming when participants perform an LDT. All participants performed an LDT but were randomly assigned to either a race stereotypes or gender stereotypes condition. We hypothesized that significant stereotype priming effects would emerge for race but not for gender.

Results and Discussion Response times from correct trials were analyzed with a 2 (Condition: LDT vs. SCT) 2 (Congruency: stereotype congruent vs. incongruent) mixed factorial ANOVA with Condition between-subjects (see Appendix B). This analysis revealed the expected Congruency main effect, F(1, 109) = 49.67, p < .001, η2 = .313, 90% CI [.197, .414], and the Congruency by Condition interaction, F(1, 109) = 30.70, p < .001, η2 = .219, 90% CI [.114, .324]. Follow-up analyses revealed a significant congruency effect in the SCT condition, F (1, 42) = 27.92, p < .001, η2 = .399, 90% CI [.204, .538]. Participants responded faster to congruent (M = 644 ms; SD = 55.8) than incongruent (M = 668 ms; SD = 52.9) targets. Participants in the LDT condition also responded more quickly to congruent (M = 677 ms; SD = 58.0) than incongruent (M = 682 ms; SD = 58.0) targets, F(1, 67) = 10.20, p = .002, η2 = .132, 90% CI [.031, .257]. The difference between incongruent and congruent trials in the SCT (M = 23.6 ms; SD = 29.3) is significantly greater than that in the LDT (M = 5.1 ms; SD = 13.1), F(1, 109) = 20.70, p < .001, η2 = .160, 90% CI [.067, .261].10 Matching our predictions, race stereotype priming was observed when participants completed a lexical decision task. Additionally, priming in the LDT condition was weaker than in the SCT condition. This strengthens the assertion that stereotype priming effects are robust when participants perform an SCT. When combined with results from Studies 1 and 2, the results suggest that stereotype priming effects are less consistent when participants perform an LDT. They also raise the possibility that stereotype dimension moderates the stereotype priming effect. Studies 4 and 5 simultaneously investigated this possibility.

Method Participants The final sample included 165 participants: 87 in the race stereotypes condition and 78 in the gender stereotypes condition. A total of 178 undergraduates participated; however, data from 13 participants were not available for analyses or excluded from the primary analyses (see Appendix B). One hundred fifty-six self-identified as Hispanic/Latino. Ages ranged from 18 to 53 years (M = 20.8). Stimuli Trials consisted of a prime followed by a target and were either a stereotype congruent pair, a stereotype incongruent pair, or a pair with a nonword target. Primes The pictures used for experimental (20 black females, 20 Hispanic females, 20 black males, and 20 Hispanic males) and practice primes were the same as those in Study 3. Targets Targets were a set of 80 words and 80 pronounceable nonwords from the ARC Nonword Database (Rastle et al., 2002). Words consisted of 20 words from each of the four stereotype categories (female, male, Black, and Hispanic). In the race condition, the target words were 20 black and 20 Hispanic stereotype words. In the gender condition, the target words were 20 female and 20 male stereotype words. Words were matched for valence and length and normed in the local population (see Appendix A).

As in Study 1, we examined the SCT using all trials and half of the trials and found no difference in the pattern of significance in these two analyses. We report the analysis that includes all SCT trials. 10 When we analyzed data from 137 participants (of the 150 total) who had data in all experimental conditions, the global analyses revealed the same Congruency main effect and Condition by Congruency interaction. The congruency effect for the SCT remained significant, but the effect for the LDT was not, F(1, 74) = 3.08, p = .083, η2 = .040. Ó 2018 Hogrefe Publishing

Social Psychology (2018), 49(1), 29–46

Procedure The procedure was identical to the LDT task used in Study 3 except for the following differences. First, participants were assigned to either the gender or race condition. In the gender condition, half the word trials were stereotype congruent (e.g., male picture: male stereotype word) and half were stereotype incongruent (e.g., male picture: female stereotype word). In the race condition, half the word trials were stereotype congruent (e.g., black picture: black stereotype word) and half were stereotype incongruent (e.g., black picture: Hispanic stereotype word). Second, the ITI was 1,200 ms not 1,500 ms. Third, trials were organized into 7 blocks that included a block of 24 practice trials and 6 experimental blocks each began with 4 practice trials followed by 80 experimental trials.

Results and Discussion Response times from correct responses were analyzed with a 2 (Condition: Race vs. Gender) 2 (Congruency: stereotype congruent vs. incongruent) mixed factorial ANOVA with Condition between-subjects. This revealed nonsignificant effects for both the Congruency main effect (Ms = 574 and 575 ms; SDs = 56.2 and 56.1 for congruent and incongruent, respectively), F(1, 163) = 0.24, p = .626, η2 = .002, 90% CI [.00, .026], and the Condition by Congruency interaction, F(1, 163) = 0.51, p = .477, η2 = .003, 90% CI [.00, .032]. Because we predicted a significant finding for race and not gender, we conducted follow-up analyses. These analyses revealed no significant priming for either race (Ms = 582 and 583 ms; SDs = 51.3 and 51.5 for congruent and incongruent, respectively), F(1, 86) = 0.62, p = .433, η2 = .007, 90% CI [.00, .063], or gender (Ms = 566 and 566 ms; SDs = 60.5 and 60.0 for congruent and incongruent, respectively), F(1, 77) = 0.32, p = .858, η2 = .004, 90% CI [.00, .057]. Contrary to predictions and results from Study 3, stereotype priming failed to reach significance for both race and gender stereotypes. This once again underscores the inconsistent nature of stereotype priming effects when an LDT is used. While Study 4 was being conducted at a university in the Southwest that has a predominantly Hispanic student demographic, we were simultaneously attempting to replicate the findings of Study 3 at another university in the Southeast with a substantial black population.

Study 5 In this study, we manipulated both participant task (SCT vs. LDT) and stereotype dimension (race vs. gender). We used both an SCT and LDT in this study because it was Social Psychology (2018), 49(1), 29–46

Katherine R. G. White et al., Task & Stereotype Priming

conducted at a new university with different participant demographics. Participants were randomly assigned to one of four conditions – Race-LDT, Race-SCT, GenderLDT, or Gender-SCT. We predicted that stereotype priming would be significant in both SCT conditions, significant but weaker in the Race-LDT condition, and absent in the Gender-LDT condition.

Method Participants The final sample included 217 participants (76.1% female) from a midsized Southeastern university (M = 21.84 years): 60 in the Race-LDT condition, 46 in the Race-SCT condition, 54 in the Gender-LDT condition, and 57 in the Gender-SCT condition. A total of 238 undergraduates participated; however, data from 21 participants were not available for analyses or excluded from the primary analyses (see Appendix B). Forty-one percent self-identified as White/Caucasian, 43.2% as Black/African American, 7.7% as Hispanic/Latino, and 9.9% as Pacific Islander, American Indian, Asian, multiracial, or said they preferred not to answer. Stimuli Trials consisted of a prime stimulus followed by a target stimulus. For participants assigned to one of the two Race conditions, trials were race stereotype-congruent, race stereotype-incongruent, or had a nonword target. Participants assigned to one of the two Gender conditions saw trials that were gender stereotype-congruent, gender stereotype-incongruent, or had a nonword target. Nonword targets only appeared in the LDT conditions. Primes Prime stimuli consisted of 120 pictures – 30 black females, 30 black males, 30 white females, and 30 white males. Pictures were from the Park Aging Mind Laboratory (Minear & Park, 2004) or were used courtesy of Michael J. Tarr (Righi, Peissig, & Tarr, 2012). Pictures were headshots on a neutral or white background. Pictures were previously rated as neutral or near neutral in attractiveness. An additional five pictures of each type were used for practice trials only. Targets Targets were words associated with race, gender, or nonwords. Race words consisted of 30 words stereotypically associated with blacks and 30 words stereotypically associated with whites (see Appendix A for complete list). Gender words consisted of 30 words stereotypically associated with men and 30 words stereotypically associated with women. An additional eight words of each type were used for practice trials only. Ó 2018 Hogrefe Publishing

Katherine R. G. White et al., Task & Stereotype Priming

Both race and gender stimuli were matched for valence and normed in the local population (see Appendix A). Nonwords were 60 pronounceable nonwords created by changing two letters in the word stimuli (e.g., BASPETHALL), rather than words from the ARC Nonword Database. Ten additional nonwords were created and used in practice trials. Procedure The procedure was identical to Study 3, which also had both LDT and SCT procedures, except for the following differences. First, participants were randomly assigned to one of the four conditions (Race-LDT, Race-Categorization, Gender-LDT, Gender-Categorization). Participants in the Race-SCT were instructed to indicate whether the target was more associated with blacks or whites. Participants in the Gender-SCT were instructed to indicate whether the target was more associated with men or women. Second, the target lasted until the participant responded but no more than 1,500 ms and the ITI was 1,200 ms.

Results and Discussion The omnibus 2 (Task: LDT, categorization) 2 (Stereotype Dimension: race, gender) 2 (Congruency: stereotype congruent vs. incongruent) ANOVA revealed significant effects for all main effects and interactions, including a Task Stereotype Dimension Congruency interaction, F(1, 213) = 7.87, p = .005, η2 = .036. This significant three-way interaction suggests that the critical withinsubject congruency effect varies across the four unique between-subject conditions. We probed this by testing the congruency effect in each condition (see Table 2). As expected, participants responded faster to congruent than incongruent targets in the Race-SCT, F(1, 45) = 15.94, p < .001, η2 = .262, 90% CI [.093, .414]. However, there was no evidence of stereotype priming in the RaceLDT, F(1, 59) = 1.56, p = .216, η2 = .026, 90% CI [.00, .121], Gender-SCT, F(1, 56) = 0.16, p = .695, η2 = .003, 90% CI [.00, .062], or the Gender-LDT, F(1, 53) = 0.002, p = .970, η2 < .001, 90% CI [.00, .002]. Thus, the results reveal significant stereotype priming only for race in a stereotype categorization task.

General Discussion Taken as a whole, results from the five experiments indicate that participant task is a significant moderator of the stereotype priming effect. When participants performed a stereotype classification task (SCT), significant stereotype priming was observed with relatively high consistency in three of four within-subject tests across both gender and Ó 2018 Hogrefe Publishing

Table 2. Mean, standard deviation, and p-values for congruent and incongruent trials in Study 5 Congruent Condition

M (SD)

Incongruent M (SD)

Significance (p)

Race – SCT

759 (83.5)

776 (88.2)

Gender – SCT

689 (75.0)

690 (75.6)

Race – LDT

691 (87.1)

694 (88.5)

Gender – LDT

683 (79.8)

683 (79.7)

< .001

race stereotypes (note that Study 5 has two “tests,” one for gender and another for race). By contrast, stereotype priming was largely absent when participants performed a lexical decision task (LDT), reaching significance only once. To further explore this issue, we completed a meta-analysis of the five studies reported here and included one pilot study referenced in the Introduction. Using a random effects model, we found that stereotype priming was significant overall (d = 0.20, SE = 0.065, p = .002, 95% CI [.07, .33]), but there was also significant heterogeneity among the studies, Q(10) = 24.14, p = .007. To test whether the heterogeneity was due to task, we ran a mixed effects model with task as a categorical moderator. The model accounted for 38.33% of the differences in effect sizes between studies, QM(1) = 4.73, p = .030. While stereotype priming was significant for both the stereotype categorization (k = 3, d = 0.42, SE = 0.12, p < .001, 95% CI [.19, .66]) and lexical decision (k = 8, d = 0.13, SE = 0.07, p = .049, 95% CI [.05, .26]) tasks, priming was significantly larger for the SCT than the LDT tasks, w2 = 4.73, p = .030. These results largely mirror those reported in a full-scale meta-analysis conducted by our laboratory (Kidder, White, Hinojos, Sandoval, & Crites, in press), with the exception that the stereotype priming effect for the LDT failed to reach significance in the full meta-analysis. As discussed earlier, we began this line of research in 2009 as a follow-up to a study that demonstrated that the N400 component of the event-related potential (ERP) is associated with stereotype priming with a task conceptually similar to the SCT (White et al., 2009). As we were conducting these studies, other researchers (Wentura & Degner, 2010; Wentura & Rothermund, 2014) were articulating a theoretical framework that can help explain the pattern of results across our studies. Wentura and Degner (2010) divide sequential priming paradigms into two categories – semantic and response priming paradigms. Differentiating between these paradigms is important because different cognitive mechanisms are thought to underlie priming effects observed in these different paradigms (Wentura & Rothermund, 2014). A critical variable that distinguishes these two types of paradigms is the judgment that people make about target stimuli. Sequential priming designs that use an LDT, pronunciation, or semantic classification task fall under the category of Social Psychology (2018), 49(1), 29–46

Katherine R. G. White et al., Task & Stereotype Priming

semantic priming paradigms because priming observed in these paradigms is thought to be driving primarily by the semantic or associative relation between primes and targets.11 Priming observed in semantic priming paradigms is traditionally attributed to spreading activation (Collins & Loftus, 1975) or the activation of overlapping mental representations (e.g., Masson, 1995; Rumelhart & McClelland, 1986) in memory. For example, seeing a female picture prime activates the concept of “female” in memory, which then facilitates responses to other items which share those semantic features. As “nurse” is stereotypically female, the pattern of activation for nurse is facilitated by the preceding concept of “female.” Sequential priming paradigms that use a task like the SCT fall under the category of response priming paradigms because the judgment about targets can also be applied to primes, which may create facilitation/competition for target responses. Priming effects observed in response priming paradigms are attributed to both memory activation and response activation that occurs when a prime activates a response that is either compatible or incompatible with the response required by the target. For example, people in a gender stereotype priming SCT have to respond “female” when they see the target word “nurse.” Seeing a female name as the prime activates the response “female” to that prime, which then makes it easier to respond “female” to the word “nurse.” Alternatively, seeing a male name as the prime activates the response “male,” which interferes with the necessary response to the target word “nurse.” It is important to note that response priming designs may involve BOTH semantic and response priming whereas semantic priming designs involve only semantic priming (although see Footnote 11). Thus, Wentura and Degner (2010) assert that priming effects in response priming paradigms may be more robust than those observed in semantic priming paradigms. The present research results align nicely with this framework. On several occasions, Wentura and Degner (2010) suggest that response priming effects are relatively robust (although not entirely consistent) and that semantic priming effects tend to be more mercurial. With only one exception in the present set of studies (Study 5, gender), significant stereotype priming was consistently observed when an SCT (response priming paradigm) was used. Moreover, the meta-analysis of these results demonstrated that the stereotype priming effect observed with an SCT is significantly stronger than priming effects observed with an LDT. Priming effects in response priming paradigms likely reflect the combined influence of semantic and response priming processes (Wentura & Rothermund,

2014). The combined influence of these two processes ostensibly makes for a stronger, more robust stereotype priming effect in response priming paradigms. This can explain the more consistent (and stronger) pattern of significant stereotype priming effects when participants performed an SCT in the present research. In contrast, semantic priming paradigms lack the contribution of response processes, which potentially makes them less robust and more difficult to detect. This could explain why significant stereotype priming was observed only once (Study 3, race) when an LDT (semantic priming paradigm) was used in the present research. The pattern of more robust priming in response priming paradigms versus semantic priming paradigms has been established for evaluative priming, but has not been directly tested for stereotype priming (Wentura & Degner, 2010; although see Banaji & Hardin, 1996 for an indirect test). The present results, therefore, affirm the assertion that response priming paradigms give rise to more robust stereotype priming effects than semantic priming paradigms. One immediate implication of this research concerns appropriate interpretation of stereotype priming results. Much of the research on stereotype priming has been devoted to testing the “automaticity” of stereotype activation (e.g., Blair, 2002). The distinction between semantic and response priming paradigms speaks directly to this issue. If priming effects are driven by memory activation (as in semantic priming paradigms), then priming is thought to operate at an early stage when a target concept is accessed from memory. This more directly captures what researchers refer to as “automatic activation.” Alternatively, if priming effects are driven largely by response compatibility (as in response priming paradigms), then priming is occurring at a later stage that is more dependent on the processing goals that people use to select and execute a response. Because both processes can occur in response priming paradigms, stereotype priming effects from response priming paradigms, like the SCT, do not necessarily provide evidence for direct associative links among social stimuli. Instead, they might only be evidence of a response strategy and therefore not optimal for the study of “automatic” stereotype activation. There has been recognition that using a task like the LDT provides stronger evidence for automaticity than tasks like the SCT (e.g., Banaji & Hardin, 1996), but many stereotype priming studies have failed to consider the importance of participant task. For example, previous studies claim that eye gaze (Macrae, Hood, et al., 2002, Study 2), stimulus familiarity (Macrae, Mitchell, et al., 2002, Study 2),

It is important to note that priming effects observed in semantic priming paradigms may also be driven by other processes. This is particularly true for the LDT, which can be impacted by expectancy effects (Neely, 1977) and strategies such as backward checking (Neely, Keefe, & Ross, 1989).

