Jpp 2016 15 issue 1

Volume 15 / Number 1 / 2016

Journal of

Personnel Psychology Editor-in-Chief Bernd Marcus Managing Editor Petra GellĂŠri Associate Editors Ian Gellatly Barbara Griffin Laurenz Meier Sandra Ohly Despoina Xanthopoulou Xi-An Zhang

How to meet people’s increasing need to develop and manage their own lives and careers “A must read for all career development professionals and students alike. The Handbook of the Life Design is a new and essential resource for those working to improve career services in line with today’s challenges and conditions.”

Sara Santilli, PhD, writing in Career Convergence, February 2015

Laura Nota / Jérôme Rossier (Editors)

Handbook of Life Design

From Practice to Theory and from Theory to Practice 2015, vi + 298 pp., hardcover US $54.00 / € 38.95 ISBN 978-0-88937-447-8 Also available as eBook Our lives and careers are becoming ever more unpredictable. The “lifedesign paradigm” described in detail in this ground-breaking handbook helps counselors and others meet people’s increasing need to develop and manage their own lives and careers.

Life-design interventions, suited to a wide variety of cultural settings, help

www.hogrefe.com

individuals become actors in their own lives and careers by activating, stimulating, and developing their personal resources. This handbook first addresses life-design theory, then shows how to apply life designing to different age groups and with more at-risk people, and looks at how to train life-design counselors.

Journal of

Personnel Psychology Volume 15, No. 1, 2016

Editor-in-Chief

Bernd Marcus, Work and Organizational Psychology, Faculty of the Humanities and Social Sciences, University of Hagen, Universita¨tsstr. 33, 58084 Hagen, Germany, Tel. +49 2331 987-2746, Fax +49 2331 987-2179, E-mail: bernd.marcus@fernuni-hagen.de

Managing Editor

Petra Gelle´ri, Work and Organizational Psychology, Faculty of the Humanities and Social Sciences, University of Hagen, Universita¨tsstr. 33, 58084 Hagen, Germany, Tel. +49 2331 987-2745, Fax +49 2331 987-2179, E-mail: jpp.editorial.office@gmail.com

Associate Editors

Ian Gellatly, University of Alberta, Canada Barbara Griffin, Macquarie University, Australia Laurenz Meier, University of Fribourg, Switzerland Sandra Ohly, University of Kassel, Germany Despoina Xanthopoulou, Aristotle University of Thessaloniki, Greece Xin-An Zhang, Shanghai Jiao Tong University, China

Editorial Board

Mike Ashton, Canada Arnold Bakker, The Netherlands Tanja Bipp, The Netherlands Gerhard Blickle, Germany Diana Boer, Germany John Campbell, USA Oliver Christ, Germany Neil Christiansen, USA Brian Connelly, Canada Jeremy Dawson, UK Nele de Cuyper, Belgium Filip De Fruyt, Belgium Evangelia Demerouti, The Netherlands Deanne den Hartog, The Netherlands Jo¨rg Felfe, Germany Steffen Giessner, The Netherlands Richard Goffin, Canada Peter Harms, USA Alex Haslam, UK Sarah Hezlett, USA Giles Hirst, Australia Stefan Ho¨ft, Germany Astrid C. Homan, The Netherlands Thomas Jønsson, Denmark Rudolf Kerschreiter, Germany Ulla Kinnunen, Finland Martin Kleinmann, Switzerland Cornelius Ko¨nig, Germany Franciska Krings, Switzerland Jonas Lang, Belgium Kibeom Lee, Canada Klaus Melchers, Germany Bertolt Meyer, Germany

Publisher

Hogrefe Publishing, Merkelstr. 3, 37085 Go¨ttingen, Germany, Tel. +49 551 99950-0, Fax +49 551 99950-425, E-mail publishing@hogrefe.com, Web http://www.hogrefe.com North America: Hogrefe Publishing, 38 Chauncy Street, Suite 1002, Boston, MA 02111, USA, Phone (866) 823-4726, Fax (617) 354-6875, E-mail publishing@hogrefe.com

Production

Christina Sarembe, Hogrefe Publishing, Merkelstr. 3, 37085 Go¨ttingen, Germany, Tel. +49 551 99950-0, Fax +49 551 99950-425, E-mail publishing@hogrefe.com

Subscriptions

Hogrefe Publishing, Herbert-Quandt-Str. 4, 37081 Go¨ttingen, Germany, Tel. +49 551 99950-900, Fax +49 551 99950-998

Advertising/Inserts

Marketing, Hogrefe Publishing, Merkelstr. 3, 37085 Go¨ttingen, Germany, Tel. +49 551 99950-423, Fax +49 551 99950-425, E-mail marketing@hogrefe.com

ISSN

ISSN-L 1866-5888, ISSN-Print 1866-5888, ISSN-Online 2190-5150

Copyright Information

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

Publication

Published in 4 issues per annual volume.

Subscription Prices

Calendar year subscriptions only. Rates for 2016: Institutions: US $324.00/1238.00/£190.00; Individuals: US $159.00/1114.00/£91.00. Postage and handling: US $16.00/112.00/£10.00. Single copies: US $81.00/159.50/£47.50 + postage and handling.

Payment

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

Abstracting Services

Abstracted/indexed in Social Sciences Citation Index (SSCI), Social SciSearch, Current Contents/Social and Behavioral Sciences (CC/S&BS), PsycINFO, PsyJOURNALS, PSYNDEX, Scopus, IBZ, IBR, ERIH.

Electronic Full-Text

The full-text of the Journal of Personnel Psychology is available online at www.psyjournals.com and also in PsycARTICLES.

John P. Meyer, Canada Karin S. Moser, UK Klaus Moser, Germany Peter Muck, Germany Laetitia Mulder, The Netherlands Cornelia Niessen, Germany Ioannis Nikolaou, Greece Lisa Penney, USA Deborah Powell, Canada Floor Rink, The Netherlands Ann Marie Ryan, USA Paul R. Sackett, USA Jesus F. Salgado, Spain Niclas Schaper, Germany Lothar Schmidt-Atzert, Germany Bert Schreurs, The Netherlands Sebastian Schuh, China Birgit Schyns, UK Meir Shemla, The Netherlands Christiane Spitzmu¨ller, USA Daan Stam, The Netherlands Thomas Staufenbiel, Germany Sebastian Stegmann, Germany H. Canan Su¨mer, Turkey Klaus J. Templer, Singapore Robert Tett, USA Christian Vandenberghe, Canada Beatrice van der Heijden, The Netherlands Rolf van Dick, Germany Chockalingam Viswesvaran, USA S. Arzu Wasti, Turkey Juergen Wegge, Germany Ingo Zettler, Denmark

Impact Factor (2014): 0.805

Journal of Personnel Psychology (2016), 15(1)

Ó 2016 Hogrefe Publishing

Contents Editorial

A Goodbye, a Welcome, and a Look Ahead Bernd Marcus

1

Original Articles

The Bidirectional Relationship Between Person-Job Fit and Work Engagement: A Three-Wave Study Leon T. de Beer, Sebastiaan Rothmann Jr., and Karina Mostert

4

Ă“ 2016 Hogrefe Publishing

Applicant Reactions as a Function of Test Length: Is There Reason to Fret Over Using Longer Tests?

Andrew B. Speer, Brandon S. King, and Michael Grossenbacher

15

Participative Supervisory Behavior and the Importance of Feeling Safe and Competent to Voice Mari Svendsen, Thomas S. Jønsson, and Christine Unterrainer

25

Spaces That Signal Identity Improve Workplace Productivity Katharine H. Greenaway, Hannibal A. Thai, S. Alexander Haslam, and Sean C. Murphy

35

Journal of Personnel Psychology (2016), 15(1)

Editorial A Goodbye, a Welcome, and a Look Ahead Bernd Marcus (Editor-in-Chief) Industrial and Organizational Psychology, University of Hagen, Institute of Psychology, Work and Organizational Psychology, Hagen, Germany

It is just one year ago that Rolf van Dick (2015) reviewed recent developments at the Journal of Personnel Psychology (JPP) and announced a number of changes, including the transition in the role of editor-in-chief. One of the many reasons I have to be grateful to Rolf is that his farewell editorial saves me from having to describe these developments and changes in great detail again. Nevertheless, reasons enough remain for me to write another editorial after such a short period of time.

First, I believe I can speak in the name of all readers, authors, reviewers, and board members, and of the team of editors and production staff, when I express my deep thanks to Rolf for his tremendous achievements as this journal’s editor-in-chief over the past eight years. In the 2 years I have now served JPP as an associate editor, I have not only been impressed with what he has achieved but no less with how he has made things happen in consistently fair, warmhearted, and respectful interaction with all stakeholders of the journal. I extend my big thank you to Johannes Ullrich and Diana Boer, who during Rolf’s term built congenial dyads with him as managing editors, a role whose value perhaps only editors-in-chief can appraise accurately. If JPP is now “recognized as a decent journal among authors around the globe,” as Rolf put it quite modestly (Van Dick, 2015, p. 1), this success has essentially been achieved by this team of editors. Fortunately, both Rolf and Diana have agreed to continue their service to the journal by joining the editorial board.

Secondly, I would like to introduce Petra Gelléri as Diana’s successor as managing editor. Together with the team of associate editors, who have fortunately all agreed to stay on board, and the people at Hogrefe, Petra, and I will try to continue the work of the outgoing editors. We are well aware how big the footprints are we try to step in, but we promise to do our best.

Finally, while I do not need to reiterate recent developments at JPP Rolf already wrote about in his latest editorial, a few further changes since then are worth mentioning.

Ó 2016 Hogrefe Publishing

Initiatives for Improving the Trustworthiness of Applied Psychological Research One issue discussed by Rolf (Van Dick, 2015) that I would like to mention again is JPP’s continued dedication to recent collaborative efforts to improve the trustworthiness of psychological research. Sadly enough, applied psychology in general and personnel psychology in particular, JPP’s focal areas, have not always appeared quite at the forefront of this movement. Fortunately, though, there are a number of signs indicating that the relatively ponderous development of our field in this respect is now beginning to gain momentum. The team at JPP will try to intensify this journal’s efforts in taking an active role in the movement toward improving the reliability and trustworthiness of applied psychological research. Along these lines, JPP introduces three specific but related initiatives, all of which take effect immediately.

Registered Reports As previously announced (Van Dick, 2015), JPP now joins the growing number of journals in a range of empirical sciences accepting submissions in the format of Registered Reports. This format involves a two-stage review process.

At Stage 1, authors submit a detailed proposal including Introduction and Method sections before they conduct their research, which is reviewed for conceptual and methodological rigor. Pending approval at Stage 1, the registered research is then conducted as planned and a full paper is prepared for Stage 2 review. At this final stage, the full paper is reviewed for adherence to the approved plan and for adequate presentation and discussion of results, but ultimate acceptance does not depend on outcomes such as observed support for hypotheses or effect sizes.

Journal of Personnel Psychology (2016), 15(1), 1–3 DOI: 10.1027/1866-5888/a000153

2

Hybrid Registered Reports In addition to Registered Reports, JPP also introduces the related but even newer format of Hybrid Registered Reports in a joint initiative with a number of renowned journals specifically from the fields of applied psychology and organizational behavior. This format differs from that of “traditional” Registered Reports in that Hybrid Registered Reports present research that has already been conducted before Stage 1 submission. Hence, the review process for Hybrid Registered Reports can be much shorter and more commensurate with the traditional approach of preparing elaborate papers only after having collected the data. Analogous to Registered Reports, however, authors of Hybrid Registered Reports submit only the Introduction and Method sections of their paper at Stage 1. If these parts of the paper are accepted in principle, there is a very high chance of ultimate success for the full paper at Stage 2.

Again, this warrants that acceptance for publication depends on the rigor of underlying theory and methods, but not on the results. Authors can now find new separate submission categories for Registered and Hybrid Registered Reports in JPP’s electronic submission portal at www. editorialmanager.com/jppsy, along with detailed guidelines for both new formats on the journal’s website at www. hogrefe.com/journals/jpp.

Replications JPP now also explicitly welcomes replications of previous research. Recently, a group of researchers initiated and presented a joint effort to study the replicability of psychological research independent of outcomes and, therefore, independent of potential publication bias (Open Science Collaboration, 2012, 2015). Yet, this collaborative initiative was almost exclusively restricted to experimental research in basic disciplines of psychology. Also, studies conducted in this context are explicitly restricted to replications using the exact same methods as the original published research (direct replications). At JPP, we try to extend these ideas to applied topics and to nonexperimental designs more common in applied settings, although replications of experiments within the scope of JPP are also welcome.

Moreover, we invite not only exact (direct) replications of previous research but may also accept studies attempting to replicate previous findings with more rigorous methods (advanced replications; e.g., multiple sources, designs allowing for causal inferences, more rigorous sampling, etc.), as well as combinations of both types of replications.

Apart from these specific differences, however, JPP’s present initiative shares with that of the Open Science

Journal of Personnel Psychology (2016), 15(1), 1–3

Editorial

Collaboration the core element that chances of publication success should not depend on outcomes. Probably the best way to guarantee this is that the team of reviewers decides on acceptance without knowing (or even being able to guess) the outcome. Therefore, JPP accepts replication studies only in the formats of either Registered Reports or Hybrid Registered Reports. Specifics regarding replications are included in the guidelines for these two submission formats.

Awards for Outstanding Achievements as Authors and Reviewers Another new development I am happy to announce is that, starting with the present volume, JPP will establish two annual awards for outstanding achievements in different roles. First, outstanding contributions by authors will be recognized by an annual Best Paper Award. Yet authorship, at least authorship of successful submissions, is inherently rewarding even if there were no additional recognition or incentive. The truly unsung heroes in the process of publishing research are the anonymous external reviewers, who often contribute substantially to the quality of the final product without being recognized at all. In an attempt to change this at least in a few of the most deserved cases, JPP will also distinguish two outstandingly helpful reviewers with its annual Best Reviewer Award. A committee composed of members of the editorial team will nominate and select candidates for both awards. Winners will be recognized in one printed issue and on the journal’s website. Although both awards are immaterial in principle, Hogrefe has kindly offered award winners a small tangible gift from its program.

New Members of the Editorial Board Finally, I am proud and happy to announce that JPP’s editorial board continues to grow in a number of respects, including not only quantity, but also internationality, scope of expertise, and reputation. With the present issue, I warmly welcome the following scholars to our editorial board (in alphabetical order): Mike Ashton (Canada) Diana Boer (Germany)

Ó 2016 Hogrefe Publishing

Editorial

Filip De Fruyt (Belgium) Deanne den Hartog (The Netherlands) Peter Harms (USA) Franciska Krings (Switzerland) Jonas Lang (Belgium) Kibeom Lee (Canada) Bertolt Meyer (Germany) Lisa Penney (USA) Ann Marie Ryan (USA) Paul R. Sackett (USA) Klaus J. Templer (Singapore) Robert Tett (USA) Rolf van Dick (Germany) Ingo Zettler (Denmark) It seems to me that exciting times lie ahead of JPP and its team of contributors serving the journal in different roles.

I look forward to sharing this experience with our readers and with everybody interested in the future of the journal and the topics it deals with. Thank you for your ongoing support.

Ó 2016 Hogrefe Publishing

3

References Open Science Collaboration. (2012). An open, large-scale, collaborative effort to estimate the reproducibility of psychological science. Perspectives on Psychological Science, 7, 657–660. doi: 10.1177/1745691612462588 Open Science Collaboration. (2015). Estimating the reproducibility of psychological science. Science, 349, aac4716. doi: 10.1126/ science.aac4716 Van Dick, R. (2015). Registered reports, advance articles online, and the way ahead [Editorial]. Journal of Personnel Psychology, 14, 1–3. doi: 10.1027/1866-5888/a000140 Published online April 12, 2016 Bernd Marcus Work and Organizational Psychology Faculty of the Humanities and Social Sciences University of Hagen Universitätsstr. 33 58084 Hagen Germany Tel. +49 2331 987-2746 Fax +49 2331 987-2179 E-mail bernd.marcus@fernuni-hagen.de

Journal of Personnel Psychology (2016), 15(1), 1–3

How abusive professionals manipulate their clients and what we can do about it “This is a very informative book on a deeply important subject.

It is essential for the professional functioning of all individuals in a fiduciary relationship with patients, clients and students.”

Michael F. Myers, MD, former Clinical Professor of Psychiatry, University of British Columbia, Vancouver

Werner Tschan

Professional Sexual Misconduct in Institutions Causes and Consequences, Prevention and Intervention 2014, xvi + 216 pp.

US $49.00 / € 34.95 ISBN 978-0-88937-444-7 Also available as eBook Professional sexual misconduct (PSM) is a problem that is notoriously difficult to address and that can be a minefield for all concerned – for victims, for the institutions where it takes place – and also because outstanding and supposedly responsible members of society may be accused of abuse.

Here, Werner Tschan, one of the world’s leading experts in dealing with PSM, outlines an up-to-date approach to PSM and professional disruptive behaviors. He describes practical ways to prevent PSM, as well as effective treatments for victims and to rehabilitate offenders or those accused.

www.hogrefe.com

Using examples from real-life cases from around the world, he also discusses how PSM is a societal problem and what we can do to stop it. Recent headline cases involving a variety of organizations – medical, media, church, schools, sport, industry – show that institutions can be ideal environments for PSM, and so great emphasis is placed in this volume on preventive measures that we can and must take at an institutional level. With clear, jargon-free writing this book is essential reading for all professionals interested in preventing and dealing with PSM, as well as of interest to victims and their families.