Social Psychology (2018), 49(1), 29–46

Ó 2018 Hogrefe Publishing

Katherine R. G. White et al., Task & Stereotype Priming

negation training (Kawakami et al., 2000, Study 3), counterstereotype exposure/expectancy (Blair & Banaji, 1996, Studies 3 and 4), and stimulus configuration (Castelli et al., 2004, Studies 1 and 2; Macrae & Cloutier, 2009, Study 1b) all moderate stereotype activation. However, because each of these studies used a response priming paradigm, their manipulations may have affected response selection/execution as opposed to memory activation (Deutsch & Gawronski, 2009). That is not to say that these manipulations definitively do not alter activation, but the present evidence is insufficient to justify this claim due to the confounding presence of response competition processes in response priming paradigms. Researchers who are primarily interested in automatic stereotype activation have two potential avenues to avoid these issues. The first option is to use a response priming paradigm and implement methods that separate out the effects of semantic and response priming processes. One method of doing this is supplementing RT measures with ERPs, which is where our research began. It may also be possible to statistically separate different processes using just behavioral measures (e.g., Conrey, Sherman, Gawronski, Hugenberg, & Groom, 2005; Voss, Rothermund, Gast, & Wentura, 2013). Another option is to use a semantic priming paradigm in place of a response priming paradigm. It is important to note, however, that although using a semantic priming paradigm removes the confounding presence of response processes, priming observed in these paradigms does not necessarily reflect purely automatic processes (see Footnote 11). The findings of the present research also illustrate that priming effects in semantic priming paradigms (or at least the LDT) appear to be relatively inconsistent and fragile (see also Wentura & Degner, 2010 for review and suggestions). We used an LDT because it was the most prevalent semantic priming task in the literature, but there are other semantic priming tasks that might yield better results, such as the pronunciation task and semantic categorization. Indeed, our full-scale meta-analysis on sequential stereotype priming found that while stereotype priming failed to reach significance for the LDT, it was significant when a semantic categorization task was used (Kidder et al., in press). Priming was nonsignificant for pronunciation tasks in the meta-analysis, but this may be due to the small sample size available. Future research could directly investigate the extent to which these conclusions extend to other tasks used in semantic priming paradigms. It is important to note that the inconsistencies observed in this set of studies do not imply that stereotype priming

does not occur in semantic priming paradigms. As discussed previously, we elected to use the LDT because this task was the most prevalent in past research, and thus presumably led to the more robust findings. It may be that other tasks which engender more semantic processing of targets will lead to greater semantic stereotype priming. It is also possible that changing aspects of the LDT itself may lead to more robust priming. For example, Wentura and Degner (2010) suggest that nonword targets be similar to word targets to encourage greater depth of processing. We did this in Study 5 and found no evidence of priming, but a more systematic exploration of this might prove fruitful.12 It might also be useful to systematically vary whether the stereotype item appears as a target or prime. In all of our studies, the prime was an individual (male/ female, black/white) and the target was a stereotype item. We chose this arrangement because this seems the most common situation that involves stereotype activation and use – encountering an unknown person and activating stereotypes to guide interactions and memory. The reverse, however, can also occur – reading the word “nurse” may activate “woman.” These and other ideas may provide avenues for future exploration. Another objective of this paper is to serve as a cautionary example to guide future research on stereotypes and other constructs. We began this research with the idea of enhancing the study of stereotypes by employing event related potentials and proceeded under the assumption that stereotype priming effects from tasks such as the lexical decision task were established and relatively easy to replicate. This journey led to a greater appreciation for the importance of replication and follow-up research on studies that attempt to use more sophisticated methodologies to parse apart theoretical underpinnings of an observed effect. The research in stereotype priming, for example, has not focused much on the theoretical underpinnings of the priming that occurs. Research in other areas has demonstrated that there are different types of priming – some are due to “lower” types of perceptual processes such as stimulus form (e.g., Adelman et al., 2014) and others due to the nature and organization of memory (Schacter, Gutchess, & Kensinger, 2009). The latter category, which is the most relevant to stereotype priming, can be divided into the broad categories of semantic priming and associative priming. Semantic priming can be further divided into subtypes that involve categories (e.g., horse and pig because are both animals), function (e.g., a broom and floor because former is used to clean the latter), and shared script/ schema (e.g., wine and restaurant because former is

Studies 1–4 used non-words from the ARC database because the first study we performed (with ERPs) used ARC non-words. We wanted to avoid changing too many methodological factors across studies as we searched for an explanation of our null results, so we chose to keep these non-words in the next several studies.

Ó 2018 Hogrefe Publishing

Social Psychology (2018), 49(1), 29–46

included in the script of the latter; Moss, Ostrin, Tyler, & Marslen-Wilson, 1995). Research has revealed that these different semantic associations may be distinct from associative relations between items that are commonly linked in language use and thought (Moss et al., 1995; Postman & Keppel, 1970). Stereotype priming may be driven by both semantic and associative aspects. That is, stereotypes are frequently recognized as one component of social-cognitive schemas (e.g., Casper, Rothermund, & Wentura, 2010) and recent priming research has demonstrated their associative nature (Verhaeghen et al., 2011). Trying to differentiate the semantic and associative nature of stereotypes is potentially important because research suggests that semantic and associative relations may give rise to different priming effects (e.g., Carson & Burton, 2001; Hutchison, 2003; Lucas, 2000; Moss et al., 1995; Voss et al., 2013; Wiese & Schweinberger, 2008). While it may be arduous to create stereotype stimulus pairs that separate the script and associative relation components, the diffusion model analysis used by Voss and colleagues (2013) could be implemented to identify their differential impact on priming (see also Conrey et al., 2005). This might be particularly important for research on “automatic” stereotype priming, since previous research suggests that associative priming appears to reflect activation processes whereas semantic priming reflects response processes (Voss et al., 2013). Automaticity was not a primary focus in the current research, but it is our hope that this paper and its insights will help other stereotype priming researchers avoid some of the pitfalls that hampered our own endeavors.

Conclusions The present research demonstrated that participant task is a significant moderator of the stereotype priming effect. Specifically, stereotype priming effects are more consistent and robust when participants perform a stereotype classification task (SCT) than a lexical decision task (LDT). This pattern of results can be understood as the difference between response and semantic priming paradigms. Response priming paradigms ostensibly produce priming effects influenced by both semantic and response priming processes, resulting in stronger priming effects. However, the confounding presence of response processes in these paradigms presents significant problems for those primarily interested in “automatic” stereotype activation. Using a semantic priming paradigm is an appealing option to address this confound, but may introduce additional complexities and concerns, including an increased chance of null results. Stereotype priming researchers are encouraged to carefully consider these matters when planning future studies. Social Psychology (2018), 49(1), 29–46

Katherine R. G. White et al., Task & Stereotype Priming

Acknowledgments Portions of this research were performed under an appointment awarded to Katherine R. G. White from the US Department of Homeland Security (DHS) Scholarship Program, administered by the Oak Ridge Institute for Science and Education (ORISE). ORISE is managed by Oak Ridge Associated Universities (ORAU) under DOE Contract Number DE-AC05-06OR23100. All opinions expressed in this paper are the authors’ and do not necessarily reflect the policies and views of DHS, DOE, or ORAU/ORISE. We thank Adriaan Spruyt for helpful advice regarding the mental tally procedure in Study 2 and Ciara Kidder for her assistance with the meta-analytic analyses.

References Adelman, J. S., Johnson, R. L., McCormick, S. F., McKague, M., Kinoshita, S., Bowers, J. S., . . . Davis, C. J. (2014). A behavioral database for masked form priming. Behavior Research Methods, 46, 1052–1067. https://doi.org/10.3758/s13428-013-0442-y Banaji, M. R., & Hardin, C. D. (1996). Automatic stereotyping. Psychological Science, 7, 136–141. https://doi.org/10.1111/ j.1467-9280.1996.tb00346.x Bartholow, B. D., Dickter, C. L., & Sestir, M. A. (2006). Stereotype activation and control of race bias: Cognitive control of inhibition and its impairment by alcohol. Journal of Personality and Social Psychology, 90, 272–287. https://doi.org/10.1037/ 0022-3514.90.2.272 Bean, M. G., Stone, J., Moskowitz, G. B., Badger, T. A., & Focella, E. S. (2013). Evidence of nonconscious stereotyping of Hispanic patients by nursing and medical students. Nursing Research, 62, 362–367. https://doi.org/10.1097/NNR.0b013e31829e02ec Bessenoff, G. R., & Sherman, J. W. (2000). Automatic and controlled components of prejudice toward fat people: Evaluation versus stereotype activation. Social Cognition, 18, 329–353. https://doi.org/10.1521/soco.2000.18.4.329 Blair, I. V. (2002). The malleability of automatic stereotypes and prejudice. Personality and Social Psychology Review, 6, 242. https://doi.org/10.1207/S15327957PSPR0603_8 Blair, I. V., & Banaji, M. R. (1996). Automatic and controlled processes in stereotype priming. Journal of Personality and Social Psychology, 70, 1142. doi: 10.1037/0022-3514.70. 6.1142 Cacciari, C., & Padovani, R. (2007). Further evidence of gender stereotype priming in language: Semantic facilitation and inhibition in Italian role nouns. Applied Psycholinguistics, 28, 277–293. https://doi.org/10.1017/S0142716407070142 Carson, D. R., & Burton, A. M. (2001). Semantic priming of person recognition: Categorial priming may be a weaker form of the associative priming effect. The Quarterly Journal of Experimental Psychology A: Human Experimental Psychology, 54A, 1155–1179. https://doi.org/10.1080/02724980143000091 Casper, C., Rothermund, K., & Wentura, D. (2010). Automatic stereotype activation is context dependent. Social Psychology, 41, 131–136. https://doi.org/10.1027/1864-9335/a000019 Castelli, L., Macrae, C. N., Zogmaister, C., & Arcuri, L. (2004). A tale of two primes: Contextual limits on stereotype activation. Social Cognition, 22, 233–247. https://doi.org/10.1521/soco.22.2.233. 35462 Chwilla, D. J., & Kolk, H. J. (2005). Accessing world knowledge: Evidence from N400 and reaction time priming. Cognitive Brain

Ó 2018 Hogrefe Publishing

Katherine R. G. White et al., Task & Stereotype Priming

Research, 25, 589–606. https://doi.org/10.1016/j.cogbrainres. 2005.08.011 Clow, K. A., & Esses, V. M. (2007). Expectancy effects in social stereotyping: Automatic and controlled processing in the Neely paradigm. Canadian Journal of Behavioural Science/Revue Canadienne Des Sciences Du Comportement, 39, 161–173. https://doi.org/10.1037/cjbs20070013 Collins, A. M., & Loftus, E. F. (1975). A spreading-activation theory of semantic processing. Psychological Review, 82, 407. https:// doi.org/10.1037/0033-295X.82.6.407 Conrey, F. R., Sherman, J. W., Gawronski, B., Hugenberg, K., & Groom, C. J. (2005). Separating multiple processes in implicit social cognition: The quad model of implicit task performance. Journal of Personality and Social Psychology, 89, 469–487. https://doi.org/10.1037/0022-3514.89.4.469 Deutsch, R., & Gawronski, B. (2009). When the method makes a difference: Antagonistic effects on “automatic evaluations” as a function of task characteristics of the measure. Journal of Experimental Social Psychology, 45, 101–114. https://doi.org/ 10.1016/j.jesp.2008.09.001 Freng, S., & Willis-Esqueda, C. (2011). A question of honor: Chief Wahoo and American Indian stereotype activation among a university based sample. The Journal of Social Psychology, 151, 577–591. https://doi.org/10.1080/00224545.2010.507265 Habibi, R., & Khurana, B. (2012). Spontaneous gender categorization in masking and priming studies: Key for distinguishing Jane from John Doe but not Madonna from Sinatra. PloS One, 7, 1–7. https://doi.org/10.1371/journal.pone.0032377 Hehman, E., Volpert, H. I., & Simons, R. F. (2014). The N400 as an index of racial stereotype accessibility. Social Cognitive and Affective Neuroscience, 9, 544–552. https://doi.org/10.1093/ scan/nst018 Higgins, E. T., & King, G. (1981). Accessibility of social constructs: Information processing consequences of individual and contextual variability. Personality, Cognition, and Social Interaction, 69, 121. Hutchison, K. A. (2003). Is semantic priming due to association strength or feature overlap? A microanalytic review. Psychonomic Bulletin & Review, 10, 785–813. https://doi.org/10.3758/ BF03196544 Karylowski, J. J., Motes, M. A., Wallace, H. M., Harckom, H. A., Hewlett, E. M., Maclean, S. L., . . . Vaswani, C. L. (2001). Spontaneous gender-stereotypical categorization of trait labels and job labels. Current Research in Social Psychology, 6, 77–90. Kawakami, K., Dion, K. L., & Dovidio, J. F. (1999). Implicit stereotyping and prejudice and the primed Stroop task. Swiss Journal of Psychology/Schweizerische Zeitschrift Für Psychologie/Revue Suisse De Psychologie, 58, 241–250. https://doi.org/10.1024/ 1421-0185.58.4.241 Kawakami, K., & Dovidio, J. F. (2001). The reliability of implicit stereotyping. Personality and Social Psychology Bulletin, 27, 212–225. https://doi.org/10.1177/0146167201272007 Kawakami, K., Dovidio, J. F., Moll, J., Hermsen, S., & Russin, A. (2000). Just say no (to stereotyping): Effects of training in the negation of stereotypic associations on stereotype activation. Journal of Personality and Social Psychology, 78, 871–888. https://doi.org/10.1037/0022-3514.78.5.871 Kidder, C., White, K., Hinojos, M., Sandoval, M., & Crites, S. L. (2017). Sequential stereotype priming: A meta-analysis. Personality and Social Psychology Review. Advance online publication. https://doi.org/10.1177/1088868317723532 Kimura, A., Wada, Y., Goto, S., Tsuzuki, D., Cai, D., Oka, T., & Dan, I. (2009). Implicit gender-based food stereotypes. Semantic priming experiments on young Japanese. Appetite, 52, 521–524. https://doi.org/10.1016/j.appet.2008.11.002

Ó 2018 Hogrefe Publishing

Lucas, M. (2000). Semantic priming without association: A metaanalytic review. Psychonomic Bulletin & Review, 7, 618–630. https://doi.org/10.3758/BF03212999 Macrae, C. N., & Cloutier, J. (2009). A matter of design: Priming context and person perception. Journal of Experimental Social Psychology, 45, 1012–1015. https://doi.org/10.1016/j.jesp.2009.04.021 Macrae, C. N., Hood, B. M., Milne, A. B., Rowe, A. C., & Mason, M. F. (2002). Are you looking at me? Eye gaze and person perception. Psychological Science, 13, 460–464. https://doi. org/10.1111/1467-9280.00481 Macrae, C. N., & Martin, D. (2007). A boy primed Sue: Feature-based processing and person construal. European Journal of Social Psychology, 37, 793–805. https://doi.org/10.1002/ejsp.406 Macrae, C. N., Mitchell, J. P., & Pendry, L. F. (2002). What’s in a forename? Cue familiarity and stereotypical thinking. Journal of Experimental Social Psychology, 38, 186–193. https://doi.org/ 10.1006/jesp.2001.1496 Masson, M. J. (1995). A distributed memory model of semantic priming. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 3–23. https://doi.org/10.1037/0278-7393.21.1.3 McNamara, T. P. (2005). Semantic priming: Perspectives from memory and word recognition. New York, NY: Psychology Press. https://doi.org/10.4324/9780203338001 Minear, M., & Park, D. C. (2004). A lifespan database of adult facial stimuli. Behavior Research Methods, Instruments & Computers, 36, 630–633. https://doi.org/10.3758/BF03206543 Moskowitz, G. B., & Li, P. (2011). Egalitarian goals trigger stereotype inhibition: A proactive form of stereotype control. Journal of Experimental Social Psychology, 47, 103–116. https://doi. org/10.1016/j.jesp. 2010.08.014 Moss, H. E., Ostrin, R. K., Tyler, L. K., & Marslen-Wilson, W. D. (1995). Accessing different types of lexical semantic information: Evidence from priming. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 863–883. https://doi. org/10.1037/0278-7393.21.4.863 Müller, F., & Rothermund, K. (2014). What does it take to activate stereotypes? Simple primes don’t seem to be enough. A replication of stereotype activation (Banaji & Hardin, 1996; Blair & Banaji, 1996). Social Psychology, 45, 187–193. https:// doi.org/10.1027/1864-9335/a000183 Neely, J. H. (1977). Semantic priming and retrieval from lexical memory: Roles of inhibitionless spreading activation and limitedcapacity attention. Journal of Experimental Psychology: General, 106, 226–254. https://doi.org/10.1037/0096-3445.106.3.226 Neely, J. H., Keefe, D. E., & Ross, K. L. (1989). Semantic priming in the lexical decision task: Roles of prospective prime-generated expectancies and retrospective semantic matching. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15(6), 1003–1019. htps://doi.org/10.1037/0278-7393.15.6.1003 Plaza, M., Boiché, J., Brunel, L., & Ruchaud, F. (2017). Sport = Male. . . But not all sports: Investigating the gender stereotypes of sport activities at the explicit and implicit levels. Sex Roles, 76, 202–217. https://doi.org/10.1007/s11199-016-0650-x Postman, L., & Keppel, G. (1970). Norms of word association. New York, NY: Academic Press. Rastle, K., Harrington, J., & Coltheart, M. (2002). 358,534: The ARC Nonword Database. The Quarterly Journal of Experimental Psychology, 55A, 1339–1362. https://doi.org/10.1080/ 02724980244000099 Righi, G., Peissig, J. J., & Tarr, M. J. (2012). Recognizing disguised faces. Visual Cognition, 20, 143–169. https://doi.org/10.1080/ 13506285.2012.654624 Rohmer, O., & Louvet, E. (2012). Implicit measures of the stereotype content associated with disability. British Journal of Social Psychology, 51, 732–740. https://doi.org/10.1111/ j.2044-8309.2011.02087.x