Fredrike Bannink

Handbook of Positive Supervision for Supervisors, Facilitators, and Peer Groups 2015, xii + 204 pp.

US $49.00 / € 34.95 ISBN 978-0-88937-465-2 Also available as eBook

Positive supervision focuses on what actually works instead of on problems and on supervisees’ strengths rather than on their deficits. The task of supervisors using this approach is – unlike the more traditional problemsolving – to create solutions with their supervisees and to teach them to apply the same approach when working with their own clients.

Essential reading for all supervisors, this book introduces a new form of supervision, based on positive psychology and solution-focused brief therapy, that is shorter, more positive and hopeful, and more cost-effective than traditional methods.

Fredrike Bannink

Handbook of Solution-Focused Conflict Management 2010, xii + 180 pp., hardcover US $49.00 / € 34.95 ISBN 978-0-88937-384-6 Also available as eBook

Benjamin Franklin once said: “Every problem is an opportunity in disguise.”

In the new and highly successful approach of solution-focused conflict management described here, the focus is on discovering these opportunities to find the “win-win” scenario. The key lies in asking eliciting questions about goals, exceptions, and competencies and in motivating clients to change.

Clients’ perspectives are considered primary, and they are empowered to formulate their own hopes for the future and to devise ways to make them happen. Focusing on the preferred future facilitates change in the desired direction.

www.hogrefe.com

This book is essential reading for all those who need to manage conflicts. It provides a detailed description of the highly successful solution-focused model, its theoretical background, and practical applications in conflict management practice: divorce, workplace, family, neighbors, personal injury cases, victim-offender conflicts.

Kenneth Cloke, one of the most inspiring mediators in the world, wrote the Foreword and Epilogue.

Original Article

The Bidirectional Relationship Between Person-Job Fit and Work Engagement A Three-Wave Study Leon T. de Beer, Sebastiaan Rothmann Jr., and Karina Mostert WorkWell Research Unit, North-West University, Potchefstroom, South Africa

Abstract: Although theorized and generally accepted, research on the relationship (specifically the direction of the causal relationship) between person-job fit and work engagement is limited and not yet clear. Theoretical arguments can be presented for both directions, but empirical evidence is scarce. The study reported here explored the bidirectional relationship between person-job fit and work engagement in a longitudinal design. A three-wave cross-lagged panel design was used in a sample of 382 participants from the financial, healthcare, and manufacturing industries in South Africa. The results of Bayesian informative hypothesis testing showed the most support for the hypothesis stating that work engagement is a stronger predictor of person-job fit over time. Conversely, the hypothesis that person-job fit is the predictor of work engagement over time received limited support.

Keywords: person-job fit, work engagement, structural equation modeling, Bayesian informative hypothesis testing

Due to global recovering economic situations and the resultant scarcity of work, the current employment market still presents numerous obstacles. Consequently, applicants have to grab employment opportunities with both hands, which may entail having to settle for positions that are not related to their level of education, specialization, and/or expectations. As a result, employees may feel discontent with their current occupational positions (Jha & Bhattacharyya, 2012). These feelings of discontent could create feelings of misfit with their position and could in turn affect work engagement levels.

The relationship between person-job fit and work engagement is important. Work engagement is a pivotal issue for employers since it is strongly related to retention and increased organizational commitment (Bakker & Demerouti, 2007; De Beer, Rothmann Jr., & Pienaar, 2012; Hakanen, Bakker, & Schaufeli, 2006). Moreover, increased employee satisfaction, health, and overall performance are all important outcomes of work engagement (Bakker & Leiter, 2010). However, in the literature, there is an almost exclusive focus on the relationship of personjob fit with outcomes such as job satisfaction, performance, and commitment (Edwards, 1991; Erdogan & Bauer, 2005; June & Mahmood, 2011; Kristof-Brown, Zimmerman, & Johnson, 2005). While these are all important organizational outcomes to investigate, most of the referred studies Journal of Personnel Psychology (2016), 15(1), 4–14 DOI: 10.1027/1866-5888/a000143

do not include or even mention work engagement (although some researchers suggest that person-job fit could be a predictor of work engagement, e.g., Leiter & Maslach, 1999).

Although it is important to investigate the relationship between person-job fit and work engagement, a major limitation in the literature is an almost exclusive focus on cross-sectional data (e.g., Laschinger & Finegan, 2005) instead of utilizing longitudinal designs. A key strength of longitudinal designs is that they allow for investigating developments and changes over time (Taris, 2000), this will assist researchers and organizations to understand the dynamics between person-job fit and work engagement (cf. Parker & Griffin, 2011) over time. One exception is the recent study of Lu, Wang, Lu, Du, and Bakker (2014), who investigated the relationship of work engagement to person-job fit through job crafting. They used a two-wave longitudinal design and found favorable support for their hypothesis, that is, that engaged employees create better person-job fit through job-crafting behaviors.

The study of Lu et al. (2014) makes an important contribution toward understanding the longitudinal relationship between work engagement and person-job fit. However, it is important to note that Lu et al. only tested the unidirectional relationship between work engagement and personjob fit. Therefore, the question remains which of the two Ă“ 2016 Hogrefe Publishing

L. T. de Beer et al., Person-Job Fit and Work Engagement

variables has the largest impact on the other over time, as arguments can be presented for both directions. Investigating the longitudinal bidirectional relationship between person-job fit and work engagement is an important endeavor for researchers and practitioners, since research findings will guide practitioners with regard to the focus of their intervention strategies to achieve the desired employee and organizational outcomes. If it is true that person-job fit predicts work engagement over time, then the focus and main concern of interventions should be on recruitment and selection practices in order to optimize person-job fit (to enhance work engagement over time).

On the other hand, if work engagement precedes personjob fit, interventions should be focused on factors influencing work engagement (so that perceptions of person-job fit may change over time).

Proposing the predictive strength of one variable over the other (e.g., person-job fit will be a stronger predictor of work engagement over time than the other way around) is not necessarily simple. Past research shows that the longitudinal relationships between work-based affective states (such as work engagement) and perceptions of fit could be very complex. For instance, Gabriel, Diefendorff, Chandler, Moran, and Greguras (2013) examined the longitudinal relationship between perceived fit and affective-based variables (including job satisfaction, positive affect, and negative affect). They report that: (i) fit can precede affect (this was primarily the case where unidirectional fit led to job satisfaction); (ii) affect can precede fit (this occurred when unidirectional positive and negative affect had subsequent effects on person-job fit); and (iii) that reciprocal relationships exist (however, this occurred more in the case of person-organization fit than person-job fit).

With regard to the relationship between person-job fit and work engagement, the question therefore arises: Which of the two variables has the largest impact on the other over time, since convincing arguments can be presented for both directions.

The objective of this study was therefore to test the bidirectional relationship between person-job fit and work engagement using three waves of data (as recommended

by Ployhart & Ward, 2011). More specifically, the aim was to test whether person-job fit is a stronger predictor of work engagement over time, or whether work engagement is a stronger predictor of person-job fit over time.

Work Engagement Work engagement is often considered to be a positive affective state which classically comprises three components: vigor, dedication, and absorption (Schaufeli, Salanova, González-Romá, & Bakker, 2002). However, absorption

Ó 2016 Hogrefe Publishing

5

has been argued to be a consequence of engagement’s two core components: vigor and dedication (Demerouti, Mostert, & Bakker, 2010; González-Romá, Schaufeli, Bakker, & Lloret, 2006). As a result, several researchers only include vigor and dedication in their studies (e.g., Bakker, Schaufeli, Leiter, & Taris, 2008; Brand-Labuschagne, Mostert, Rothmann Jr., & Rothmann, 2013; De Beer, Pienaar, & Rothmann Jr., 2014; Langelaan, Bakker, Van Doornen, & Schaufeli, 2006). Vigor (an energy component) can be defined as an energy aspect of work engagement, characterized by willingness to invest effort, high levels of activity, and mental resilience shown in the process of work (Schaufeli & Bakker, 2004; Schaufeli et al., 2002). Dedication (an identification component) is characterized by a sense of significance, enthusiasm, inspiration, pride, and challenge that can be seen as an attitudinal aspect of work engagement (Schaufeli & Bakker, 2004; Schaufeli et al., 2002). From previous research, it is clear that work engagement leads to various desirable and sought-after outcomes such as decreased turnover of employees (therefore retention of scarce skills) and organizational commitment, which in turn results in financial gain for employers (Bakker & Demerouti, 2007; Bakker, Demerouti, & Sanz-Vergel, 2014; De Beer et al., 2012; Hakanen et al., 2006).