Social Psychology (2018), 49(1), 29–46

Rumelhart, D. E., & McClelland, J. L. (1986). The PDP Research Group 1986: Parallel distributed processing: Explorations in the microstructure of cognition (Vols. 1 and 2). Cambridge MA: Foundations, MIT Press/Bradford Books. Sassenberg, K., & Moskowitz, G. B. (2005). Don’t stereotype, think different! Overcoming automatic stereotype activation by mindset priming. Journal of Experimental Social Psychology, 41, 506–514. https://doi.org/10.1016/j.jesp.2004.10.002 Schacter, D. L., Dobbins, I. G., & Schnyer, D. M. (2004). Specificity of priming: A cognitive neuroscience perspective. Nature Reviews Neuroscience, 5, 853–862. https://doi.org/10.1038/nrn1534 Schacter, D. L., Gutchess, A. H., & Kensinger, E. A. (2009). Specificity of memory: Implications for individual and collective remembering. In P. Boyer, J. V. Wertsch, P. Boyer, & J. V. Wertsch (Eds.), Memory in mind and culture (pp. 83–111). New York, NY: Cambridge University Press. https://doi.org/ 10.1017/CBO9780511626999.006 Spruyt, A., De Houwer, J., Everaert, T., & Hermans, D. (2012). Unconscious semantic activation depends on feature-specific attention allocation. Cognition, 122, 91–95. https://doi.org/ 10.1016/j.cognition.2011.08.017 Spruyt, A., De Houwer, J., & Hermans, D. (2009). Modulation of automatic semantic priming by feature-specific attention allocation. Journal of Memory and Language, 61, 37–54. https://doi.org/10.1016/j.jml.2009.03.004 Spruyt, A., De Houwer, J., Hermans, D., & Eelen, P. (2007). Affective priming of nonaffective semantic categorization responses. Experimental Psychology (formerly “Zeitschrift für Experimentelle Psychologie”), 54, 44–53. https://doi.org/ 10.1027/1618-3169.54.1.44 Stewart, T. L., Latu, I. M., Kawakami, K., & Myers, A. C. (2010). Consider the situation: Reducing automatic stereotyping through situational attribution training. Journal of Experimental Social Psychology, 46, 221–225. https://doi.org/10.1016/ j.jesp.2009.09.004 Verhaeghen, P., Aikman, S. N., & Van Gulick, A. E. (2011). Prime and prejudice: Co-occurrence in the culture as a source of automatic stereotype priming. British Journal of Social Psychology, 50, 501–518. https://doi.org/10.1348/014466610X524254 Voss, A., Rothermund, K., Gast, A., & Wentura, D. (2013). Cognitive processes in associative and categorical priming: A diffusion model analysis. Journal of Experimental Psychology: General, 142, 536–559. https://doi.org/10.1037/a0029459 Wentura, D., & Degner, J. (2010). A practical guide to sequential priming and related tasks. In B. Gawronski & B. K. Payne (Eds.), Handbook of implicit social cognition: Measurement, theory, and applications (pp. 95–116). New York, NY: Guilford Press. Wentura, D., & Rothermund, K. (2014). Priming is not priming is not priming. In D. C. Molden & D. C. Molden (Eds.), Understanding priming effects in social psychology (pp. 49–69). New York, NY: Guilford Press. White, K. R. G., & Crites, S. L. (2009). Gender stereotype priming in RTs and the N400. Unpublished raw data. El Paso, TX: University of Texas at El Paso. White, K. R., Crites, S. L., Taylor, J. H., & Corral, G. (2009). Wait, what? Assessing stereotype incongruities using the N400 ERP component. Social Cognitive and Affective Neuroscience, 4, 191–198. https://doi.org/10.1093/scan/nsp004 Wiese, H., & Schweinberger, S. R. (2008). Event-related potentials indicate different processes to mediate categorical and associative priming in person recognition. Journal of Experimental Psychology: Learning, Memory, and Cognition, 34, 1246–1263. https://doi.org/10.1037/a0012937 Wittenbrink, B. (2007). Measuring attitudes through priming. In B. Wittenbrink & N. Schwarz (Eds.), Implicit measures of attitudes (pp. 17–58). New York, NY: Guilford Press.

Social Psychology (2018), 49(1), 29–46

Katherine R. G. White et al., Task & Stereotype Priming

Wittenbrink, B., Judd, C. M., & Park, B. (1997). Evidence for racial prejudice at the implicit level and its relationship with questionnaire measures. Journal of Personality and Social Psychology, 72, 262. https://doi.org/10.1037/0022-3514. 72.2.262 Received December 7, 2016 Revision received July 24, 2017 Accepted August 31, 2017 Published online February 7, 2018 Katherine R. White Department of Psychology Kennesaw State University Kennesaw, GA 30144 USA kwhit162@kennesaw.edu

Appendix A Primes and Target Stimuli for Studies 1–5 Study 1 Primes Pictures of unknown males and females were gathered from a public rating forum called “Rate My Picture” on https://www.MySpace.com. Pictures were initially gathered and then screened for unusual or distracting features (e.g., green hair, busy background). The remaining pictures were rated for attractiveness by 47 individuals on a scale from 3 to +3. The present study used 54 pictures of males and 54 pictures of females as primes, all neutral in attractiveness. Names were chosen from a list of approximately 600 male and female names that were rated by 20 individuals along the dimensions of masculinity/femininity and familiarity. The masculinity/femininity of names was rated on a scale from 3 (very masculine) to +3 (very feminine). Familiarity was rated on a scale from 3 (very unfamiliar) to +3 (very familiar). The 54 most masculine and familiar names were chosen to serve as male primes, and the 54 most feminine and familiar female names were chosen to serve as female primes. The average masculinity rating for male names was 2.55, with a range of 2.2 to 2.95. The average familiarity rating for male names was 1.69, with a range of 1–2.76. For female names, the average femininity rating was 2.62, with a range of 2.4–2.95. The average familiarity rating for female names was 1.63, with a range of 1–2.76. Male Names: JACOB, RICARDO, MARCOS, RAUL, CHRISTOPHER, JESUS, JOSEPH, ROBERT, ANTONIO, ARTURO, EDUARDO, GEORGE, HUGO, MANUEL, MARK, MATTHEW, PETER, ROBERTO, ALFREDO, DAVID, HECTOR, PABLO, PAUL, PEDRO, Ó 2018 Hogrefe Publishing

Katherine R. G. White et al., Task & Stereotype Priming

ANDREW, BENJAMIN, FRANK, JONATHAN, RUBEN, STEVE, CARLOS, EDWARD, ENRIQUE, ISAAC, JAVIER, JOHN, MIKE, OSCAR, PATRICK, RAYMOND, SERGIO, WILLIAM, BRIAN, JAMES, JORGE, MIGUEL, VICTOR, AARON, ADAM, ALAN, ANTHONY, CESAR, JOSHUA, RICK Female Names: ELIZABETH, JESSICA, VICTORIA, ISABELA, MELISSA, SAMANTHA, ALICIA, BRITTANY, KATHERINE, MONICA, ALEJANDRA, DIANA, JENNIFER, LILIANA, LISA, MARIA, REBECCA, SILVIA, ALICE, ANNIE, CYNTHIA, EMILY, KATIE, SANDRA, VALERIE, VERONICA, AMANDA, ANITA, ANNA, CRYSTAL, DENISE, EMMA, EVA, JACQUELINE, JULIA, LUCY, NATALIE, PAMELA, STEPHANIE, SUZANNE, THERESA, VANESSA, BRENDA, CHRISTINA, CINDY, CLAUDIA, GLORIA, GRACE, JULIE, LYDIA, MEGAN, OLIVIA, RACHEL, ROSA

Targets Gender stereotypical words were obtained in a series of steps. First, 18 undergraduate students were asked to list personality traits, occupations, and objects stereotypically associated with men and women. Second, these were entered into a database and the most frequently generated words were identified. Lastly, these 70 words were rated by an independent set of 44 undergraduate students for masculinity/femininity (1 = very masculine, 7 = very feminine) and valence ( 3 to +3). The 8 most feminine traits (M = 5.45) and 8 most feminine nontraits (M = 6.23; e.g., occupations, objects, activities) were chosen to serve as female stereotypical targets. The 8 most masculine traits (M = 2.34) and 8 most masculine nontraits (M = 1.66) were chosen to serve as male stereotypical targets. Male and female targets were significantly different in their masculinity/femininity ratings, t = 19.88, p < .002, and were matched for valence, p = .404, length, p = .544, and frequency, p = .287. Female Traits: GOSSIPY, NURTURING, EMOTIONAL, DELICATE, TALKATIVE, SENSITIVE, CARING, VULNERABLE Female Nontraits: DRAMA, LIPSTICK, PURSE, MANICURE, ESTROGEN, NURSE, SECRETARY, BEAUTICIAN Male Traits: MACHO, TOUGH, ROUGH, VIOLENT, COCKY, STRONG, AGGRESSIVE, DOMINANT Male Nontraits: TESTOSTERONE, TRUCKS, PLUMBER, FOOTBALL, SOLDIER, CONSTRUCTION, MECHANIC, MUSCLES Nonwords: WRYNKED, KREMBTH, EEBS, STRYLTH, PEUNTHS, SKWUXTE, SCWIETH, PHROOGNS, SPRUPCE, SPRIRGUE, SPOMFS, FROUGHGE, PRAFES, CWERSH, SWALS, SHRUPES, SPENCHED,

Ó 2018 Hogrefe Publishing

TWILS, GLAUGHTH, GHUGNTH, SPLERLTE, TROUZE, ZYMPH, BLEIGG, REIGUED, SNOLFS, GHWUCSTS, PSANCSED, THROOMS, WEUSED, PLAUGHGUES, GHWOURV, CURDGED, GWURP, SKIRCHED, SPRURVES, TWOMF, DWEAMTHS, SCKRERGNS, SKWYLMB, TRAUGHBED, SOCH, GHEILTS, ZEIDGE, LURCED, BONTHS, SNOURMB, JERGN, GNOWLE, WRERCHED, CRINSE, FROINED, WOWGE, CRIGHV, GEELLS, PRAUGHD

Study 2 Same as Study 1.

Study 3 Primes Pictures of unknown males and females were gathered at the University of Texas at El Paso and were used courtesy of Michael Zárate. They were head and shoulder pictures of individuals on a neutral background.

Targets Race stereotypical words were obtained in a series of steps. First, we identified a set of items stereotypically associated with blacks, whites, and Hispanics. Some stereotype terms were found in published research, and we added to this set by asking approximately 10 individuals in El Paso, Texas (primarily undergraduate students) to generate stereotype words for blacks, whites, and Hispanics. This resulted in a list of 270 potential items. Second, we then had 42 undergraduate students rate these 270 items. Half rated them on a 7-point Hispanic/black dimension and a 7-point valence dimension; and half rated the items on a 7-point white/black dimension and a 7-point valence dimension. We then elected to focus the study on black/Hispanic stereotypes and selected the strongest set of 20 stereotype items for each group that were closely matched in stereotypicality (M = 1.94 vs. 6.08 for Hispanic and black terms, respectively, on a scale with 1 = Hispanic, 4 = Race Neutral, 7 = Black) and valence (M = 4.59 and 4.56 for Hispanic and black terms, respectively, on a scale with 1 = Negative, 4 = Neutral, 7 = Positive). Hispanic: CHILI, DAY LABORER, MAIDS, FAMILY, MACHO, SHORT, TRADITION, ENCHILADAS, LATIN AMERICA, MARIACHI, TEQUILA, SOCCER, CUMBIA, DEPORTATION, IMMIGRANT, SAMBA, TANGO, ACCENT, CONSTRUCTION, CATHOLICS Black: SOUL, ATHLETE, RAPPER, BLACK PANTHERS, SWAGGER, TALL, SPORTS, FOOTBALL, CHURCH CHOIR, RUNNER, LOUD, JAZZ, R&B, SEGREGATION,

Social Psychology (2018), 49(1), 29–46

Katherine R. G. White et al., Task & Stereotype Priming

SLAVERY, HIP HOP, BASKETBALL, AFRICA, NFL, BLING Nonwords: PSOUDES, KNEULLS, FRIBBS, VOLMS, SMOIGNS, PRANKSED, SNURFS, STEAVES, TWARKED, BLORTHED, PHREICS, SHROONS, STROOBS, WOFTS, WHOLS, SPIRPED, CRAWFED, QUEACKED, MAUGED, THAIPS, CLULTHS, SWARDES, DWEIGUES, TEMPCED, BREPTH, TOVS, SNYGUES, WHOOKES, SCREALTS, PHREGUED, SWOIST, PHARDES, YUSKS, PHREABS, KNARVES, BRUIFFTH, MEEFTH, BLAUMS, THROOPPS, RERNS, SCROOMTH, GNULKED, HIEFED, SHUNCHED, QUAIDS, SNAIDS, WOSCS, PSOUGHBED, KEAMMED, DRIGHDS, BREUNDS, GEMPCED, SLAMBTH, PHRYGGED, SLOMES, WEGUED, SHRUMPED, SHROUND, FLINCE, SHOURNED, VAPSE, FENTH, CROICED, SOYS, MICKED, WUMPS, PLOOD, RHIRD, SWOUST, SHARCED, CLUFT, RAWNED, PHRARS, FOWD, PLECT, SNOSSED, SPRYPT, GRUCT, MAFFED

Study 4 Primes Same as Study 3.

masculinity/femininity ratings (MM = 1.74, MF = 6.26), t = 65.04, p < .001, but were matched for valence (MM = 4.38, MF = 4.52), t = 0.53, p = .601. Nonwords were created by changing one or two letters in the original target word to create a pronounceable nonword. Men: mustache, priest, bald, necktie, president, cologne, muscular, mechanic, baseball, hockey, boxing, plumber, jock, pirate, umpire, handsome, cigars, football, beer, CEO, wrestling, carpenter, sheriff, basketball, hairy, politician, sword, army, janitor, hunting Women: bossy, needy, weak, apron, laundry, crying, nurse, kitchen, gossip, pink, sewing, babysitter, homemaker, pretty, secretary, earrings, knitting, pantyhose, barbie, nanny, ballet, petite, anorexia, lipstick, mascara, dolls, skirt, makeup, perfume, bikini Gender Nonwords: masteshe, proest, bild, neftie, plesodent, cafogne, muscitar, muchinic, besepall, hachey, bexong, prumber, juck, pirase, umvare, hindtome, cagabs, faotvall, buer, CQO, wrasnling, caypejter, scerlff, boseetball, hawry, pofitinian, swird, almy, jabifor, hantilg, boesy, naegy, wepk, apoon, liungry, creeng, norse, kutshen, gresip, penk, syweng, bajysipter, hokesaker, poetty, searatury, eurrinks, kniftigg, partyyose, berjie, nuniy, bahlut, pahite, anirebia, libstock, madcera, dopls, skibt, maferp, perbome, bukoni

Targets The Hispanic & black stereotype words and nonwords were the same as Study 3. The process for identifying gender stereotype words was identical to that described in Study 3, but the original files and data from the pilot study could not be located so we are unable to report mean stereotypicality and valence for the male and female stereotype words. The words used in the study are in the table below. Female: GOSSIP, SEWING, ANOREXIA, PETITE, EARRINGS, PINK, KNITTING, BARBIE, MASCARA, DOLLS, BALLET, MAKEUP, MAIDS, SKIRT, PRETTY, PANTYHOSE, PERFUME, NANNY, LIPSTICK, BIKINI Male: PRIEST, UMPIRE, CARPENTER, CONSTRUCTION, BOXING, BASEBALL, HOCKEY, MECHANIC, PRESIDENT, HUNTING, FOOTBALL, NBA, BALD, PLUMBER, NECKTIE, MACHO, NFL, COLOGNE, MUSTACHE, HANDSOME

Study 5 Gender Fourteen participants provided valence (1 = very negative, 7 = very positive) and masculinity/femininity (1 = very associated with men, 7 = very associated with women) ratings for the gender target stimuli. The 30 most masculine and 30 most feminine stimuli were then selected for use in the study. These stimuli were significantly different regarding their

Social Psychology (2018), 49(1), 29–46

Race Sixteen participants provided valence (1 = very negative, 7 = very positive) and race association (1 = very associated with Blacks, 7 = very associated with Whites) ratings for the race target stimuli. The 30 words most with Blacks and 30 words most associated with Whites were then selected for use in the study. These stimuli were significantly different regarding their race association ratings (MB = 2.32, MW = 5.73), t = 37.21, p < ..001, but were matched for valence (MB = 3.44, MW = 3.95), t = 1.66, p = .102. Nonwords were created by changing one or two letters in the original target word to create a pronounceable nonword. Blacks: rappers, ghetto, dreads, weaves, basketball, gangs, rims, streetwise, foodstamps, poverty, rhythm, soulful, welfare, violence, mistreated, watermelon, dancing, unemployed, uneducated, prison, projects, collards, chicken, struggle, fighting, jazz, lips, reunions, ribs, dangerous Whites: country, beach, uptight, snobby, lawyers, power, pink, politics, swimming, boss, suburbs, trailers, superiority, patriotic, shotgun, honkie, Abercrombie, farming, presidents, majority, skiing, republican, pale, hunting, privileged, racist, golf, tennis, tanning, redneck Race Nonwords: roopers, guegto, drefts, woales, baspethall, golgs, riws, straetcise, foolslamps, pazerty,

Ó 2018 Hogrefe Publishing

Katherine R. G. White et al., Task & Stereotype Priming

rhuuhm, syulkul, wilfire, vielesce, mustgeated, witertelon, domcing, unimzloyed, unewudated, plicon, prokacts, coltords, chopken, steohgle, feghking, jalz, lups, reinisns, rebs, dangproos, cquttry, boach, uppight, sbobly, laquers, puwtr, ponk, pofitids, swomling, byss, sipurbs, treifers, sapertority, periotic, shuttun, hyngie, azerqrombie, ferping, presihebts, mipority, swuing, reduclican, paie, honjing, pyivilewed, rikist, guff, ternas, talping, reffeck

we examined both the untransformed RT data and also RT data that was subjected to an inverse transform. The pattern of results and significance for these analyses were identical across all five studies. We report the analyses on untransformed RT data for ease of interpretation. Reported below are the numbers associated with these data preparation steps for each of the five studies.