Person-Job Fit Person-job fit can generally be categorized into two broader subsections, that is, supplementary fit and complementary fit. Supplementary person-job fit is found where a person “supplements, embellishes, or possesses characteristics which are similar to other individuals” in the work environment (Kristof-Brown et al., 2005; Muchinsky & Monahan, 1987, p. 269). Complementary person-job fit, the focus of the current research, can be regarded as the degree to which an employee estimates how he/she “fits” or “matches” a current position – that is, referring to the knowledge, skills, and abilities (KSAs) that an employee should possess in order to perform his/her work tasks assigned by employers successfully so that specified outcomes can be reached (cf. Cable & DeRue, 2002; Edwards, 1991; Kristof, 1996; Muchinsky & Monahan, 1987; O’Reilly, Chatman, & Caldwell, 1991). Person-job fit is associated with several key organizational outcomes, including job performance (Edwards, 1991; June & Mahmood, 2011; KristofBrown et al., 2005) and job satisfaction (Erdogan & Bauer, 2005; Kristof-Brown et al., 2005).

The recruitment and selection process assists organizations to match or fit individuals to certain positions, at least according to the criteria of the organization. However, it has been found that the subjective perception of personjob fit (i.e., individuals’ subjective perception of their

Journal of Personnel Psychology (2016), 15(1), 4–14

6

congruence with the organization) is a more accurate indicator of employees’ attitude than objective indicators (e.g., qualification) (Caplan, 1987). Therefore, if the subjective perception of employees could be influenced positively, then the perception of fit with the job could be increased.

As a result, failure in the recruitment and selection process might not be a dead end, but interventions could be implemented to affect person-job fit perceptions.

Person-Job Fit as Predictor of Work Engagement Over Time The first case that could be made is that person-job fit will be a stronger predictor of work engagement over time (compared to the other way around – i.e., that work engagement will be a stronger predictor of person-job fit over time). This specific relationship can be argued within the framework of two theories, namely the self-determination theory (Deci, Ryan, & Williams, 1996) and the job demands-resources (JD-R) model (Bakker & Demerouti, 2007; Bakker et al., 2014).

According to the self-determination theory (Deci, Vallerand, Pelletier, & Ryan, 1991), every individual has certain needs connected to autonomy, relatedness, and competence, which increases intrinsic motivation (Deci & Ryan, 1985), enhances well-being, and increases commitment. With regard to autonomy, different employees might enjoy different types of support (e.g., more participation in decision-making concerning their work and position). To develop relatedness, organizations might implement team building sessions or create opportunities where employees could interact with each other for social support and satisfy the basic need to belong. To increase employee competence, organizations could provide and encourage skills development and training opportunities (cf. Greguras & Diefendorff, 2009). When an organization provides resources to satisfy these needs, increased person-job fit will be evident due to the congruence created for the employee in the position and environment (Deci et al., 1996; Yu, 2009). In other words, employees whose needs and requirements are satisfied or matched by an employer (by person-job resource-fitting) in accordance with autonomy, competence, and relatedness, should show higher levels of work engagement (Hakanen et al., 2006).

Research suggests that the relationship between personjob fit and work engagement exists because employees who fit their job may increase their work engagement through positive meaningful work by matching the employee’s self-concept with job tasks and behaviors (Hamid & Yahya, 2011; Scroggins, 2008). Thus, when employees have good fit with their jobs, demonstrate the specific skills and abilities relevant to such positions, and are more congruent to

Journal of Personnel Psychology (2016), 15(1), 4–14

L. T. de Beer et al., Person-Job Fit and Work Engagement

the job, they should be able to apply their available resources to perform well and be more engaged in their work. According to the JD-R model, a balance between job demands and job resources leads to work engagement and an activated state of well-being (Bakker & Demerouti, 2007). Employers could fit job resources to employees’ needs, which could in turn improve the person-job fit perception of the employee. Autonomy, supervisor and colleague support, growth, and development opportunities are all job resources that could be discussed and enhanced during annual performance reviews or job redesign/enrichment interventions. This could lead to increased motivation and proactive behavior in the creation of a satisfying position and work environment (Bakker, 2010). As a result, employees may be more able and likely to feel engaged and perform well on the job (Hackman, 1980).

Work Engagement as Predictor of Person-Job Fit Over Time Although it seems likely that person-job fit will be a strong predictor of work engagement over time, an equally strong argument can be made that work engagement will be a strong predictor of person-job fit over time by drawing on the broaden-and-build theory (Fredrickson, 2001). The basic premise of this theory is that positive affect broadens attention, cognition, and action – instigating a “building” of personal resources such as physical resources (e.g., skills), intellectual resources (e.g., knowledge), and psychological resources (e.g., resilience) (Fredrickson & Branigan, 2005). Therefore, employees who are engaged (a positive affective state) could alter their environment to be more in line with their preferences, which could increase perceptions of fit with the job. This is in line with the findings of Yu (2009), who argues that the direction from the majority of cross-sectional studies might have been misinterpreted and that work-based affect would cause person-environment fit rather than the other way around. As an example, Yu (2009) illustrates how affect (in this case, job dissatisfaction) could influence fit through the hedonistic perspective. Employees are motivated to experience positive affect, but respond to negative affect in a corrective manner, that is, either cognitive (e.g., by focusing on issues such as compensation) or behavioral (e.g., by quitting their job). Thus, employees who experience job dissatisfaction would have a lowered person-job fit perception due to the negative affective emotionality of the dissatisfaction (Cable & DeRue, 2002; Carless, 2005).

Since job satisfaction has been argued to be a more passive state of well-being compared to work engagement, which is considered an active state of well-being (Bakker, 2011), it is therefore possible that work engagement

Ó 2016 Hogrefe Publishing

L. T. de Beer et al., Person-Job Fit and Work Engagement

(a positive state) could also have an effect on person-job fit over time as it also results in positive affect (Bakker & Demerouti, 2007; Schaufeli & Taris, 2005; Sonnentag, Mojza, Binnewies, & Scholl, 2008). Indeed, in support of this argument, the recent two-wave study by Lu et al.

(2014) found that work engagement, through job crafting (a proactive behavior), can increase perceptions of person-job fit.

The Present Study Given the above-mentioned arguments, there is reason to expect relationships in both directions over time (i.e., person-job fit leading to work engagement, and conversely, work engagement leading to person-job fit). Therefore, the following two competing informative research hypotheses are presented for investigation of the data (see the method section for the statistical expression of hypotheses): HA: Person-job fit is a stronger positive predictor of work engagement than engagement is of person-job fit across all time intervals in a three-wave study.

HB: Work engagement is a stronger positive predictor of person-job fit than person-job fit is of work engagement across all time intervals in a three-wave study.

Bayesian informative hypothesis testing was used to test both of these hypotheses simultaneously. By using Bayes factors to compete the two hypotheses in the data, the researcher allows for both hypotheses to be potentially true, but is able to present substantive support for which one of the two informative hypotheses received the most support as predictor of the other over time.

Method Null-Hypothesis Testing in Cross-Lagged Panel Designs In the classic cross-lagged panel longitudinal design, variables in question were auto-regressed over time to act as control variables. Given these control paths, the variables were then also regressed on each other over time (crosslagged paths, as suggested by Taris, 2000). The null hypothesis was then tested for each of the cross-lagged paths. Rejection of these null hypotheses was then used as evidence that a relationship existed over time. This design is quite popular for testing time-lagged effects or even causality; although whether true causality can be determined has also been questioned (cf. Rogosa, 1980). Ă“ 2016 Hogrefe Publishing

7

The classic cross-lagged panel method is simple in the case of two-wave longitudinal designs because only one lagged path exists between each variable and the other over time. However, our three-wave longitudinal design which is complicated matters, because three cross-lagged relationships existed between each variable and the other over time: time 1 to time 2, time 2 to time 3, and time 1 to time 3. Use of the above-mentioned classic method in the current study would have resulted in the null hypothesis being tested for three cross-lagged paths, and testing one hypothesis with multiple null-hypothesis tests would have increased the risk for error (Banerjee, Chitnis, Jadhav, Bhawalkar, & Chaudhury, 2009) and would not address the expected hypotheses directly (cf. Van de Schoot, Hoijtink, & Jan-Willem, 2011).

Nested model comparison is an alternative method of hypothesis testing in a cross-lagged context (Kline, 2011).

All cross-lagged paths from person-job fit to engagement could be constrained to zero in one model, and vice versa in another model. These models could be competed against each other by means of a chi-square difference test. This results in the difference being tested with a single hypothesis test. The problem with this approach, in our case, was that we did not expect any one of the crosslagged paths to be zero. Thus, if we were to follow the nested model approach, we would essentially have been specifying hypotheses that were not a true reflection of our theoretical argumentation and the research hypotheses specified.

We were interested in determining which one of the two variables has the strongest cross-lagged paths over all measurement intervals, given the data. Another option, which we applied, was to make use of inequality-constrained hypotheses (Van de Schoot, Hoijtink, & Deković, 2010).