Study 1

Appendix B Data Preparation for Studies 1-5 Except as noted below, we used an identical procedure to screen and prepare data for analyses across all five studies. First, responses within 250 ms of target onset were eliminated. Second, we eliminated data from participants whose overall accuracy was more than two standard deviations below the mean accuracy for their between-subject condition (if the experiment had a between-subject manipulation). Third, response times (RTs) greater than 2 standard deviations from each participantâ&#x20AC;&#x2122;s mean response latency were replaced by the mean plus 2 SD values for that participant (see Wittenbrink, 2007 for similar criterion). Fourth, we eliminated data from participants if one or more within-subject conditions contained less than 1/3 of available trials. This is done to insure that each within-subject condition had enough trials to obtain a stable average for that condition. Participants sometimes have a response bias (e.g., respond â&#x20AC;&#x153;nonwordâ&#x20AC;? in LDT, or â&#x20AC;&#x153;maleâ&#x20AC;? in SCT) that results in fewer trials in certain conditions, but the bias is not strong enough to lower their overall accuracy rate below the 2 SD accuracy threshold for excluding all of their data. After these steps, response times from correct responses were averaged in each condition. In LDT conditions, the nonword trials were not examined. In addition to the analyses reported in the text, we did two other sets of analyses on the data in all five studies to support the results reported in the text. First, to ascertain whether steps were followed to exclude data impacted findings, we analyzed the data from all participants (including those who had low accuracy or were missing more than 2/3 of available trials in a within-subject condition). The pattern of results and significance of these analyses are identical to those reported in the primary analyses for all five studies, save for one finding in Study 3 that was significant in the primary analyses and marginal in the analysis with all participants (see Footnote 10). Second,

Ă&#x201C; 2018 Hogrefe Publishing

Data from 11 participants were not available because of technical problems with hardware/software during the experiment. Data from 12 participants excluded due to low accuracy: Four participants in the LDT condition (M = 94.7%; SD = 5.0%); three in the pre-primed LDT condition (M = 94.6%; SD = 5.7%), and five in the gender categorization condition (M = 90.9%; SD = 11.2%). Data from five participants excluded due to low number of trials in one or more conditions (1 in LDT, 1 in preprimed LDT, and 3 in gender categorization). For the participants included in the primary analyses, 5.3% of trials were eliminated because response occurred within 250 ms of target onset. Although we begin data preparation by removing fast responses, we report the percentage of fast response only for participants included in the analyses. A handful of participants engage in a practice of responding quickly (and randomly) to targets so they can quickly finish the experiment (accuracy for fast responses is approximately chance across all participants). Thus, including participants who are removed due to low accuracy inflates percentage for participants whose data is analyzed.

Study 2

Data from two participants were not available because of technical problems with hardware/software during the experiment. Data from 11 participants excluded due to low accuracy: Eight participants in the ethnicity tally condition (M = 87.9%; SD = 11.2%) and three in the gender tally condition (M = 89.1%; SD = 8.8%). Data from 15 participants excluded due to low number of trials in one or more conditions (11 in ethnicity tally and 4 in the gender tally). For the participants included in the primary analyses, 13.7% of trials were eliminated because response occurred within 250 ms of target onset.

Social Psychology (2018), 49(1), 29â&#x20AC;&#x201C;46

Katherine R. G. White et al., Task & Stereotype Priming

Study 3

Data from three participants were not available because of technical problems with hardware/software during the experiment. Data from 22 participants excluded due to low accuracy: Five participants in the LDT condition (M = 85.3%; SD = 14.4%) and 17 in the SCT condition (M = 77.1%; SD = 18.0%). The cutoff number used for elimination in the SCT condition is one change from the typical procedure. Because of the low mean accuracy and high standard deviation in this condition, the two standard deviation cutoff fell below chance levels of responding. So, we used the same cutoff in the SCT condition as used in the LDT condition (56.5%). Data from six participants excluded due to low number of trials in one or more conditions (2 in SCT and 4 in LDT). For the participants included in the primary analyses, 8.7% of trials were eliminated because response occurred within 250 ms of target onset.

Data from two participants excluded due to low number of trials in one or more conditions (1 in gender task and 1 in race task). For the participants included in the primary analyses, 0.9% of trials were eliminated because response occurred within 250 ms of target onset.

Study 5

Data from 20 participants excluded due to low accuracy: thirteen in the race-SCT condition (M = 78.7%; SD = 16.5%); two in the gender-SCT condition (M = 91.5%; SD = 9.0%); one in the race-LDT condition (M = 93.4%; SD = 6.9%); four in the gender-LDT condition (M = 91.7%; SD = 10.7%); Data from one participant (gender-LDT condition) excluded due to low number of trials in one or more conditions. For the participants included in the primary analyses, 0.9% of trials were eliminated because response occurred within 250 ms of target onset.

Study 4

Data from 11 participants excluded due to low accuracy: nine in the gender-LDT condition (M = 87.9%; SD = 13.8%) and two in the race-LDT condition (M = 91.3%; SD = 8.3%).

Social Psychology (2018), 49(1), 29â&#x20AC;&#x201C;46

Ă&#x201C; 2018 Hogrefe Publishing

Original Article

Less Power, Greater Conflict Low Power Increases the Experience of Conﬂict in Multiple Goal Settings Petra C. Schmid Department of Management, Technology, and Economics, Swiss Federal Institute of Technology (ETH Zurich), Zurich, Switzerland

Abstract: Power facilitates goal pursuit, but how does power affect the way people respond to conflict between their multiple goals? Our results showed that higher trait power was associated with reduced experience of conflict in scenarios describing multiple goals (Study 1) and between personal goals (Study 2). Moreover, manipulated low power increased individuals’ experience of goal conflict relative to high power and a control condition (Studies 3 and 4), with the consequence that they planned to invest less into the pursuit of their goals in the future. With its focus on multiple goals and individuals’ experiences during goal pursuit rather than objective performance, the present research uses new angles to examine power effects on goal pursuit. Keywords: power, goal conflict, self-efficacy, multiple goals

The experience of power helps people to pursue their goals effectively (Galinsky, Gruenfeld, & Magee, 2003; Guinote, 2007b; Magee & Smith, 2013), but what if they want to pursue multiple goals that draw on the same resources? The present paper examined the effect of psychological power on the experience of resource conflict. Resource conflict occurs when goals are incompatible due to competition for the same limited resources such as time and money (Kruglanski et al., 2002; Riediger & Freund, 2004; Riediger, Freund, & Baltes, 2005). For example, a person may wish to enroll in a leadership course to increase his/her chances for a promotion but at the same time may want to spend more time with family – both goals draw on the resource of time. Such situations typically generate a subjective experience of goal conflict, which is associated with greater stress, worry, and negative affect (Carver & Scheier, 1998; Emmons & King, 1988; Hirsh, Mar, & Peterson, 2012; Inzlicht, Bartholow, & Hirsh, 2015). Power is defined as the ability to control resources that another person requires or desires (Galinsky, Rucker, & Magee, 2015), and may be operationalized as a person’s position in a hierarchy. Power may also represent a psychological property that reflects an individual’s representation of his/her power relative to others (e.g., A. Anderson & Galinsky, 2006; Hershcovis et al., 2017). The psychological experience of power has considerable effects on how people behave and how they pursue goals (A. Anderson & Galinsky, 2006). Past research on the effect of psychological Ó 2018 Hogrefe Publishing

power on goal pursuit has focused largely on objective performance outcomes (e.g., Galinsky et al., 2003; Guinote, 2007a; Magee, Galinsky, & Gruenfeld, 2007; Schmid, Kleiman, & Amodio, 2015; Schmid & Schmid Mast, 2013; Schmid, Schmid Mast, & Mast, 2015) and neglected what people experienced when pursuing goals. Furthermore, these studies mainly examined the pursuit of a single goal although people typically pursue multiple goals in real life. The present studies add new perspectives to the literature on psychological power and goal pursuit by focusing on the pursuit of multiple goals and individuals’ experiences during goal pursuit rather than objective performance. Specifically, our main aim was to investigate whether relatively higher psychological power is associated with reduced experience of resource conflict, and to explore the consequences of such effects for goal pursuit. Because the experience of goal conflict typically lowers people’s investments into the pursuit of their goals and decreases their well-being (Emmons & King, 1988), this research may provide important insights into the effect of psychological power on effective goal pursuit and, more broadly, on people’s affectivity.

Goal Conflict People may pursue goals that can be characterized as pursuing a rewarding state (e.g., spending more time with Social Psychology (2018), 49(1), 47–62 https://doi.org/10.1027/1864-9335/a000327

the family) and goals that include avoiding an undesired state (e.g., avoiding a non-sympathetic coworker). As such, incompatibility between goals may be experienced in so-called “approach-approach conflicts,” which refer to situations in which one wishes to approach two different goals at the same time (e.g., when one desires to invest more hours at work to increase the salary and at the same time wants to spend more time with the family), approachavoidance conflicts (e.g., when one wants to avoid a nonsympathetic coworker, and at the same time needs the coworker’s expertise to finish a project), and avoidanceavoidance conflicts (e.g., when one wants to avoid a nonsympathetic coworker, and at the same time wants to avoid that others get the impression that one is not social and collaborative; Lewin, 1935). Goal conflict is not bad per se, and sometimes even beneficial for goal attainment (e.g., Kleiman, Hassin, & Trope, 2014). For instance, it has been proposed and shown that to be able to successfully pursue a goal in the presence of distractors one first needs to detect that there is a goal conflict (Botvinick, Braver, Barch, Carter, & Cohen, 2001; Myrseth & Fishbach, 2009). The goal-shielding theory suggests that when multiple goals are in conflict, goals may inhibit one another, as part of a potentially automatic and nonconscious process. Consequentially, resources may be invested into a focal goal at the expense of alternative goals (Shah, Friedman, & Kruglanski, 2002). Thus, detecting goal conflict may be important for the attainment of focal goals. The focus of the present paper lies on individuals’ subjective and conscious experience of resource conflict, which likely follows initial (potentially unconscious) processes of conflict detection and intergoal inhibition. While the early detection of goal conflict may be beneficial for the pursuit of a focal goal (Botvinick et al., 2001), subjectively experienced conflict between one’s multiple goals has been associated with reduced well-being and reduced willingness to invest into one’s goals (e.g., Emmons & King, 1988).

Power and Goal Conflict Many studies have shown that psychological power facilitates the pursuit of a single goal (e.g., Galinsky et al., 2003; Guinote, 2007b; Schmid & Schmid Mast, 2013). According to the social distance theory of power (Magee & Smith, 2013), this may be the case because high-power individuals engage more abstract mental representations (see also Smith & Trope, 2006), which facilitates selfcontrol (Fujita, Trope, Liberman, & Levin-Sagi, 2006). Self-control is required to keep one’s behavior in line with a goal in the presence of distractors, temptations, and alternative goals. Social Psychology (2018), 49(1), 47–62

P. C. Schmid, Power and Goal Conflict

Indeed, several studies revealed that manipulated high power increased people’s ability to focus on goal-relevant information in the presence of distractors compared to low power (Guinote, 2007b; Schmid, Kleiman, et al., 2015; Smith, Jostmann, Galinsky, & van Dijk, 2008). Using event-related potential (ERP) methods, a study revealed further that high-power and low-power participants did not significantly differ in how they processed conflict evoked by distracting information (i.e., as indexed by the N2r ERP component); however, only high-power participants were able to use this conflict signal to control their behavior (Schmid, Kleiman, et al., 2015). Importantly, in this study, goal conflict was evoked by irrelevant information on a trial-by-trial basis, and conflict processing was measured as a (potentially nonconscious) cognitive process that occurred just milliseconds after the presentation of conflict trials and before goal-consistent responses were given. Conflict processing in this study is different from the (conscious) subjective experience of conflict that people may report when they have time to deliberate about how much their multiple relevant goals are in conflict with one another. Relatively little research has investigated how power affects the pursuit of multiple relevant goals. Cai and Guinote (2017) found that lack of power decreased people’s ability to multitask, suggesting that the powerless may have difficulty resolving conflict between multiple goals. This may be the case because the experience of power affects individuals’ task strategies in multiple goal situations. Indeed, Schmid, Schmid Mast, et al. (2015) showed that when performing a cognitively demanding dual task, high-power participants prioritized one task over the other (i.e., performed relatively well on one task and relatively badly on the other task). Low-power participants, in contrast, performed the two tasks equally well/poorly. Moreover, Guinote (2008) found that manipulated high power led people to pursue their multiple goals as a function of situational affordances (e.g., they planned outdoor activities during the summer and indoor activities during the winter) while low-power participants more likely pursued multiple goals in parallel, independent of situational affordances. These studies focused on objective outcomes during multiple goal pursuit. It is also possible that both goal prioritization and sequential rather than parallel pursuit of multiple goals may serve the powerful as strategies to decrease the subjective experience of goal conflict. To summarize, optimal self-regulation and successful goal pursuit might require the detection of conflict, which facilitates goal focus and conflict resolution (Botvinick et al., 2001; Myrseth & Fishbach, 2009). When answering questions about conflict, both powerful and powerless people might initially and potentially unconsciously detect conflict (in line with Schmid, Kleiman, et al., 2015), but Ó 2018 Hogrefe Publishing

P. C. Schmid, Power and Goal Conflict

because the powerful are better at subsequently inhibiting influences from distractors and alternative goals (Schmid, Kleiman, et al., 2015; Schmid, Schmid Mast, et al., 2015), we expected that the powerful would subjectively experience and report less conflict than the powerless.