The hypothesis would then be that person-job fit is a better predictor of engagement, given all cross-lagged paths, or the complement (opposite), that engagement is a better predictor of person-job fit, given all cross-lagged paths.

Inequality-constrained hypotheses can be tested with the likelihood ratio test (Van de Schoot, Hoijtink, Hallquist, & Boelen, 2010) or with Bayesian informative hypothesis testing, which will provide support for each of the hypotheses in the form of a Bayes factor.

Bayesian Informative Hypothesis Testing and Bayes Factors Informative hypothesis testing provides researchers with the ability to test hypotheses with inequality expressions (Hoijtink, 2012). These inequality expressions may be expressed in terms of multiple parameters of a structural equation model, including means, beta coefficients, and Journal of Personnel Psychology (2016), 15(1), 4–14

8

even the residuals (Van de Schoot et al., 2012). Informative hypotheses are tested by calculating Bayes factors, which can be compared against a range of other hypotheses, including the null hypothesis, the unconstrained hypothesis, the complement of the hypothesis (the exact opposite), or other hypotheses concerning alternative expectations (Hoijtink, 2012). The Bayes factor takes into account model fit as well as the complexity of the hypothesis, that is, researchers are rewarded for being more specific about their expectations.

Testing inequality constraints presents exciting possibilities in the case of cross-lagged models. To determine which factor is likely to be the causal factor, a hypothesis could be formulated that assumes one variable as a stronger predictor of another over time. This could be extended to threeor more waved cross-lagged designs by hypothesizing that one variable is a stronger predictor of another over all given time points. This methodology was therefore applied in the current study.

Procedure and Participants To help achieve the research aims, a longitudinal survey design was used. The first survey was conducted in 2011, after 1 year (12 months; SD = 1), and again 1 year later (12 months; SD = 1). The three waves were merged into one longitudinal dataset containing the common participants from each of the time points.

Employees of all ages and backgrounds were sampled.

The majority (253; 66%) of participants were female, while 131 (34%) of the participants were male. Employees between 18 years and 65 years of age were included, with the average age being 40.31 (SD = 10.29). Data was collected from participants in organizations from three sectors: finance (108; 28.1%), healthcare (139; 36.2%), and manufacturing (137; 35.7%).

Anonymity and confidentiality of the participants were assured. No identifying information was shared with the organization or any other parties during or after the research. All ethical guidelines for research with human subjects were adhered to, and informed consent was obtained from all participants. All the participants received a link to the Internet-based survey via e-mail. Unique assessment numbers linked to e-mail addresses were used to link the surveys conducted over time with each of the participants. All data and data connections were encrypted to further ensure privacy and security.

Initially, 520 employees were sampled from the different sectors. During the first year of implementation, 475 participants completed the survey (response rate: 84.9%), in the second year of implementation, 423 participants completed the survey, and in the third year, 384 employees completed the survey (Attrition: 19.8%).

Journal of Personnel Psychology (2016), 15(1), 4–14

L. T. de Beer et al., Person-Job Fit and Work Engagement

To ascertain whether the attrition could be described as missing completely at random (MCAR), the means of person-job fit and engagement were compared between the participants who had left the study and the participants who continued with the study (Enders, 2010). It was evident that there were negligible mean differences between the groups (d < 0.20; Cohen, 1988).

Measures The South African Employee Health and Wellness Survey (SAEHWS; Rothmann & Rothmann, 2007) was used to measure the constructs. The SAEHWS is a self-administered survey applied in organizational climate research, and can be completed online and/or by means of paperand-pencil implementation. The following constructs were used for this study: Work engagement (α = .92; .91; .92) was measured by items from its two core components vigor and dedication as a one-factor structure (De Bruin & Henn, 2013; De Bruin, Hill, Henn, & Muller, 2013) totaling eight items (four items to measure vigor and four to measure dedication), for example, “I am full of energy in my work” (vigor) and “I find my work full of meaning and purpose” (dedication).

Person-job fit (α = .86; .84; .82) as a one-factor structure was measured by four items, for example, “The requirements of my job match my specific talents and skills.” This scale is based on the measure of Saks and Ashforth (1997).

All items for person-job fit were measured on a 6-point Likert-type scale ranging from (1) Strongly disagree to (6) Strongly agree. Engagement was measured on a 7-point scale ranging from (0) Never to Always (6).

Analysis Latent variable modeling with Bayesian estimation was conducted with Mplus 7.11 (Muthén & Muthén, 2013).

Firstly, a measurement model was specified in which latent variables were created for each of the time points (in one model). The items were specified for categorical data estimation (modeling multiple thresholds rather than single intercepts) because many of the items were measured on Likert-type scales (Rhemtulla, Brosseau-Liard, & Savalei, 2012). Thresholds and factor loadings were fixed between the three groups to ensure the comparability between time points (Kline, 2011).

Consequently the cross-lagged model was specified (see Figure 1), with auto-regression control paths and crosslagged paths between time periods (Taris, 2000). Ó 2016 Hogrefe Publishing

L. T. de Beer et al., Person-Job Fit and Work Engagement

9

Figure 1. The cross-lagged structural model.

The cross-lagged paths were included for year 1 to year 2, year 2 to year 3, and for year 1 and year 3. The parameters of the Markov chain Monte Carlo (MCMC) estimation were exported and saved using the BPARAMETERS option in Mplus (Muthén & Muthén, 2013).

Figure 1 shows the cross-lagged structural model as specified for the BPARAMETER output.

Subsequently, Bayes factors (BF) were computed from the parameter output file to investigate support for each of the informative hypotheses. The procedure described in Van de Schoot et al. (2012) was used. Below, the Bayes factor is shown as a function of model fit and complexity, expressed as:

BFHi vs: Hu ¼

fi ci

504 ci

ð2Þ

where ni is the number of iterations needed for stable results for hypothesis i and ci is the hypothesis complexity. It is important to note that these iterations exclude the burn-in phase, and as a default, Mplus uses 50% of the iterations as burn-in. Therefore to be safe, at least twice the original number of iterations derived from the Ó 2016 Hogrefe Publishing

BFHi1 vs: Hi2 ¼

BFHi1 vs: Hu BFHi2 vs: Hu

ð3Þ

where BFHi1 vs. Hu is the Bayes factor for Hi1 against the unconstrained hypothesis and BFHi2 vs. Hu is the Bayes factor for Hi2 against the unconstrained hypothesis.

Statistically, the research hypotheses were specified as: HA:

ð1Þ

where fi represents the model fit and ci, the complexity of the hypothesis Hi. This Bayes factor compares the given hypothesis Hi against the unconstrained hypothesis, which states that “something is going on, but not the null hypothesis” (Van de Schoot, Verhoeven, & Hoijtink, 2013, p. 86). Complexity refers to the number of equivalent hypotheses to the specified hypothesis Hi, given all possible hypotheses, while model fit refers to the number of MCMC iterations that matched the specific hypothesis Hi, given the total number of iterations. Guidelines regarding the minimum number of iterations needed for stable results are provided by Van de Schoot et al.

(2012). The authors of the current research article derived a rule of thumb for the approximate calculation of the number of iterations needed for informative hypothesis testing:

ni

formula above, and an increase to 100,000 iterations were specified in Mplus.

Finally, to compare one hypothesis (Hi1) against another hypothesis (Hi2), the following formula was used (Hoijtink, 2012):

Be2p3 < Bp2e3 ^ Be1p3 < Bp1e3 ^ Be1p2 < Bp1e2

ð4Þ

Be2p3 > Bp2e3 ^ Be1p3 > Bp1e3 ^ Be1p2 > Bp1e2

ð5Þ

HB:

where Beipj refers to the standardized beta of engagement at time point i on person-job fit at time point j and Bpiej refers to the standardized beta of person-job fit at time point i on work engagement at time point j.

The level of practical significance of correlation coefficients was set at r 0.30 for a medium effect, and r 0.50 for a large effect (Cohen, 1988). Bayesian analysis also presents credibility intervals for estimates at the 95% credibility interval, which are analogously similar to bootstrapped confidence intervals (Zyphur & Oswald, 2015).

The lower and upper credibility interval should also not cross zero. Connected to this is a p-value, and it is important to note that the p-value presented in the output file was one-tailed – therefore “statistical significance” was set at p < .025 and not p < .05 as the case would have been with two-tailed p-value.

Results The results are discussed in the following order: Parameter trace plots, measurement invariance, correlation matrix for Journal of Personnel Psychology (2016), 15(1), 4–14

10

L. T. de Beer et al., Person-Job Fit and Work Engagement

Table 1. Correlation matrix for the latent variables r Variable

1

1. Engagement (T3)

2

3

4

5

6

1.00

2. Person-job fit (T3)

0.64**

1.00

3. Engagement (T2)

0.76**

0.53**

1.00

4. Person-job fit (T2)

0.27

0.64**

0.40*

1.00

5. Engagement (T1)

0.52**

0.57**

0.71**

0.54**

1.00

6. Person-job fit (T1)

0.35*

0.49*

0.51**

0.58**

0.66**

1.00

Lower

Upper

Notes. r = correlation coefficient; T1 = time 1; T2 = time 2; T3 = time 3; * = medium effect; ** = large effect.