Self-Efficacy as a Mediator? A way by which high power might reduce people’s experience of goal conflict relative to low power is by increasing self-efficacy (A. Anderson & Galinsky, 2006; Schmid & Schmid Mast, 2013; Weick & Guinote, 2010). Self-efficacy refers to the extent to which an individual believes that s/he is able to complete a task and to reach a goal, and greatly affects people’s performance and their motivation to engage in goal-directed behavior (Abele & Spurk, 2009; Bandura, 1997; Judge & Bono, 2001; Lent, Brown, & Hackett, 1994). Self-efficacious people may cope more effectively with the stressful situation of having conflicting goals, such as in work-family conflict (Cinamon, 2006). Because power provides people with confidence in their skills and abilities (Fast, Sivanathan, Mayer, & Galinsky, 2012), it has been argued that power may also increase self-efficacy beliefs (A. Anderson & Galinsky, 2006). Indeed, the experience of power has been related to greater self-efficacy specifically (A. Anderson & Galinsky, 2006; Schmid & Schmid Mast, 2013), as well as to related concepts such as greater optimism about future outcomes (A. Anderson & Galinsky, 2006), to greater perception of personal control (Fast, Gruenfeld, Sivanathan, & Galinsky, 2009), and to the experience of demanding situations as challenging rather than threatening (Scheepers, de Wit, Ellemers, & Sassenberg, 2012). Although high power has been related to greater self-efficacy than low power in past research, self-efficacy did not typically mediate the power effect on cognition and behavior (A. Anderson & Galinsky, 2006; Schmid & Schmid Mast, 2013; Slabu & Guinote, 2010; Weick & Guinote, 2010). Nevertheless, given the empirical research linking power to self-efficacy (e.g., A. Anderson & Galinsky, 2006; Schmid & Schmid Mast, 2013; Weick & Guinote, 2010) and self-efficacy to enhanced goal pursuit processes and reduced goal conflict (e.g., Cinamon, 2006; Judge & Bono, 2001), we tested the possibility that relatively greater power is associated with reduced experience of goal conflict through increased self-efficacy in the present research. The focus lay on the general feeling that one is able to achieve goals (i.e., generalized self-efficacy), and the specific belief that one is able to achieve one’s own most important goals (i.e., goal-related self-efficacy).

Ó 2018 Hogrefe Publishing

Overview of Studies The main goal of this research was to investigate how individuals’ psychological power is related to their experience of conflict between multiple goals. Goal conflict was assessed by focusing on general goal incompatibility (Studies 1 and 4) and more specifically, on resource conflict (Studies 1–4). Studies 1 and 2 focused on how individual differences in trait power relate to the experience of goal conflict. In Study 1, two conflicting goals were presented in the form of a scenario, ensuring that all participants were exposed to the same goal conflict. In Study 2, participants rated the extent to which they perceived that their own goals would compete for the same resources. In Study 3, power was manipulated to determine a causal effect of high versus low power on the experience of resource conflict. In Study 4, power was also manipulated and a control group was included. Moreover, it was tested whether the power effect on experienced goal conflict has consequences for individuals’ intentions to further pursue their goals. All studies included a measure of self-efficacy (generalized or related to their goals) to test whether it mediates the effect of power on experienced goal conflict. For all studies, we report all data exclusions (if any), all manipulations, and all measures. Sample sizes for the four studies were determined a priori based on power analyses using G*Power (Faul, Erdfelder, Lang, & Buchner, 2007), setting α-error probability to .05 and power to .80. Studies 1, 2, and 4 included online samples recruited through Amazon’s Mechanical Turk (mTurk), and we expected small-to-medium-sized effects (r = .2), given the lack of control we had over participants’ environments and the relatively great heterogeneity of the samples. Study 3 was a laboratory study for which we estimated a medium-sized effect (r = .3), because the study environment was greatly standardized and the sample more homogeneous. According to these analyses, sufficient statistical power would be achieved by including 191 participants each in Studies 1 and 2, 102 participants in Study 3, and 318 participants in Study 4. In all studies, some additional participants were recruited to account for possible outliers or noncompliant participants who would have to be excluded from analyses.

Study 1 The goal of Study 1 was to examine whether individual differences in trait power are related to the extent to which people experience two approach goals as conflicting with one another. The two conflicting approach goals were presented within a scenario, thus keeping actual goal

Social Psychology (2018), 49(1), 47–62

conflict constant. Both goals were described as being relevant, but one goal was presented as the primary goal and the other goal as the secondary goal. This provided participants with an option to reduce experienced goal conflict through prioritization. Our main interest focused on participants’ subjective experience of goal conflict in this situation. It was further tested whether participants’ generalized self-efficacy could explain a potential power effect on the experience of goal conflict.

P. C. Schmid, Power and Goal Conflict

(2) Your focal goal is to be one of the 30 best runners at a half-marathon. At the same time, you are considering accepting a job at the museum to earn some extra money. (3) Your focal goal is to party your 30th birthday. You are also planning to get a tattoo. Participants could only proceed to the next question if they answered this question correctly (i.e., if they chose answer option (2). If the answer was wrong, they had to reread the scenario.

Method Participants Participants were 197 US citizens recruited over mTurk and paid $1 for study completion. Of this sample, 57.4% were female and participants’ average age was 34.58 years, SD = 10.45. Procedure After providing consent, participants completed two different blocks in randomized order. In the goal conflict block, participants imagined themselves in a scenario with two conflicting goals and answered questions concerning these goals and the extent to which they perceived these goals to be in conflict with one another. In the trait scale block, participants completed scales assessing their generalized self-efficacy, trait power, and other traits. Finally, all participants provided information about their demographics, education, and profession. Goal Conflict Scenario Participants were asked to imagine themselves in a situation with two goals that were in conflict with one another because they competed for the resource time. This resource conflict represented an approach-approach conflict: The focal goal was to train hard in order to be among the top 30 runners at a half-marathon and the secondary goal was to take up a part-time job at the museum to earn some extra money. Participants were made aware of a potential conflict between the two goals; accepting the part-time job would mean training fewer hours per week. See Electronic Supplementary Material (ESM 1) for the exact scenario. After reading the scenario, participants had to pass an attention check concerning the goal scenario. Specifically, participants were asked to check all statements that correctly represent the role/situation that they have been asked to imagine. Participants were given three answer options: (1) Your focal goal is to make it in the top 30 at a mathematics competition. At the same time, you are considering accepting a job at the museum to earn some extra money.

Social Psychology (2018), 49(1), 47–62

Goal Prioritization After reading the scenario, participants were asked whether they would accept or decline the job offer from the museum. If participants chose to accept the job, they were further asked to indicate how many hours (between 6 and 18 hrs in full hour increments) they would spend working as a museum tour guide. Goal Conflict Study 1 included measures of perceived goal incompatibility and more specifically, resource conflict. Goal incompatibility was assessed using an adaptation of Emmons’ (1986) scale. Specifically, participants were asked to indicate whether working as a museum tour guide has a helpful, harmful, or no effect on their focal goal to be one of the 30 best runners at the half-marathon and whether training for the half-marathon has a helpful, harmful, or no effect on their financial goal (museum job) on a 7-point scale (1 = very harmful, 4 = neither harmful nor helpful, 7 = very helpful). The scale was recoded such that greater values indicated greater goal incompatibility. In order to assess resource conflict, participants were asked how often, in their opinion, it would happen that because of the pursuit of the athletic goal (half-marathon) they would not invest as much time and energy into the financial goal (museum job) as they would like (adapted from the Intergoal Relations Questionnaire (IRQ) int; Riediger & Freund, 2004). The question was also asked the other way around (i.e., how often it may happen that they cannot pursue the financial goal because of the athletic goal). Answers were given on a 5-point scale (1 = never, 3 = sometimes, 5 = very often). Goal Satisfaction Participants reported on a 7-point scale (1 = not at all, 7 = extremely) how satisfying it would be if they achieved (a) the focal goal and (b) the secondary goal. These data were not of focal interest for this paper but were nevertheless assessed to inform a debate in the power literature

Ó 2018 Hogrefe Publishing

P. C. Schmid, Power and Goal Conflict

Table 1. Zero-order correlation coefficients (r) for all variables in Study 1 1

1. Trait power 2. Goal incompatibility

.14†

3. Resource conflict: focal goal takes resources away

.07

4. Resource conflict: secondary goal takes resources away

.17*

.26**

.04

5. Generalized self-efficacy

.57**

.12

.03

.16*

6. # Hours invested in secondary goal

.03

.02

.54**

.11

.02

4.71

3.22

2.62

3.49

3.14

11.72

1.08

1.33

1.09

1.17

0.50

4.28

†

Note. p < .10; *p < .05; **p < .01.

concerning the effect of power on reward orientation (here anticipated satisfaction). See ESM 2 for more details and analysis. Trait Power The Generalized Sense of Power Scale (C. Anderson, John, & Keltner, 2012) was used to assess individuals’ trait power. This scale includes eight items (four reverse-scored) in which participants reported to what extent they agree or disagree with several statements on a 7-point scale (1 = disagree strongly, 7 = agree strongly). A sample item was: “In my relationships with others, I can get people to listen to what I say.” Self-Efficacy The Generalized Self-Efficacy Scale is a 10-item scale designed to assess optimistic self-beliefs to cope with a variety of difficult demands in life (e.g., “I can manage to solve difficult problems if I try hard enough”; Schwarzer & Jerusalem, 1995). Participants indicated whether several statements applied to them on a 4-point scale (1 = not at all true, 4 = exactly true). Supplementary Trait Scales In order to obscure our interest in power, the trait power scale was presented along with two other trait scales that were not analyzed further: a three-item screening assessment for social phobia (Connor, Kobak, Churchill, Katzelnick, & Davidson, 2001) and the Ten-Item Personality Inventory (Gosling, Rentfrow, & Swann, 2003). Demographics, Education, and Occupation Participants indicated their educational level (no high school degree, high school degree, some college, associate’s degree, bachelor’s degree, master’s degree, doctoral degree, other) and occupational status (student, employee, owner of a business or company, self-employed, nonpaid volunteer, unemployed, retired, other). Moreover, they indicated whether they supervised other people at their job. If they did, they were also asked how many people they Ó 2018 Hogrefe Publishing

currently supervised. Participants who were currently employed also indicated where they positioned themselves in the hierarchy of their company by moving a slider between 0 (at the bottom) and 100 (at the top).

Results Preliminary Analyses The majority of participants (64%) did not supervise other people at their job. On average, participants reported occupying mid-rank positions; however, there was great variability (M = 45.09, SD = 29.74; 14 participants could not provide their hierarchical rank because they were unemployed or retired). Reliability analyses showed good internal consistency for the trait power scale (α = .909) and the generalized selfefficacy scale (α = .921). The two goal incompatibility items correlated strongly, r(195) = .48, p < .001, 95% CI [.36, .59] and therefore, a composite measure was computed with greater values indicating greater goal incompatibility. The two resource conflict items did not significantly correlate, r(195) = .04, p = .614, 95% CI [ .21, .12], and were therefore not averaged. Table 1 provides an overview over the zero-order correlations between the most relevant variables, as well as their means and SDs. Power and Experienced Goal Conflict We computed correlations to test our main hypothesis that higher trait power would be associated with reduced experience of goal conflict. As expected, greater trait power was associated (by trend) with reduced experience of goal incompatibility (i.e., pursing one goal was perceived as less harmful for the pursuit of the other goal), r(195) = .14, p = .057, 95% CI [ .01, .28]. Because the two resource conflict items did not significantly correlate, separate analyses were computed for the two items. These analyses showed that participants with greater trait power experienced that the pursuit of the secondary goal (the museum job) took Social Psychology (2018), 49(1), 47–62

relatively fewer resources away from the pursuit of the primary goal (the half-marathon), r(195) = .17, p = .015, 95% CI [ .32, .02]. Participants’ sense of power was unrelated to the extent to which they perceived that the pursuit of the primary goal took resources away from the pursuit of the secondary goal, r(195) = .07, p = .349, 95% CI [ .08, .21].1 Self-Efficacy Mediation Analysis Greater trait power was associated with greater self-efficacy, r(195) = .57, p < .001, 95% CI [.45, .67], in line with past research. While self-efficacy was not significantly associated with the experience of goal incompatibility, r(195) = .12, p = .101, 95% CI [ .02, .26], it was associated with reduced experience of resource conflict (i.e., the pursuit of the secondary goal was perceived as taking fewer resources away from the primary goal), r(195) = .16, p = .025, 95% CI [ .30, .01]. A mediation analysis was run using Hayes’ (2013) process macro with 5,000 bootstrapping resamples, trait power as the predictor, self-efficacy as the mediator, and resource conflict as the outcome measure. The mediation effect was not significant: 95% CI [ .18, .05]. Supplementary Analyses Additional analyses tested whether trait power was associated with greater goal prioritization. Goal prioritization could be expressed by rejecting the museum job offer and by investing less time into the museum job. Only five participants (2.5%) rejected the museum job offer, which was why analyses focused on the amount of time they would invest into the museum job. No evidence for prioritization was found; trait power was not related to the number of hours participants planned to invest into the museum job, r(195) = .033, p = .647, 95% CI [ .19, .13]. These findings suggest that in this study, the relationship between trait power and reduced goal conflict could not be explained by greater goal prioritization.

Discussion As expected, individuals with relatively greater trait power experienced less conflict between two approach goals in a scenario in terms of experienced goal incompatibility (marginal effect) as well as in terms of experienced conflict for resources (i.e., they experienced that the pursuit of the secondary goal would take less resources away from the focal goal). Presenting participants with a standardized goal 1

P. C. Schmid, Power and Goal Conflict

conflict scenario had the advantage that the goal conflict was held constant across all participants. As such, we could show that trait power was associated with greater subjectively experienced goal conflict, independent of actual goal conflict. However, the disadvantage of using scenarios is that goals may not be relevant to participants, which may affect their capability to estimate how much conflict they would experience. In Studies 2–4, we therefore focused on resource conflict between participants’ personal goals. Although greater trait power was associated with greater generalized self-efficacy and self-efficacy was in turn related to reduced experience of resource conflict, selfefficacy did not act as a mediator in this study. It is possible that this was the case because self-efficacy was measured as a generalized tendency rather than specific to the task or goal to achieve – a hypothesis that was tested in Study 2.

Study 2 The main goal of Study 2 was to further examine the relationship between trait power and the experience of resource conflict between important personal goals. When using scenarios in Study 1, experienced goal conflict was hypothetical, and the extent to which participants felt conflicted in this situation may have depended on the extent to which participants could relate to the described goals. By focusing on potential conflict between personal goals, we may show that people who experience greater power actually experience less goal conflict in their lives. In this study, we did not specify whether participants should focus on their approach or avoidance goals to ensure that participants thought of goals that were personally important to them. It is however possible that individuals’ level of trait power is related to the type of goals they pursue. Indeed, high power has been associated with increased motivation to approach positive states and low power with greater motivation to avoid negative states (Keltner, Gruenfeld, & Anderson, 2003; Smith & Bargh, 2008). To control for this possibility, participants in Study 2 were asked to indicate whether their goals pertained to pursuing something (approach goal) or avoiding something (avoidance goal). Participants’ current goal progress was also assessed. This was done because people tend to show increased motivation and effort when they are close to goal completion (e.g., Brown, 1948; Förster, Higgins, & Idson, 1998; Hull, 1932), which could lead to greater experience of resource conflict.

This correlation was not significant possibly because participants took the pursuit of the focal goal as a given and focused on whether the secondary goal took away resources. Indeed, a t-test showed that participants perceived that the secondary goal took away more resources from the focal goal (M = 3.49, SD = 1.17), than the focal goal did from the secondary goal (M = 2.62, SD = 1.09), t(196) = 7.51, p < .001, 95% CI [ 1.10, 0.64].

Social Psychology (2018), 49(1), 47–62

Ó 2018 Hogrefe Publishing

P. C. Schmid, Power and Goal Conflict

As in Study 1, we tested whether the link between trait power and reduced resource conflict could be explained by self-efficacy. Here, self-efficacy was indexed as participants’ confidence that they could achieve their personal goals if they tried hard enough, which may be more strongly linked to experienced conflict between the personal goals than the general self-efficacy beliefs that were assessed in Study 1.

Method Participants Participants were 202 US citizens recruited through mTurk and paid $1 for participation. Five participants were excluded from analyses; one participant did not report his goals (i.e., entered meaningless words), and four participants had outlying scores on the main dependent variable of goal conflict (i.e., values exceeded 1.5 the interquartile range). Of the remaining sample, 44.7% were female and average age was 34.22 years, SD = 9.70. Procedure Participants provided consent before listing the three most important personal goals that they currently pursued. Next, they reported to what extent their three goals competed for the same resources and answered questions concerning the nature of their goals. Finally, participants completed the trait power scale and other trait scales in randomized order, and provided information about their demographics, education, and current occupation. Goal Conflict Participants listed the three most important goals that they currently pursued. These goals were paired with one another, resulting in three different goal pairs, and resource conflict was subsequently assessed. Specifically, for each of the three pairs of goals, participants indicated how much these goals were in conflict with one another such that they may not be able to invest as much time and effort into the pursuit of one goal because of the pursuit of the other goal on a 7-point scale (1 = not at all conflicting, 7 = very much conflicting). Goal Satisfaction As in Study 1, participants were asked to indicate how satisfying it would be for them to attain each of their three goals. Analyses are presented in the ESM 2. Goal Characteristics To control for goal characteristics, participants indicated how much progress they had already made on each of their three goals by moving a slider on a scale from 0 to 100, and rated to what extent each of their goals described avoiding Ó 2018 Hogrefe Publishing

something versus pursuing something on a 7-point scale (1 = avoiding, 4 = neither avoiding nor pursuing, 7 = pursuing). Trait Power See Study 1. Goal-Related Self-Efficacy For each of their personal goals, participants indicated how likely they would achieve their goal if they tried hard enough on a 7-point scale (1 = very unlikely, 7 = very likely). Trait Scales See Study 1. Demographics, Education, and Occupation See Study 1.

Results Preliminary Analyses Most participants (62%) did not supervise other people at their work, and on average they held a mid-rank hierarchical position (M = 43.73, SD = 28.42; 18 participants were unemployed or retired and could not provide their hierarchical rank). The most frequent goals participants reported were losing weight, exercising, paying loans and saving money, completing an education, finding a job, and buying a house. Reliability analyses showed acceptable internal consistency for the trait power scale (α = .918), the goal conflict items (α = .665), the goal-related self-efficacy items (α = .729), and the goal satisfaction items (α = .764). Items were therefore averaged for the respective variables. Table 2 shows the zero-order correlations between the relevant variables, as well as their means and SDs. Power and Experienced Goal Conflict We first tested our main hypothesis that greater trait power would be associated with reduced experience of resource conflict between goals. This was indeed the case, r(195) = .15, p = .037, 95% CI [ .28, .01]. Next, a Table 2. Zero-order correlation coefficients (r) for all variables in Study 2 1

1. Trait power 2. Resource conflict

.15*

3. Goal-related self-efficacy

.28**

.17*

4. Goal progress

.19**

.01

5. Avoidance-approach goal

.11

.13†

.19** .45**

.09

4.60

2.38

6.08

37.86

5.85

1.22

1.39

1.00

19.72

1.31

Note. †p < .10; *p < .05; **p < .01.

Social Psychology (2018), 49(1), 47–62

regression analysis was computed with trait power as a predictor (mean-centered), resource conflict as the outcome, and goal characteristics as control variables (i.e., goal progress and goal ratings on the avoidance-approach dimension; all mean-centered). Results showed that the effect of trait power on resource conflict remained significant, β = .17, t = 2.03, p = .044, 95% CI [ .33, .004]. Self-Efficacy Mediation Effects In line with past research, greater trait power was associated with increased self-efficacy, r(195) = .28, p < .001, 95% CI [.13, .42]. Greater self-efficacy was in turn associated with reduced experience of resource conflict, r(195) = .17, p = .018, 95% CI [ .31, .04]. Using Hayes’ (2013) process macro with 5,000 bootstrapping resamples, a significant mediation effect emerged, 95% CI [ .10, .002], showing that trait power was associated with enhanced feelings of self-efficacy in relation to the pursuit of participants’ own goals, b = .23, SE = .06, t = 4.07, p < .001, which in turn was marginally related to reduced experience of resource conflict, b = .19, SE = .10, t = 1.87, p = .062.

Discussion The primary result of Study 2 was that people with relatively greater sense of power experienced their own goals as less in conflict with one another in terms of resources. This association remained stable after controlling for current goal progress and avoidance-approach characteristics of the goal. As such, this study provided further evidence for our hypothesis that greater sense of power is associated with the experience of less conflict between goals. It remains open whether this effect is causal or just correlational. Unlike in Study 1, Study 2 showed that self-efficacy significantly mediated the link (albeit marginally) between trait power and experienced resource conflict. Importantly, in Study 1, self-efficacy was measured as a generalized state and not specific to the pursuit of the conflicting goals, while in Study 2, self-efficacy was assessed with respect to participants’ listed goals. This could explain the inconsistent effects. An alternative explanation is that self-efficacy plays a more prominent role when conflict happens between goals that people actually pursue (as was the case in Study 2) rather than between imagined goals that participants may not necessarily want to pursue (as was the case in Study 1). This possibility was tested in Study 3.

Study 3 Study 3 was conducted in the laboratory and aimed to examine whether there is a causal effect between people’s Social Psychology (2018), 49(1), 47–62

P. C. Schmid, Power and Goal Conflict

sense of power and experienced conflict between personal goals. Study 3 also aimed to clarify the role of self-efficacy in the effect of power on experienced goal conflict. In this study, we focused again on generalized self-efficacy to see whether it mediated the effect of power on experienced goal conflict when the goal conflict included goals that participants personally pursued.

Method Participants We recruited 113 students through the participant pool of ETH Zurich and the University of Zurich. Twelve participants were excluded from analyses: five did not understand their assigned role (i.e., three low-power participants indicated that they were in charge in their role and two high-power participants did not feel in charge in their role; these participants were also outliers on the power manipulation check), two had outlying scores exceeding 1.5 the interquartile range on the main dependent variable goal conflict, three indicated that they had already attained at least one of the five goals (these already attained goals cannot currently conflict with the other goals), and two showed no variability in their questionnaire answers (i.e., reported 1 on all items), which suggested noncompliance. Of the remaining participants, 63.5% were female and mean age was 23.23 years, SD = 5.37. Procedure Participants were run in group sessions of 18–33 individuals. After providing consent, participants listed their five most important personal goals on a sheet of paper. Next, participants were randomly assigned to the high-power or low-power condition, and their power was manipulated accordingly. After the power manipulation, participants put the list with their personal goals next to them and answered a series of questions on the computer concerning the nature of each goal and the extent to which they experienced resource conflict between their goals. Using this procedure (i.e., reporting goals before power was manipulated) ensured that participants’ personal goals were not affected by the power manipulation, and thus, any potential effects of the power manipulation on perceived resource conflict likely reflected a change in participants’ subjective experience of resource conflict. Finally, participants completed the manipulation check and several trait/state scales, and provided demographics. Power Manipulation Power was manipulated through an essay-writing task (e.g., as in Dubois, Rucker, & Galinsky, 2010). All participants were asked to type a short text ( 600 characters) about how they would feel in a specific role. In the high-power Ó 2018 Hogrefe Publishing

P. C. Schmid, Power and Goal Conflict

condition, participants were asked to imagine themselves as a boss at a firm who is evaluating subordinates and the productivity of the team they supervise. In the low-power condition, participants imagined themselves as employees waiting to receive their evaluations from their superiors. To ensure that participants understood the roles they were asked to imagine in terms of their level of power, they were asked to indicate to what extent they agreed with the following two statements on a 5-point scale (1 = not at all, 5 = very much): “In my role, I had influence over others,” and “In my role, I was in charge.” Moreover, following past research (e.g., A. Anderson & Galinsky, 2006; Galinsky et al., 2003; Schmid, Kleiman, et al., 2015), the essays’ contents were coded on expression of feelings of power/powerlessness using a 5-point scale (1 = not at all, 5 = very much). Two independent coders first rated the essays of the first 30 participants. There was acceptable interrater reliability (for expressions of highpower feelings, r = .84; and for expressions of low-power feelings, r = .81); thus, a single coder completed the coding of all remaining essays. Goal Conflict Resource conflict was assessed by pairing participants’ five personal goals with one another, resulting in 10 different pairs of goals. For each pair, participants were asked two questions: “How often can it happen that because of goal A you cannot invest as much time, money, and energy in goal B as you would like?” and “How often can it happen that because of goal B you cannot invest as much time, money, and energy in goal A as you would like?” Participants responded on a 5-point scale (1 = never, 5 = very often). Goal Characteristics For each goal, participants completed the same measures of goal progress and avoidance-approach characteristics as in Study 2. In addition, time sensitivity was assessed by asking participants whether they had a deadline (official or selfdetermined) for the attainment of their goals on a 6-point scale (1 = the deadline is very close, 6 = there is no deadline), assuming that there would be more goal conflict if participants pursued goals with close deadlines. These measures reflect goal characteristics and served as control variables. Trait Power See Study 1. Generalized Self-Efficacy See Study 1. Self-Control and Mood In addition, participants completed the 10-Item Self-Scoring Self-Control Scale (Tangney, Baumeister, & Boone, 2004), as well as a mood scale. These measures were not central to Ó 2018 Hogrefe Publishing

Table 3. Zero-order correlation coefficients (r) between trait scales and dependent variables in Study 3 1

1. Trait power 2. Resource conflict

.01

3. Generalized self-efficacy 4. Goal progress

.38**

.02

.13

.01

.09

.07

.03

.15

.06

.07

.32**

.12

.11

5. Avoidance-approach goal 6. Time sensitivity

.05

5.01

2.17

2.91 38.21

5.77 4.12

0.78

0.48

0.45 11.99

0.92 1.01

Note. †p < .10; *p < .05; **p < .01.

the aim of this research; we added these variables for exploratory reasons. See ESM 3 for more details and analyses.

Results Preliminary Analyses Students’ goals predominantly consisted of finishing their studies, doing well on exams, losing weight, and exercising. Reliability analysis showed acceptable internal consistency for the trait power scale (α = .769), the generalized self-efficacy scale (α = .852), and the goal conflict items (α = .711). Items were therefore averaged for the respective variables. See Table 3 for correlation coefficients between trait measures and dependent variables as well as means and SDs. To confirm that we successfully manipulated participants’ feelings of power, t-tests were conducted on participants’ expressions of feelings of power and powerlessness in their essays. High-power participants expressed significantly greater feelings of power (M = 3.54, SD = 1.07) than low-power participants (M = 1.40, SD = 0.68), t(99) = 12.20, p < .001, 95% CI [ 2.49, 1.79], r = .77. High-power participants also expressed significantly lower feelings of powerlessness in their essays (M = 1.74, SD = 0.91) than low-power participants (M = 2.76, SD = 0.90), t(99) = 5.68, p < .001, 95% CI [0.67, 1.38], r = .50. Power and Experienced Goal Conflict Our main hypothesis concerned the causal effect of power on experienced resource conflict. As expected, high-power participants experienced their goals as less conflicting (M = 2.06, SD = .45) compared to low-power participants (M = 2.26, SD = .49), t(99) = 2.13, p = .036, 95% CI [.01, .39], r = .21. An analysis of covariance (ANCOVA) showed that power effects remained significant after controlling for goal characteristics (i.e., goal progress, ratings on the avoidance-approach dimension, and time sensitivity), F(1, 96) = 4.11, p = .045, ηp2 = .04. Social Psychology (2018), 49(1), 47–62

Self-Efficacy Mediation Effects Power did not significantly affect self-efficacy, t(99) = 0.24, p = .813, 95% CI [ .20, .16], r = .02, suggesting that selfefficacy did not mediate the power effect on experienced goal pursuit.

Discussion While Studies 1 and 2 showed that one’s psychological power across different life situations (i.e., trait power) was associated with reduced experience of goal conflict, Study 3 provided evidence for a causal effect – situationally induced high power reduced the experience of resource conflict compared to low power. This suggests that individuals’ experience of resource conflict can be reduced by making them feel powerful. Because participants first listed their goals on a sheet of paper, then their power was manipulated, and finally they took their sheet of paper next to them and completed questions about their goals, participants’ goals (and thus actual conflict between their goals) could not have been altered as a consequence of the manipulation. Thus, this study provides further evidence that power alters individuals’ subjective experience of goal conflict independent of actual goal conflict. A potential disadvantage of this procedure is that it may have caused some disruption in the flow of the experiment: participants were asked to change from thinking about goals to thinking about power and then again about their goals, which made them also change from doing tasks on paper to the computer and back. Cai and Guinote (2017) showed that low power hinders people’s ability to switch between tasks, and it is thus possible that the procedure of this study already evoked some experience of goal conflict in low-power participants. Study 3 further probed the role of generalized selfefficacy in the power effect on reduced goal conflict. However, the mediation effect was not significant. Together with findings from Study 1, this suggests that generalized self-efficacy cannot explain the effect of power on reduced experience of goal conflict. Findings of Study 2, however, showed a significant (although weak) mediation effect of self-efficacy when it was assessed in relation to the pursuit of personal goals. In Study 4, measures of both generalized and goal-related self-efficacy were included, in order to test for the robustness of the effects.

Study 4 The goal of Study 4 was threefold: (1) we wanted to provide further evidence for a causal effect of power on the experience of goal conflict, Social Psychology (2018), 49(1), 47–62

P. C. Schmid, Power and Goal Conflict

(2) we included a control group to inform the literature about the potential direction of the power effect, and (3) we investigated potential behavioral consequences of the power effect on experienced goal conflict. Moreover, in Study 4, we decided to have a more fluent procedure in which all measures were assessed online and the power manipulation was separated from the sections that involved personal goals. This choice was made to reduce the possibility that the procedure of the experiment itself would evoke greater goal conflict in low-power participants because they are less flexible in switching between tasks (Cai & Guinote, 2017). Past research has shown that when people experience goal conflict, they invest less into the pursuit of their goals (Boudreaux & Ozer, 2013; Emmons & King, 1988). In Study 4, we therefore examined whether the effect of power on experienced goal conflict has consequences for individuals’ goal commitment in the future. We hypothesized that high-power participants would experience less goal conflict than low-power participants, and, consequently, plan to invest more into the pursuit of their goals in the future and pursue their goals more actively in the future. We did not have hypotheses concerning comparisons with the control group. Past research on power and goal pursuit revealed inconsistent results in terms of whether high power facilitated goal pursuit relative to the control group or whether low power hindered it – even in studies where methods were quite comparable (cf. Schmid, Kleiman, et al., 2015; Smith et al., 2008). This may be the case because it is unclear how the control condition should look in order to provide a good comparison (see Schmid, Kleiman, et al., 2015, for a similar argument). However, by including a control group in the present study and comparing results with past work, we can help clarify this question.

Method Participants Study 4 included 345 participants recruited over mTurk and paid $2 for participation. Similar to the previous studies, 14 participants were excluded because they had outlying scores exceeding 1.5 the interquartile range on the goal conflict variables and five participants did not understand the power or lack of power associated with their assigned role (i.e., four low-power participants indicated that they were in charge in their role and one high-power participant did not feel in charge in the assigned role). Of the remaining participants, 57.4% were female and mean age was 35.07 years, SD = 11.63. Ó 2018 Hogrefe Publishing

P. C. Schmid, Power and Goal Conflict

Procedure After providing consent, participants’ power was manipulated. As part of an online survey, participants then listed their three most important personal goals and indicated to what extent their goals competed for the same resources and were incompatible. Next, they indicated how self-efficacious they felt concerning the attainment of their goals, answered several questions concerning the nature of their goals, and indicated their behavioral intentions concerning the pursuit of their goals in the near future. Finally, participants completed the generalized self-efficacy scale, manipulation checks, and provided information about their demographics, education, and current occupation. Power Manipulation Power was manipulated using the same essay-writing task as in Study 3. Moreover, a control group was added that was instructed to imagine and write a short text about the experience of being in a supermarket and purchasing groceries for the coming week. The minimum length of the essays was shortened to 300 characters. We used the same manipulation checks as in Study 3, that is, we asked whether they understood their role and coded essays on content expressing feelings of power and powerlessness. Interrater reliability was acceptable for the first 30 essays (for expressions of high-power feelings, r = .80; and for expressions of low-power feelings, r = .83); thus, a single coder completed the coding of all remaining essays. Goal Conflict In Study 4, resource conflict was assessed as in Study 2. Moreover, to assess goal incompatibility, participants were additionally asked whether the pursuit of one goal had a helpful, harmful, or no effect on the pursuit of the other goal for all three pairs of goals in both directions (e.g., asking for the effect of the pursuit of goal A on goal B, as well as the effect of the pursuit of goal B on goal A). Answers were given on a 7-point scale (1 = very harmful, 4 = neither harmful nor helpful, 7 = very helpful). Scores were averaged and recoded such that greater values indicated greater goal incompatibility. Goal-Related Self-Efficacy See Study 2. Goal Characteristics To control for goal characteristics, participants completed the same measures of goal progress, avoidance-approach characteristics of the goal, and time sensitivity as in Study 3. Generalized Self-Efficacy See Study 1. Ó 2018 Hogrefe Publishing

Behavioral Intentions Participants further answered two questions concerning their behavioral intentions in relation to each of their three goals. The first item was: “How much time and energy do you plan to invest into the pursuit of your goal in the near future?” Answers were given on a 7-point scale (1 = very little, 7 = a lot). The second item was: “To what extent do you plan to pursue your goal actively in the near future?,” again using a 7-point scale (1 = not at all, 7 = very much). Demographics, Education, and Occupation See Study 1.