Table 2. Parameter estimates with 95% credibility intervals Structural regressions

Standard estimate

SD

p

Engagement (T1) ? Engagement (T2)

0.66

0.06

.001

0.55

0.77

Engagement (T2) ? Engagement (T3)

0.79

0.04

.001

0.71

0.86

Person-job fit (T1) ? Person-job fit (T2)

0.40

0.07

.001

0.27

0.54

Person-job fit (T2) ? Person-job fit (T3)

0.47

0.06

.001

0.35

0.57

Engagement (T1) ? Person-job fit (T2)

0.27

0.07

.001

0.13

0.40

Engagement (T1) ? Person-job fit (T3)

0.15

0.06

.012

0.02

0.27

Engagement (T2) ? Person-job fit (T3)

0.24

0.07

.001

0.10

0.37

Person-job fit (T1) ? Engagement (T2)

0.07

0.06

.134

0.06

0.19

Person-job fit (T1) ? Engagement (T3)

0.05

0.05

.201

0.15

0.06

Person-job fit (T2) ? Engagement (T3)

0.02

0.06

.377

0.13

0.09

Notes. SD = posterior standard deviation; p = one-tailed; lower and upper = 95% credibility interval.

Table 3. Results of the informative hypothesis testing Hypothesis Complexity Iterations required Model fit Bayes factor HA

.13

4032

.00

0.01

HB

.13

4032

.98

7.88

the latent variables, parameter estimates for the specified research model, and the informative hypothesis testing.

Two MCMC chains were used in the Bayesian estimation (default setting). Investigation of the parameter trace plots revealed that mixing took place with no inordinate variations; therefore, chain convergence was achieved. Results for measurement invariance for participants over all three of the time waves revealed strong measurement invariance (configural, metric, and scalar) indicating that the measurement was invariant for each participant over time.

Table 1 shows the correlation matrix for the latent variables.

The correlation values revealed large practically significant correlations between the same corresponding variables over time, for example, Engagement T3 and Engagement T2 (r = 0.76); Person-job fit T3 and Person-job fit T2 (r = 0.64). All of the variables were practically significantly correlated with each other except for the association between Engagement T3 and Person-job fit T2 (r = 0.27) Journal of Personnel Psychology (2016), 15(1), 4–14

which was approaching medium practical significance.

The association between Engagement T1 and Person-job fit was also large (r = 0.66).

Table 2 shows the parameter estimates for the research model, that is, standardized beta coefficients and associated values.

As can be seen in Table 2, the corresponding latent variables over time had relatively stable predictive relationships with their corresponding counterparts. The strongest control path was between Engagement T2 and Engagement T3 (β = 0.79; 95% CI [0.71; 0.86]). All of the predictive relationships from Engagement to Person-job fit did not include zero: Engagement T1 positively predicted Person-job fit T2 (β = 0.27; 95% CI [0.13; 0.40]), and Person-job fit T3 (β = 0.15; 95% CI [0.02; 0.27]). Moreover, Engagement T2 also positively predicted Person-job fit T3 (β = 0.24; 95% CI [0.10; 0.37]). All of the predictive relationships from Person-job fit to Engagement over all of the time points crossed zero at the 95% credibility level.

Table 3 shows the results of the hypotheses compared against the unconstrained hypothesis.

The results of the analyses revealed that HB (BF = 7.88) was the hypothesis with the most support. The aforementioned hypothesis was the hypothesis indicating that work engagement is a stronger predictor of person-job fit over Ó 2016 Hogrefe Publishing

L. T. de Beer et al., Person-Job Fit and Work Engagement

time than the complement hypotheses. Indeed, the inverse hypothesis (HA) that person-job fit has a stronger relationship with work engagement, over time, had basically no support in any circumstances given the data (BF = 0.01).

Additionally, comparing the hypotheses directly with each other, it was evident that HB compared to HA resulted in a BF of 788; indicating that HB received 788 times more support (was more likely a hypothesis) in the data, given the complexity.

Discussion The main aim of the study reported here was to investigate the bidirectional relationships of person-job fit and work engagement over time. This was accomplished by implementing a three-wave cross-lagged panel design, testing inequality-constrained informative hypotheses with a Bayesian methodology, that is, calculating Bayes factors to investigate support for each of the two expected hypotheses to determine which hypothesis received the most support.

The informative hypothesis that work engagement is a stronger predictor of person-job fit over time received the most support. The explanation for why the longitudinal relationship from work engagement to person-job fit exists was consistent with broaden-and build theory in that positive affect leads to other positive outcomes. This finding was also in line with results from the recent two-wave study

by Lu et al. (2014), who investigated and found a relationship from work engagement to person-job fit through job crafting. Previous research (Tims, Bakker, Derks, & Van Rhenen, 2013) also found that employees implementing job-crafting strategies, specifically on job resources, are more engaged, which results in a positive reciprocal relationship (gains spiral). Therefore, employees who possess adequate job resources (e.g., autonomy, participation in decision-making, role clarity, and social support) and who are engaged in their work may be better positioned to meet their own needs satisfaction – which is important for person-job fit, potentially through job crafting. Job crafting can be described as the process through which employees shape their own jobs, for example, actively changing the content and design of the position by choosing tasks and negotiating different job content (Bakker, 2011; Wrzesniewski & Dutton, 2001). Indeed, research has found that employees who are engaged are also proactive in shaping their work environments, thereby not only making full use of current job resources, but also creating their own resources to remain engaged (Bakker, Demerouti, & Xanthopoulou, 2011), which results in an apparent positive gains spiral (Llorens, Schaufeli, Bakker, & Salanova, 2007). Additionally, Bakker (2011) Ó 2016 Hogrefe Publishing

11

suggests that through job crafting, employees may be able to increase their person-job fit. Therefore it stands to reason that person-job fit may be a partial outcome of work engagement through the practice of resource-crafting behavior – which seems to be an obvious intervening process (i.e., mediator).

Even though a relationship is purported to exist between person-job fit and engagement (cf. Leiter & Maslach, 1999), the informative hypothesis that investigated person-job fit as a stronger predictor of work engagement over time (HA) received virtually no support in the data. According to Hamid and Yahya (2011), limited research has been done on the topic of person-job fit and work engagement, and research regarding person-job fit has shown conflicting findings. For example, Hirschi (2012) presupposed that work engagement as an outcome of calling to work (passion for the specific job or career; the job as one’s purpose in life) was conditional of the level of person-job fit, but found this was not a prerequisite. This indicates that person-job fit is not necessarily a prerequisite for an employee to experience more work engagement, but that an employee enters (or has) a job with a certain level of inherent person-job fit.

What then happens to the level of person-job fit in the future is highly dependent on what happens with that employee’s work engagement as per the current study’s results. Furthermore, research has found that activated well-being states can lead to feelings of misfit (Warr & Inceoglu, 2012), but also that person-job fit is strongly related to job satisfaction (Erdogan & Bauer, 2005; Kristof-Brown et al., 2005), and that job satisfaction should be considered a more passive state when compared to the work engagement state, which is seen as an active state (Bakker, 2011). Perhaps, person-job fit is not necessarily such an activator and mobilizer of resources to foster work engagement when compared to the mobilization of resources for work engagement – evidently through active job crafting (Lu et al., 2014) – to affect noticeable change in person-job fit, but it may remain important for job satisfaction. Another possibility for the little support for HA is that the current study did not consider person-job misfit – studies have found that misfit and uncertainty could lead to creative and proactive employee action to reduce uncertainty (e.g., Warr & Inceoglu, 2012). There is a possibility that management and organizations could misconstrue this finding and think to gain by instilling for instance job insecurity or in actively attempting to increase feelings of misfit in employees to reach organizational outcomes. But research has not yet shown what the longer-term effect could be on employee well-being and health may be if misfit as a potential strategy (going along with the misfit) is (un)successful?

Conversely, past research has clearly shown that work engagement is a positive state that can be influenced and Journal of Personnel Psychology (2016), 15(1), 4–14

12

improved, mainly by providing employees with adequate job resources (Bakker & Demerouti, 2007; De Beer et al., 2012; Hakanen et al., 2006). Therefore, if work engagement can affect person-job fit over time, and it is possible to enhance employees’ work engagement through job resources, it is possible to enhance their fit perceptions over time. This is in line with research discussed in the Literature section by Gabriel et al. (2013), who found that the relationship between affective states and fit perceptions is complex and that the relationship is not simplistic. When performance management, job enrichment, or job redesign strategies are implemented (Cunningham & Eberle, 1990; Hackman, 1980), the employee may request and bargain for job resources in conjunction with the employer in order to increase his or her perception of person-job fit.