Results Preliminary Analyses In this sample, 66% of all participants did not supervise other people at their job. On average, they occupied a mid-level hierarchical rank in their firm, M = 43.16, SD = 30.73. Internal consistency was acceptable for the generalized self-efficacy scale (α = .910), the resource conflict items (α = .602), the goal incompatibility items (α = .763), and the behavioral intention items (for planned resource investment, α = .689 and for planned active goal pursuit, α = .669). Items were therefore averaged for the respective variables. The two measures assessing goal-relevant behavioral intentions correlated highly, r(324) = 74, p < .001, and were therefore combined (i.e., averaged). Means, SDs, and correlation coefficients between trait measures and dependent variables are presented in Table 4. To test whether we successfully manipulated power, oneway analyses of variance (ANOVAs) were conducted on participants’ expressions of feelings of power and powerlessness in their essays. The manipulation was effective: the power effect on expressions of power feelings was significant, F(2, 323) = 217.03, p < .001, ηp2 = .57. Highpower participants expressed significantly greater feelings of power (M = 3.76, SD = 0.91) than low-power participants (M = 1.63, SD = 0.92), t(323) = 18.21, p < .001, 95% CI [ 2.36, 1.90], r = 0.71, and control participants (M = 1.64, SD = 0.70), t(323) = 18.25, p < .001, 95% CI [1.90, 2.36], r = .71. The difference between low-power participants and control participants was not significant, t(323) = 0.11, p = .907, 95% CI [ .24, .21], r = .01. The power condition effect on expressions of powerlessness was also significant, F(2, 323) = 107.38, p < .001, ηp2 = .40, showing that high-power participants expressed significantly lower feelings of powerlessness in their essays (M = 1.31, SD = 0.71) than low-power participants (M = 2.78, SD = 1.06), t(323) = 12.51, p < .001, 95% CI [1.24, 1.71], r = .57. Control participants (M = 1.33, SD = 0.75) expressed significantly less feelings of powerlessness than low-power Social Psychology (2018), 49(1), 47–62

P. C. Schmid, Power and Goal Conflict

Table 4. Zero-order correlation coefficients (r) between trait scales and dependent variables in Study 4 1

1 Resource conflict 2. Goal incompatibility

.44**

3. Planned goal investment

.06

.21**

5. Generalized self-efficacy

.06

.15**

.36**

6. Goal-Related self-efficacy

.06

.19**

.48**

.45**

7. Goal progress

.01

.06

.22**

.14**

.30**

8. Avoidance-approach goal

.12*

.14*

.40**

.21**

.37**

.00

9. Time sensitivity

.06

.00

.01

.04

.05

.08

.03

2.63

2.18

5.93

3.26

6.06

42.43

6.22

4.98

1.49

1.11

1.01

0.50

1.15

20.61

1.06

1.46

Note. *p < .05; **p < .01.

participants, t(323) = 12.79, p < .001, 95% CI [1.23, 1.68], r = .58. The high-power group and the control group did not significantly differ in their expressions of powerlessness, t(323) = 0.21, p = .837, 95% CI [ .25, .21], r = 0.01. Power and Experienced Goal Conflict To test whether power influenced individuals’ experience of goal conflict, we computed separate one-way ANOVAs for resource conflict and goal incompatibility. The power condition effect on resource conflict was not significant, F(2, 323) = 0.78, p = .458, ηp2 = .01. However, power condition significantly altered individuals’ experience of goal incompatibility, F(2, 323) = 10.34, p < .001, ηp2 = .06. As expected, high-power participants experienced their goals as less incompatible (M = 2.01, SD = 1.16) compared to low-power participants (M = 2.56, SD = 0.84), t(323) = 3.72, p < .001, 95% CI [.26, .85], r = .20. Control participants also experienced less goal incompatibility (M = 1.97, SD = 1.21) than low-power participants, t(323) = 4.11, p < .001, 95% CI [.31, .87], r = .22. The difference between high-power participants and controls was not significant, t(323) = 0.24, p = .808, 95% CI [ .33, .25], r = .05. An ANCOVA showed that power effects on goal incompatibility remained significant after controlling for goal characteristics (i.e., goal progress, ratings on the avoidance-approach dimension, and time sensitivity), F(2, 320) = 11.01, p < .001, ηp2 = .06. Self-Efficacy Mediation Effects To examine whether generalized self-efficacy and goal-related self-efficacy mediated the power effect on experienced goal conflict, we first conduced ANOVAs on self-efficacy. However, power condition did not significantly affect generalized self-efficacy, F(2, 323) = 1.08, p = .340, ηp2 = .01, nor goal-relevant self-efficacy,

Social Psychology (2018), 49(1), 47–62

F(2, 323) = 0.59, p = .558, ηp2 < .01. The mediation models were therefore not calculated. Consequences for Behavioral Intentions We tested whether the power effect on experienced goal incompatibility had consequences for individuals’ behavior intentions. As expected, experienced goal incompatibility was correlated with planned investment of resources, r(324) = .25, p < .001, 95% CI [ .35, .13]. Next, we tested the simple mediation model using Hayes’ (2013) process macro with 5,000 bootstrapping resamples. Power was entered dummy-coded (0 = low power, 1 = high power). The mediation effect was significant, 95% CI [.05, .29] and showed that low power increased the experience of goal conflict, which resulted in fewer plans to invest into the pursuit of their goals in the near future (see Figure 1).

Discussion Study 4 showed that low-power participants experienced their goals as more incompatible with one another than high-power and control participants. High-power participants and controls did not significantly differ. The lowpower effect on experienced incompatibility of goals had consequences for behavioral intentions: the more participants felt their goals were incompatible, the less they planned to invest in the pursuit of those goals in the near future. These findings are consistent with past work that suggested that the experience of conflict between one’s goal may be associated with decreased goal investments (Emmons & King, 1988). Surprisingly, power did not significantly affect participants’ experience of resource conflict. Moreover, selfefficacy did not mediate the power effect on goal conflict in this study – neither when it was measured with a goalspecific scale, nor when it was measured as a general tendency.

Ó 2018 Hogrefe Publishing

P. C. Schmid, Power and Goal Conflict

Figure 1. Mediation model illustrating that low power increased the experience of goal conflict, which resulted in fewer planned goal investments. Unstandardized coefficients and SEs in parentheses are presented. ***p < .0001.

General Discussion The main goal of this research was to investigate the relationship between individuals’ sense of power and their experience of multiple goals as being incompatible and competing for the same resources. As predicted, results of four studies showed that lower sense of power was related to enhanced experience of goal conflict. This result manifested both when measuring sense of power as a trait (Study 1 and Study 2) and when experimentally manipulating power (Study 3 and Study 4), providing evidence for a causal relationship between low power and the experience of goal conflict. Power predicted reduced experience of goal conflict when actual goal conflict was held constant across participants (i.e., by having participants imagine a specific conflict between two goals) as well as when looking into conflict between participants’ personal goals while controlling for goal characteristics. Study 4 provided further evidence that the low-power effect on goal conflict has consequences for behavioral intentions. A mediation analysis revealed that low power enhanced the experience of goal incompatibility, which in turn was associated with having fewer plans to invest in the pursuit of their goals in the near future. These findings suggest that the powerless may persist less on their multiple goals because they experience conflict between the goals. As such, our results may provide a new explanation for past findings that suggest that high power facilitates goal pursuit relative to low power (Galinsky et al., 2003; Guinote, 2007b).

An Effect of Low Power? Study 4’s findings point to an effect of low power (which significantly differed from the control group) rather than of high power (which did not significantly differ from the control group). However, as highlighted before, it is not unusual that control groups fluctuate in power research, Ó 2018 Hogrefe Publishing

sometimes suggesting that effects are driven by the highpower group and other times by the low-power group. It is therefore possible that findings based on comparisons with control groups depend on the sample and the experimental settings and do not necessarily reflect a direction of the effect. Thus, our results should not be over-interpreted. Nevertheless, including control groups in power research can be helpful to determine the factors that influence whether effects are stronger on the high-power side or the low-power side. Eventually, the field can be informed about whether these control groups provide meaningful results and should be included at all, and if yes, how comparison groups should look like to be most informative.

The Role of Self-Efficacy Another aim of this research was to examine whether the power effect on reduced goal conflict could be explained by heightened self-efficacy beliefs. Past research has shown that although power increased self-efficacy, the effect of power on cognition and behavior could not be explained by differences in self-efficacy (A. Anderson & Galinsky, 2006; Schmid & Schmid Mast, 2013; Slabu & Guinote, 2010; Weick & Guinote, 2010). In the present research, self-efficacy (generalized and goal-related) also did not mediate the power effect on experienced goal conflict in Studies 1, 3, and 4. In Study 2, a significant but weak selfefficacy mediation effect was found, in which greater sense of power was associated with increased self-efficacy to attain one’s personal goals, which in turn was marginally related to reduced experience of conflict between these goals. Taken together, our findings and past research suggest that power may be related to heightened self-efficacy (at least in some contexts), but although self-efficacy plays an important role in goal pursuit, power effects on goal pursuit processes are seemingly not well explained by self-efficacy. It is important to note that we examined self-efficacy as a general feeling that one is able to achieve goals, and the specific belief that one is able to achieve one’s most important goals. We did, however, not directly ask for selfefficacy beliefs to resolve goal conflict. Maybe a clearer effect would have emerged had we focused on people’s beliefs to deal with conflict rather than their beliefs to be able to pursue single goals. It is also possible that it is not self-efficacy per se that explains the power effect on experienced goal conflict, but related and more specific constructs such as perceived personal control and subjective experience of available resources. Psychological power has been previously associated with greater subjective experience of personal control (Fast et al., 2009) and greater perception of resources that one is ready to deal Social Psychology (2018), 49(1), 47–62

with demanding situations (Scheepers et al., 2012). These constructs – perceived control and perception of resources – are somewhat related to the concept of self-efficacy and may be more directly linked to the experience of resource conflict.

Future Directions Our results consistently showed that lower power was related to greater experience of resource conflict, with the consequence that they planned to invest in the pursuit of this goal in the future. Moreover, this research started to examine potential mechanisms by exploring the roles of self-efficacy, self-control, and mood in this effect. Some questions concerning the underlying processes remain open for future research. For instance, the role of goal prioritization in the power effect on experienced goal conflict needs to be clarified. According to the social distance theory of power (Magee & Smith, 2013), power makes people focus on and prioritize superordinate goals. If the powerful put their multiple goals in a hierarchical structure, this may explain why they experience less goal conflict. Thus, their strategy would be a goal-shielding strategy (see Shah et al., 2002) in which they prioritize and focus on one goal while inhibiting alternative goals. In Study 1, goal prioritization was assessed by asking participants how many hours they would invest in the secondary goal. Participants’ trait power was not significantly related to this measure. Although this study did not provide evidence for the prioritization of the focal goal at the expense of the secondary goal, it also did not rule out such a possibility. It is possible that the goal conflict presented in the scenario in Study 1 was insufficient to produce a prioritizing effect. Indeed, past research has shown that when they had sufficient cognitive resources to perform two tasks at the same time, high-power participants adopted a multitasking strategy (Schmid, Schmid Mast, et al., 2015); only when resources were insufficient to pursue two tasks simultaneously did high-power participants (but not low-power participants) prioritize one task over the other. It would also be interesting to investigate how the psychological experience of power affects people’s experience of conflict in specific goal conflict situations. It is possible that power would affect the experience of goal conflict differently depending on whether goals pertain to approach or avoidance goals. According to the approach/inhibition theory of power (Keltner et al., 2003), the powerful are thought to focus on pursuing rewards, while the powerless are thought to be more focused on avoiding non-rewards. Thus, the powerful may experience more conflict in approachapproach conflicts because they are more motivated to pursue approach goals, while the powerless may experience

Social Psychology (2018), 49(1), 47–62

P. C. Schmid, Power and Goal Conflict

more conflict in avoidance-avoidance conflicts, because they care more about avoiding bad situations. Study 1 focused on an approach-approach conflict, while in Studies 2–4 the type of conflict was not specified. However, participants indicated to what extent their goals were approach versus avoidance goals and on average, they listed more approach goals. Thus, approach-approach conflicts were more common in the present research. If the powerless are indeed more motivated to avoid bad situations than the powerful, then the power effects on reduced experience of conflict should be even stronger in avoidance-avoidance conflicts. Future research may also further clarify the consequence of the power effect on reduced resource conflict for performance and goal attainment. Past research suggests that individuals who experience high-power perform better in many different tasks (Kang, Galinsky, Kray, & Shirako, 2015; Lammers, Dubois, Rucker, & Galinsky, 2013; Magee et al., 2007; Schmid, Kleiman, et al., 2015; Schmid & Schmid Mast, 2013; Smith et al., 2008), including tasks that require multitasking (Cai & Guinote, 2017). Is this because the high-power participants experienced less conflict between their goals? Our results in Study 4 indeed suggest that the powerful plan to invest more into their goals, which may lead to better performance. However, more research that directly addresses this question is required.

Conclusions Our main finding was that low power increased individuals’ experience of resource conflict relative to high power, which in turn may lead to fewer plans of goal-directed behavior in the near future. This result significantly advances the literature on power that has until now focused on the effect of power on the pursuit of single goals. While past research showed that high power facilitates the pursuit of a single goal relative to low power, the current research suggests that feeling powerless may also hinder goal pursuit in multiple goal situations by decreasing one’s experience of goal conflict. By focusing on subjective experiences rather than performance outcomes, this research also takes a different angle in studying the power effect on goal pursuit. More broadly, in light of past research linking the experience of goal conflict with hesitant goal pursuit and negative affect and stress (Emmons & King, 1988), this research also provides important insights into the effect of power on effective goal pursuit and people’s affectivity. Electronic Supplementary Material The electronic supplementary material is available with the online version of the article at https://doi.org/10.1027/ 1864-9335/a000327 Ó 2018 Hogrefe Publishing

P. C. Schmid, Power and Goal Conflict

ESM 1. Text (.docx). Goal conflict scenario used in Study 1. ESM 2. Text (.docx). Description and analyses of the goal satisfaction variable in Studies 1 and 2. ESM 3. Text (.docx). Description and analyses of mood and self-control in Study 3.

References Abele, A. E., & Spurk, D. (2009). The longitudinal impact of selfefficacy and career goals on objective and subjective career success. Journal of Vocational Behavior, 74, 53–62. https://doi. org/10.1016/j.jvb.2008.10.005 Anderson, A., & Galinsky, A. D. (2006). Power, optimism, and risktaking. European Journal of Social Psychology, 36, 511–536. https://doi.org/10.1002/ejsp.324 Anderson, C., John, O. P., & Keltner, D. (2012). The personal sense of power. Journal of Personality, 80, 313–344. https://doi.org/ 10.1111/j.1467-6494.2011.00734.x Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman. Botvinick, M., Braver, T., Barch, D., Carter, C., & Cohen, J. (2001). Conflict monitoring and cognitive control. Psychological Review, 108, 624–652. https://doi.org/10.1037/0033-295X.108.3.624 Boudreaux, M. J., & Ozer, D. J. (2013). Goal conflict, goal striving, and psychological well-being. Motivation and Emotion, 37, 433–443. https://doi.org/10.1007/s11031-012-9333-2 Brown, J. S. (1948). Gradients of approach and avoidance responses and their relation to level of motivation. Journal of Comparative and Physiological Psychology, 41, 450–465. https://doi.org/10.1037/h0055463 Cai, R. A., & Guinote, A. (2017). Doing many things at a time: Lack of power decreases the ability to multitask. British Journal of Social Psychology, 56, 475–492. https://doi.org/10.1111/bjso. 12190 Carver, C. S., & Scheier, M. F. (1998). On the self-regulation of behavior. New York, NY: Cambridge University Press. Cinamon, R. G. (2006). Anticipated work-family conflict: Effects of gender, self-efficacy, and family background. The Career Development Quarterly, 54, 202–215. https://doi.org/10.1002/ j.2161-0045.2006.tb00152.x Connor, K. M., Kobak, K. A., Churchill, L. E., Katzelnick, D., & Davidson, J. R. T. (2001). Mini-SPIN: A brief screening assessment for generalized social anxiety disorder. Depression and Anxiety, 14, 137–140. https://doi.org/10.1002/da.1055 Dubois, D., Rucker, D. D., & Galinsky, A. D. (2010). The accentuation bias: Money literally looms larger (and sometimes smaller) to the powerless. Social Psychological and Personality Science, 1, 199–205. https://doi.org/10.1177/1948550610365170 Emmons, R. A. (1986). Personal strivings: An approach to personality and subjective well-being. Journal of Personality and Social Psychology, 51, 1058–1068. https://doi.org/10.1037/ 0022-3514.51.5.1058 Emmons, R. A., & King, L. A. (1988). Conflict among personal strivings: Immediate and long-term implications for psychological and physical well-being. Journal of Personality and Social Psychology, 54, 1040–1048. https://doi.org/10.1037/ 0022-3514.54.6.1040

Ó 2018 Hogrefe Publishing

Fast, N. J., Gruenfeld, D. H., Sivanathan, N., & Galinsky, A. D. (2009). Illusory control: A generative force behind power’s far-reaching effects. Psychological Science, 20, 502–508. https://doi.org/10.1111/j.1467-9280.2009.02311.x Fast, N. J., Sivanathan, N., Mayer, N. D., & Galinsky, A. D. (2012). Power and overconfident decision-making. Organizational Behavior and Human Decision Processes, 117, 249–260. https://doi.org/10.1016/j.obhdp.2011.11.009 Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39, 175–191. https://doi.org/10.3758/BF03193146 Förster, J., Higgins, E. T., & Idson, L. C. (1998). Approach and avoidance strength during goal attainment: Regulatory focus and the “goal looms larger” effect. Journal of Personality and Social Psychology, 75, 1115–1131. https://doi.org/10.1037/ 0022-3514.75.5.1115 Fujita, K., Trope, Y., Liberman, N., & Levin-Sagi, M. (2006). Construal levels and self-control. Journal of Personality and Social Psychology, 90, 351–367. https://doi.org/10.1037/00223514.90.3.351 Galinsky, A. D., Gruenfeld, D. H., & Magee, J. C. (2003). From power to action. Journal of Personality and Social Psychology, 85, 453–466. https://doi.org/10.1037/0022-3514. 85.3.453 Galinsky, A. D., Rucker, D. D., & Magee, J. C. (2015). Power: Past findings, present considerations, and future directions. In J. Simpson, M. Mikulincer, & P. R. Shaver (Eds.), APA handbook of personality and social psychology, Vol. 3: Interpersonal relations (pp. 421–460). Washington, DC: American Psychological Association. Gosling, S. D., Rentfrow, P. J., & Swann, W. B. Jr. (2003). A very brief measure of the big five personality domains. Journal of Research in Personality, 37, 504–528. https://doi.org/10.1016/ S0092-6566(03)00046-1 Guinote, A. (2007a). Power affects basic cognition: Increased attentional inhibition and flexibility. Journal of Experimental Social Psychology, 43, 685–697. https://doi.org/10.1016/j.jesp. 2006.06.008 Guinote, A. (2007b). Power and goal pursuit. Personality and Social Psychology Bulletin, 33, 1076–1087. https://doi.org/10.1177/ 0146167207301011 Guinote, A. (2008). Power and affordances: When the situation has more power over powerful than powerless individuals. Journal of Personality and Social Psychology, 95, 237–353. https://doi. org/10.1037/a0012518 Hayes, A. F. (2013). An introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: Guilford Press. Hershcovis, M. S., Neville, L., Reich, T. C., Christie, A. M., Cortina, L. M., & Shan, J. V. (2017). Witnessing wrongdoing: The effects of observer power on incivility intervention in the workplace. Organizational Behavior and Human Decision Processes, 142, 45–57. https://doi.org/10.1016/j.obhdp.2017.07.006 Hirsh, J. B., Mar, R. A., & Peterson, J. B. (2012). Psychological entropy: A framework for understanding uncertainty-related anxiety. Psychological Review, 119, 304–320. https://doi.org/ 10.1037/a0026767 Hull, C. L. (1932). The goal-gradient hypothesis and maze learning. Psychological Review, 39, 25–43. https://doi.org/10.1037/ h0072640 Inzlicht, M., Bartholow, B. D., & Hirsh, J. B. (2015). Emotional foundations of cognitive control. Trends in Cognitive Sciences, 19, 126–132. https://doi.org/10.1016/j.tics.2015.01.004

Social Psychology (2018), 49(1), 47–62

Judge, T. A., & Bono, J. E. (2001). Relationship of core selfevaluation 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 Kang, S. K., Galinsky, A. D., Kray, L. J., & Shirako, A. (2015). Power affects performance when the pressure is on: Evidence for low-power threat and high-power lift. Personality and Social Psychology Bulletin, 17, 726–735. https://doi.org/10.1177/ 0146167215577365 Keltner, D., Gruenfeld, D. H., & Anderson, A. (2003). Power, approach, and inhibition. Psychological Review, 110, 265–284. https://doi.org/10.1037/0033-295X.110.2.265 Kleiman, T., Hassin, R. R., & Trope, Y. (2014). The control-freak mind: Stereotypical biases are eliminated following conflictactivated cognitive control. Journal of Experimental Psychology: General, 143, 498–503. https://doi.org/10.1037/a0033047 Kruglanski, A., Shah, J. Y., Fishbach, A., Friedman, R., Chun, W. Y., & Sleeth-Keppler, D. (2002). A theory of goal systems. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 34, pp. 331–378). New York, NY: Academic Press. Lammers, J., Dubois, D., Rucker, D. D., & Galinsky, A. D. (2013). Power gets the job: Priming power improves interview outcomes. Journal of Experimental Social Psychology, 49, 776–779. https://doi.org/10.1016/j.jesp.2013.02.008 Lent, R. W., Brown, S. D., & Hackett, G. (1994). Toward a unifying social cognitive theory of career and academic interest, choice, and performance. Journal of Vocational Behavior, 45, 79–122. https://doi.org/10.1006/jvbe.1994.1027 Lewin, K. (1935). A dynamic theory of personality. New York, NY: Ronald Press. Magee, J. C., Galinsky, A. D., & Gruenfeld, D. H. (2007). Power, propensity to negotiate, and moving first in competitive interactions. Personality and Social Psychology Bulletin, 33, 200–212. https://doi.org/10.1177/0146167206294413 Magee, J. C., & Smith, P. K. (2013). The social distance theory of power. Personality and Social Psychology Review, 17, 158–186. https://doi.org/10.1177/1088868312472732 Myrseth, K. O. R., & Fishbach, A. (2009). Self-control: A function of knowing when and how to exercise restraint. Current Directions in Psychological Science, 18, 247–252. https://doi.org/10.1111/ j.1467-8721.2009.01645.x Riediger, M., & Freund, A. M. (2004). Interference and facilitation among personal goals: Differential associations with subjective well-being and persistent goal pursuit. Personality and Social Psychology Bulletin, 30, 1511–1523. https://doi.org/10.1177/ 0146167204271184 Riediger, M., Freund, A. M., & Baltes, P. B. (2005). Managing life through personal goals: Intergoal facilitation and intensity of goal pursuit in younger and older adulthood. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 60, 84–91. https://doi.org/10.1093/geronb/ 60.2.P84 Scheepers, D., de Wit, F., Ellemers, N., & Sassenberg, K. (2012). Social power makes the heart work more efficiently: Evidence from cardiovascular markers of challenge and threat. Journal of Experimental Social Psychology, 48, 371–374. https://doi. org/10.1016/j.jesp.2011.06.014

Social Psychology (2018), 49(1), 47–62

P. C. Schmid, Power and Goal Conflict

Schmid, P. C., Kleiman, T., & Amodio, D. M. (2015). Power effects on cognitive control: Turning conflict into action. Journal of Experimental Psychology: General, 144, 655–663. https://doi. org/10.1037/xge0000068 Schmid, P. C., & Schmid Mast, M. (2013). Power increases performance in a social evaluation situation as a result of decreased stress responses. European Journal of Social Psychology, 43, 201–211. https://doi.org/10.1002/ejsp.1937 Schmid, P. C., Schmid Mast, M., & Mast, F. W. (2015). Prioritizing – The task strategy of the powerful? The Quarterly Journal of Experimental Psychology, 68, 2097–2105. https://doi.org/ 10.1080/17470218.2015.1008525 Schwarzer, R., & Jerusalem, M. (1995). Generalized Self-Efficacy Scale. In S. W. J. Weinman & M. Johnston (Eds.), Measures in health psychology: A user’s portfolio. Causal and control beliefs (pp. 35–37). Windsor, UK: NFER-NELSON. Shah, J. Y., Friedman, R., & Kruglanski, A. W. (2002). Forgetting all else: On the antecedents and consequences of goal shielding. Journal of Personality and Social Psychology, 83, 1261–1280. https://doi.org/10.1037/0022-3514.83.6.1261 Slabu, L., & Guinote, A. (2010). Getting what you want: Power increases the accessibility of active goals. Journal of Experimental Social Psychology, 46, 344–349. https://doi.org/ 10.1016/j.jesp.2009.10.013 Smith, P. K., & Bargh, J. A. (2008). Nonconscious effects of power on basic approach and avoidance tendencies. Social Cognition, 26, 1–24. https://doi.org/10.1521/soco.2008.26.1.1 Smith, P. K., Jostmann, N. B., Galinsky, A. D., & van Dijk, W. W. (2008). Lacking power impairs executive functions. Psychological Science, 19, 441–447. https://doi.org/10.1111/j.14679280.2008.02107.x Smith, P. K., & Trope, Y. (2006). You focus on the forest when you’re in charge of the trees: Power priming and abstract information processing. Journal of Personality and Social Psychology, 90, 578–596. https://doi.org/10.1037/0022-3514. 90.4.578 Tangney, J. P., Baumeister, R. F., & Boone, A. L. (2004). High selfcontrol predicts good adjustment, less pathology, better grades, and interpersonal success. Journal of Personality, 72, 271–324. https://doi.org/10.1111/j.0022-3506.2004.00263.x Weick, M., & Guinote, A. (2010). How long will it take? Power biases time predictions. Journal of Experimental Social Psychology, 46, 595–604. https://doi.org/10.1016/j.jesp.2010. 03.005 Received July 20, 2017 Revision received September 29, 2017 Accepted September 29, 2017 Published online February 7, 2018 Petra C. Schmid Department of Management, Technology, and Economics Swiss Federal Institute of Technology (ETH Zurich) Weinbergstrasse 56/58 8092 Zurich Switzerland petraschmid@ethz.ch

Ó 2018 Hogrefe Publishing

News and Announcements Call for Papers â&#x20AC;&#x153;Ego Depletion and Self-Control: Conceptual and Empirical Advancesâ&#x20AC;? A Special Issue of Social Psychology Guest Editors: Junhua Dang1 and Martin S. Hagger2 1

Lund University, Sweden

Curtin University, Australia and University of JyvĂ¤skylĂ¤, Finland

Social Psychology is an international journal with a current Impact Factor of 2.602 (2016). The journal publishes original empirical and theoretical contributions to basic research in social psychology. The conceptualization of self-control capacity as a domain-general limited resource, and the accompanying state of low self-control resources, known as the ego depletion effect, has received considerable attention in social psychology literature. The popularity of the limited resource model lies in its elegant parsimony and inherent intuitive appeal. In addition, research on ego depletion has contributed to debates on the mechanisms underpinning self-control, catalyzed considerable interest in selfcontrol, and spurned substantive body of research testing the effect. However, recent research has raised numerous questions with respect to the replicability of the effect, and the viability and adequacy of the limited resource account. The aim of this special issue is to bring together advances in research in the field of ego depletion that tackle outstanding theoretical and empirical questions. Contributions to the special issue are expected to move research in the field of ego depletion forward by providing a platform for new empirical research in the field, and theoretical contributions that seek to provide novel explanations that will form the basis of future empirical research that will contribute to a greater understanding of the ego depletion effect. Questions addressed by contributions in the special issue are expected to include, but are not limited to, the following: What are the specific properties of tasks that will reliably lead to ego depletion? Ă&#x201C; 2018 Hogrefe Publishing

Is the sequential-task experimental paradigm fit-forpurpose as a means to test ego depletion? What factors reliably moderate the ego depletion effect? What are the most viable explanations for the mechanisms behind the effect? What are the neural correlates of ego depletion and how can this shed light on current theories of ego depletion?

Guidelines and Article Types This special issue follows standard guidelines of Social Psychology, for details please refer to the â&#x20AC;&#x153;Instructions to authorsâ&#x20AC;? available at http://www.hogrefe.com/j/sp. Accordingly, there will be three different types of articles: Original Articles that report empirical and/or theoretical contributions, Research Reports that present innovative empirical findings, and Replications that report successful or failed replications of existing research. In line with the new policy of Social Psychology, authors are encouraged to pre-register their studies and/or share their materials and data.

Timeline and Submission Manuscripts should be submitted through Social Psychologyâ&#x20AC;&#x2122;s submission portal http://www.editorialmanager. com/sopsy, indicating in a cover letter that the submission Social Psychology (2018), 49(1), 63â&#x20AC;&#x201C;64 https://doi.org/10.1027/1864-9335/a000335

Call for Papers

is for the special issue on “Ego Depletion and Self-Control.” Please direct any editorial questions to the Guest Editors: Junhua Dang (dangjunhua@gmail.com). Martin S. Hagger (martin.hagger@curtin.edu.au).

Papers acceptable for publication but cannot be published in this special issue may be considered for publication in a regular issue of Social Psychology, unless authors explicitly decline this option.

Deadline for submission of full manuscripts is August 31, 2018. All manuscripts will be subject to peer review.

Social Psychology (2018), 49(1), 63–64

Ó 2018 Hogrefe Publishing

Instructions to Authors Social Psychology is a publication dedicated to international research in social psychology as well as a forum for scientific discussion and debate. Social Psychology publishes innovative and methodologically sound research and serves as an international forum for scientific discussion and debate in the field of social psychology. Topics include all basic social psychological research themes, methodological advances in social psychology, as well as research in applied fields of social psychology. The journal focuses on original empirical contributions to social psychological research, but is open to theoretical articles, critical reviews, and replications of published research. The journal welcomes original empirical and theoretical contributions to basic research in social psychology, to social psychological methods, as well as contributions covering research in applied fields of social psychology, such as economics, marketing, politics, law, sports, the environment, the community, or health. Preference will be given to original empirical and experimental manuscripts, but theoretical contributions, critical reviews, and replications of published research are welcome as well. Social Psychology aims to increase transparency and openness of the research process and encourages authors to share their data and materials and if possible, pre-register their studies. Social Psychology publishes the following types of article: Original Articles, Research Reports, Replications. Manuscript submission: All manuscripts should in the first instance be submitted electronically at http://www.editorialmanager. com/sp. Detailed instructions to authors are provided at http:// www.hogrefe.com/j/sp 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

Ó 2018 Hogrefe Publishing

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 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 ‘‘Guidelines on sharing and use of articles in Hogrefe journals’’ on the journal’s web page at http://www.hogrefe.com/j/sp September 2016

Social Psychology (2018), 49(1)

European Journal of Health Psychology

in Now sh Engli

Editor-in-Chief Claus Vögele, University of Luxembourg, Luxembourg Editorial Assistant Nicole Knoblauch, Luxembourg

ISSN-Print 2512-8442 ISSN-Online 2512-8450 ISSN-L 2512-8442 4 issues per annum (= 1 volume)

Subscription rates (2018) Libraries / Institutions US $343.00 / € 268.00 Individuals US $120.00 / € 94.00 Postage / Handling US $16.00 / € 12.00

www.hogrefe.com

Associate Editors Verena Klusmann, Bremen, Germany Arnold Lohaus, Bielefeld, Germany Britta Renner, Konstanz, Germany Christel Salewski, Hagen, Germany Silke Schmidt, Greifswald, Germany Heike Spaderna, Trier, Germany

About the Journal The European Journal of Health Psychology was founded to provide a platform for research in health psychology, and for its application in a wide range of contexts. Health psychology is the scientific discipline within psychology that aims to promote and preserve health, to prevent diseases and contribute to their treatment by identifying disease-relevant aetiological processes, and to improve health provision.

Call for Papers The European Journal of Health Psychology invites you and/or your working group to submit papers on all aspects of the field!

The European Journal of Health Psychology strives to promote theory and practice in the analysis of psychological approaches to health and disease. Its aim is, therefore, to publish high quality empirical or experimental research as well as sound practice-oriented articles, current methodological developments, and comprehensive critical reviews of the scientific literature.

Electronic Full Text The full text of the journal – current and past issues (from 1999 onward) – is available online at econtent.hogrefe.com/loi/zgp (included in subscription price). A free sample issue is also available there.

The journal has been publishing high-quality, innovative research since 1993 (until 2016 as Zeitschrift für Gesundheitspsychologie, ISSN 0943-8149).

Manuscript Submissions All manuscripts, including Electronic Supplementary Material (ESM), should be submitted online at www.editorialmanager.com/zgp, where full instructions to authors are also available.

Abstracting Services The journal is abstracted / indexed in Social Sciences Citation Index (SSCI), Social Scisearch, Journal Citation Report/Social Sciences Edition, PsycInfo, PsycLit, PsyJOURNALS, PSYNDEX, Scopus, IBZ, IBR, and European Reference List for the Humanities (ERIH). Impact Factor (Journal Citation Reports®, Clarivate Analytics): 2016 = 0.909

Using movies to help learn about mental illness

“I have been a fan of Movies and Mental Illness from the first edition.” Steven Pritzker, PhD, psychology professor (Saybrook University) and former Hollywood script writer

Danny Wedding / Ryan M. Niemiec

Movies and Mental Illness

Using Films to Understand Psychopathology 4th edition 2014, xviii + 456 pp. US $59.00 / € 42.95 ISBN 978-0-88937-461-4 Also available as eBook Films can be a powerful aid to learning about mental illness and psychopathology – for students of psychology, psychiatry, social work, medicine, nursing, counselling, literature or media studies, and for anyone interested in mental health. Movies and Mental Illness, written by experienced clinicians and teachers who are themselves movie aficionados, has established a great reputation as a uniquely enjoyable and highly memorable text for learning about psychopathology. The new edition has been fully updated to include DSM-5 and ICD-10 diagnoses.

www.hogrefe.com

The core clinical chapters each use a fabricated case history and MiniMental State Examination along with synopses and discussions about specific movies to explain, teach, and encourage discussion about all the most important mental health disorders. Each chapter also includes: Critical Thinking Questions; “Authors’ Picks” (Top 10 Films); What To Read if You Only Have Time to Read One Book or Article; and Topics for Group Discussions.

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 individual

www.hogrefe.com

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.