The gain in job resources then leads to increased motivation. From the current results, it was clear that this effect was achieved through work engagement, and not the other way around.

In the context of the current research, it was shown that, longitudinally, there was substantial support for the hypothesis that work engagement is a strong positive predictor of person-job fit over time, and not vice versa. However, it is important to consider a multidimensional view of this finding in the context of work engagement’s broader nomological net as suggested by Parker and Griffin (2011). This finding does not invalidate the importance of the personjob fit, but rather explains the importance of work engagement for maintaining and improving motivation and person-job fit perceptions of employees.

Limitations and Recommendations for Future Research The strength of this study was its longitudinal design.

However, it would have been interesting to investigate the relationship of work engagement with person-job fit, over time, with an additional fourth and fifth wave of data.

Moreover, while the current research project was constrained to one measurement per year, other researchers could endeavor to replicate these findings with shorter waves of measurement, for example, monthly or bimonthly.

Even though it was the specific aim of this study, the study only focused on person-job fit and did not take into account the broader concept of person-organization fit, which collectively makes up person-environment fit. Additionally, this study only considered two variables’ influence on each other over time, and did not include, for example, job crafting, which could have addressed a more complex reality. Future studies should therefore endeavor to include more applicable variables in their investigations to provide a broader perspective on the findings of this study.

Journal of Personnel Psychology (2016), 15(1), 4–14

L. T. de Beer et al., Person-Job Fit and Work Engagement

Furthermore, this study consisted of a sample with participants from three specific sectors. However, some caution is still advised in extrapolating the results to other sectors indiscriminately. Future researchers therefore also have some opportunity to investigate this finding in the context of other sectors, or perhaps specifically in certain types of professions (e.g., knowledge workers vs. others).

Recommendations to Management This study showed that there is more support for work engagement as a positive predictor of person-job fit than person-job fit is of work engagement. It is therefore possible to influence an employee’s perception of person-job fit positively through the state of work engagement, that is, evidence shows that improved person-job fit is an outcome of work engagement. Thus, employees enter organizations with a certain level of person-job fit (subjected to the scrutiny of recruitment and selection practices that attempt to optimize the likelihood of person-job fit). But what happens to this level of fit perception after appointment (the maintenance or increased perception of fit) is highly dependent on the work engagement levels of the employee. Therefore, to achieve better person-job fit for employees, employers, consultants, and organizations should focus on enhancing the work engagement of employees. To achieve this, organizations should foster job resources that match employees’ needs. Organizations could consider running yearly climate surveys in order to identify job demands and job resources that are associated with work engagement in the specific organization, and attempt to intervene in those variables to ensure work engagement. Furthermore, as recent research has shown that job crafting increases resources and therefore work engagement (Bakker, 2011; Lu et al., 2014; Wrzesniewski & Dutton, 2001), organizations could consider intervening at the individual level in an attempt to “teach” job-crafting behavior to employees. This could lead to increased work engagement and therefore also increased person-job fit.

References Bakker, A. B. (2010). Engagement and job crafting: Engaged employees create their own great place to work. In S. Albrecht (Ed.), Handbook of engagement: Perspectives, issues, research and practice (pp. 229–244). Northampton, MA: Edwin Elgar.

Bakker, A. B. (2011). An evidence-based model of work engagement. Current Directions in Psychological Science, 20, 265–269.

Bakker, A. B., & Demerouti, E. (2007). The job demands-resources model: State of the art. Journal of Managerial Psychology, 22, 309–328.

Bakker, A. B., Demerouti, E., & Sanz-Vergel, A. I. (2014). Burnout and Work Engagement: The JD–R Approach. Annual Review of Organizational Psychology and Organizational Behavior, 1, 389–411. doi: 10.1146/annurev-orgpsych-031413-091235

Ó 2016 Hogrefe Publishing

L. T. de Beer et al., Person-Job Fit and Work Engagement

Bakker, A. B., Demerouti, E., & Xanthopoulou, D. (2011). How do engaged employees stay engaged? Ciencia & Trabajo, 13, 135–142.

Bakker, A. B., & Leiter, M. P. (2010). Work engagement: A handbook of essential theory and research. New York, NY: Psychology Press.

Bakker, A. B., Schaufeli, W. B., Leiter, M. P., & Taris, T. W. (2008).

Work engagement: An emerging concept in occupational health psychology. Work & Stress, 22, 187–200.

Banerjee, A., Chitnis, U. B., Jadhav, S. L., Bhawalkar, J. S., & Chaudhury, S. (2009). Hypothesis testing, type I and type II errors. Industrial Psychiatry Journal, 18, 127–131.

Brand-Labuschagne, L., Mostert, K., Rothmann Jr., S. R., & Rothmann, J. C. (2013). Burnout and work engagement of South African blue-collar workers: The development of a new scale. Southern African Business Review, 16, 58–93.

Cable, D. M., & DeRue, D. S. (2002). The convergent and discriminant validity of subjective fit perceptions. Journal of Applied Psychology, 87, 875–883.

Caplan, R. D. (1987). Person-environment fit theory and organizations: Commensurate dimensions, time perspectives, and mechanisms. Journal of Vocational Behavior, 31, 248–267.

Carless, S. A. (2005). Person-job fit versus person-organization fit as predictors of organizational attraction and job acceptance intentions: A longitudinal study. Journal of Occupational and Organizational Psychology, 78, 411–429.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (Rev. ed.). Orlando, FL: Academic Press.

Cunningham, J. B., & Eberle, T. (1990). A guide to job enrichment and redesign. Personnel, 67, 56–61.

De Beer, L., Rothmann Jr., S., & Pienaar, J. (2012). A confirmatory investigation of a job demands-resources model using a categorical estimator. Psychological Reports, 111, 528–544.

De Beer, L. T., Pienaar, J., & Rothmann Jr., S. (2014). Job burnout, work engagement and self-reported treatment for health conditions in South Africa. Stress and Health. Advance online publication. doi: 10.1002/smi.2576 De Bruin, D. P., Hill, C., Henn, C. M., & Muller, K. P. (2013).

Dimensionality of the UWES-17: An item response modelling analysis. SA Journal of Industrial Psychology, 39, 8.

De Bruin, G. P., & Henn, C. M. (2013). Dimensionality of the 9-item Utrecht work engagement scale (UWES–9). Psychological Reports, 112, 788–799.

Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and selfdetermination in human behavior. New York, NY: Plenum.

Deci, E. L., Ryan, R. M., & Williams, G. C. (1996). Need satisfaction and the self-regulation of learning. Learning and Individual Differences, 8, 165–183.

Deci, E. L., Vallerand, R. J., Pelletier, L. G., & Ryan, R. M. (1991).

Motivation and education: The self-determination perspective.

Educational Psychologist, 26, 325–346.

Demerouti, E., Mostert, K., & Bakker, A. B. (2010). Burnout and work engagement: A thorough investigation of the independency of both constructs. Journal of Occupational Health Psychology, 15, 209–222.

Edwards, J. R. (1991). Person-job fit: A conceptual integration, literature review, and methodological critique. In C. L. Cooper & I. T. Robertson (Eds.), International review of industrial and organizational psychology (Vol. 6, pp. 283–357). New York, NY: Wiley.

Enders, C. K. (2010). Applied missing data analysis. New York, NY: Guilford Press.

Erdogan, B., & Bauer, T. N. (2005). Enhancing career benefits of employee proactive personality: The role of fit with jobs and organisations. Personnel Psychology, 58, 859–891.

Ó 2016 Hogrefe Publishing

13

Fredrickson, B. L. (2001). The role of positive emotions in positive psychology: The broaden-and-build theory of positive emotions. The American Psychologist, 56, 218.

Fredrickson, B. L., & Branigan, C. (2005). Positive emotions broaden the scope of attention and thought-action repertoires.

Cognition & Emotion, 19, 313–332.

Gabriel, A. S., Diefendorff, J. M., Chandler, M. M., Moran, C. M., & Greguras, G. J. (2013). The dynamic relationships of work affect and job satisfaction with perceptions of fit. Personnel Psychology, 67, 389–420. doi: 10.1111/peps.12042 González-Romá, V., Schaufeli, W. B., Bakker, A. B., & Lloret, S. (2006). Burnout and work engagement: Independent factors or opposite poles? Journal of Vocational Behavior, 68, 165–174.

Greguras, G. J., & Diefendorff, J. M. (2009). Different fits satisfy different needs: Linking person-environment fit to employee commitment and performance using self-determination theory.

Journal of Applied Psychology, 94, 465–477.

Hackman, J. R. (1980). Work redesign and motivation. Professional Psychology, 11, 445–455.

Hakanen, J. J., Bakker, A. B., & Schaufeli, W. B. (2006). Burnout and work engagement among teachers. Journal of School Psychology, 43, 495–513.

Hamid, S. N. A., & Yahya, K. K. (2011). Relationship between person-job fit and person-organization fit on employees’ work engagement: A study among engineers in semiconductor companies in Malaysia. In Annual Conference on Innovations in Business & Management London, UK (pp. 1–30).

Hirschi, A. (2012). Callings and work engagement: Moderated mediation model of work meaningfulness, occupational identity, and occupational self-efficacy. Journal of Counseling Psychology, 59, 479–485. doi: 10.1037/a0028949 Hoijtink, H. (2012). Informative hypotheses: Theory and practice for behavioral and social scientists. London, UK: Chapman and Hall/CRC. Jha, S., & Bhattacharyya, S. S. (2012). Study of perceived recruitment practices and their relationships to job satisfaction.

Synergy, 10, 63–76.

June, S., & Mahmood, R. (2011). The relationship between personjob fit and job performance: A study among the employees of the service sector SMEs in Malaysia. International Journal of Business, Humanities and Technology, 1, 95–105.

Kline, R. B. (2011). Principles and practice of structural equation modeling. New York, NY: Guilford Press.

Kristof, A. L. (1996). Person-organization fit: An integrative review of its conceptualizations, measurement, and implications.

Personnel Psychology, 49, 1–49.

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, and person-supervisor fit.

Personnel Psychology, 58, 281–342. doi: 10.1111/j.17446570.2005.00672.x Langelaan, S., Bakker, A. B., Van Doornen, L. J., & Schaufeli, W. B. (2006). Burnout and work engagement: Do individual differences make a difference? Personality and Individual Differences, 40, 521–532.

Laschinger, H. K. S., & Finegan, J. (2005). Empowering nurses for work engagement and health in hospital settings. Journal of Nursing Administration, 35, 439–449.

Leiter, M. P., & Maslach, C. (1999). Six areas of worklife: a model of the organizational context of burnout. Journal of Health and Human Services Administration, 21, 472–489.

Llorens, S., Schaufeli, W., Bakker, A., & Salanova, M. (2007). Does a positive gain spiral of resources, efficacy beliefs and engagement exist? Computers in Human Behavior, 23, 825–841.

Journal of Personnel Psychology (2016), 15(1), 4–14

14

Lu, C. Q., Wang, H. J., Lu, J. J., Du, D. Y., & Bakker, A. B. (2014).

Does work engagement increase person-job fit? The role of job crafting and job insecurity. Journal of Vocational Behavior, 84, 142–152.

Muchinsky, P. M., & Monahan, C. J. (1987). What is personenvironment congruence? Supplementary versus complementary models of fit. Journal of Vocational Behavior, 31, 268–277.

Muthén, L. K., & Muthén, B. O. (2013). Mplus User’s Guide (7th ed.). Los Angeles, CA: Muthén & Muthén.

O’Reilly, C. A., Chatman, J., & Caldwell, D. F. (1991). People and organizational culture: A profile comparison approach to assessing person-organization fit. Academy of Management Journal, 34, 487–516.

Parker, S. K., & Griffin, M. A. (2011). Understanding active psychological states: Embedding engagement in a wider nomological net and closer attention to performance. European Journal of Work and Organizational Psychology, 20, 60–67.

Ployhart, R. E., & Ward, A. K. (2011). The “quick start guide” for conducting and publishing longitudinal research. Journal of Business and Psychology, 26, 413–422.

Rhemtulla, M., Brosseau-Liard, P. É., & Savalei, V. (2012). When can categorical variables be treated as continuous? A comparison of robust continuous and categorical SEM estimation methods under suboptimal conditions. Psychological Methods, 17, 354–373.

Rogosa, D. (1980). A critique of cross-lagged correlation. Psychological Bulletin, 88, 245–258.

Rothmann, S., & Rothmann, J. C. (2007). The South African Employee Health and Wellness Survey (SAEHWS). User manual (4th ed.). Potchefstroom, South Africa: Afriforte.

Saks, A. M., & Ashforth, B. E. (1997). A longitudinal investigation of the relationships between job information sources, applicant perceptions of fit, and work outcomes. Personnel Psychology, 50, 395–426.

Schaufeli, W. B., & Bakker, A. B. (2004). Job demands, job resources and their relationship with burnout and engagement: A multi-sample study. Journal of Organizational Behaviour, 25, 293–315.

Schaufeli, W. B., Salanova, M., González-Romá, V., & Bakker, A. B. (2002). The measurement of engagement and burnout: A two sample confirmatory factor analytic approach. Journal of Happiness Studies, 3, 71–92.

Schaufeli, W. B., & Taris, T. W. (2005). The conceptualization and measurement of burnout: Common ground and worlds apart.

Work and Stress, 19, 256–262.

Scroggins, W. A. (2008). The relationship between employee fit perceptions, job performance, and retention: Implications of perceived fit. Employee Responsibilities and Rights Journal, 20, 57–71.

Sonnentag, S., Mojza, E. J., Binnewies, C., & Scholl, A. (2008).

Being engaged at work and detached at home: A week-level

Journal of Personnel Psychology (2016), 15(1), 4–14

L. T. de Beer et al., Person-Job Fit and Work Engagement

study on work engagement, psychological detachment, and affect. Work & Stress, 22, 257–276.

Taris, T. (2000). Longitudinal data analysis. Thousand Oaks, CA: Sage.

Tims, M., Bakker, A. B., Derks, D., & Van Rhenen, W. (2013). Job crafting at the team and individual level: Implications for work engagement and performance. Group & Organization Management, 38, 427–454. doi: 10.1177/1059601113492421 Van de Schoot, R., Hoijtink, H., & Deković, M. (2010). Testing inequality constrained hypotheses in SEM models. Structural Equation Modeling, 17, 443–463.

Van de Schoot, R., Hoijtink, H., Hallquist, M. N., & Boelen, P. A. (2012). Bayesian evaluation of inequality-constrained hypotheses in SEM models using Mplus. Structural Equation Modeling, 19, 1–17.

Van de Schoot, R., Hoijtink, H., & Jan-Willem, R. (2011). Moving beyond traditional null hypothesis testing: Evaluating expectations directly. Frontiers in Psychology, 2, 1–5. doi: 10.3389/ fpsyg.2011.00024 Van de Schoot, R., Verhoeven, M., & Hoijtink, H. (2013). Bayesian evaluation of informative hypotheses in SEM using Mplus: A black bear story. European Journal of Developmental Psychology, 10, 81–98.

Warr, P., & Inceoglu, I. (2012). Job engagement, job satisfaction, and contrasting associations with person-job fit. Journal of Occupational Health Psychology, 17, 129–138.

Wrzesniewski, A., & Dutton, J. E. (2001). Crafting a job: Revisioning employees as active crafters of their work. Academy of Management Review, 26, 179–201.

Yu, K. Y. T. (2009). Affective influences in person-environment fit theory: Exploring the role of affect as both cause and outcome of P-E fit. Journal of Applied Psychology, 94, 1210–1226.

Zyphur, M. J., & Oswald, F. L. (2015). Bayesian estimation and inference: A user’s guide. Journal of Management, 41, 390–420. doi: 10.1177/0149206313501200

Received February 2, 2014 Revision received November 3, 2014 Accepted December 1, 2014 Published online April 12, 2016 Leon T. de Beer Private Bag X6001 WorkWell Research Unit North-West University Hoffman Street Potchefstroom, 2520 South Africa Tel. +27 18 299-1347 Fax +27 87 231-5396 E-mail DeBeer.Leon@nwu.ac.za

Ó 2016 Hogrefe Publishing

Original Article

Applicant Reactions as a Function of Test Length Is There Reason to Fret Over Using Longer Tests?

Andrew B. Speer,1 Brandon S. King,2 and Michael Grossenbacher3 1

American Family Insurance, Madison, WI, USA

2

Advocate Health Care, Downers Grove, IL, USA Wonderlic Inc., Vernon Hills, IL, USA

3

Abstract: This study investigated how the length of preemployment assessments affects applicant reactions to the testing process and organization. Using a between-subjects design, participants took one of four assessments (short personality, long personality, short cognitive, long cognitive) where they were incentivized to perform well, followed by a survey assessing perceptions of procedural justice, organizational attractiveness, and likelihood of accepting a job offer. Longer tests did not worsen applicant reactions for either personality or cognitive tests, and in fact individuals taking a longer cognitive assessment reported more favorable applicant reactions. Implications are discussed.

Keywords: applicant reactions, test length, test time, procedural justice, testing

An important concern for human resource (HR) professionals is the ability to attract the most qualified candidates.

Remaining attractive during recruitment and selection increases the chances that candidates will accept job offers, pass on positive word-of-mouth, and form desirable work attitudes such as job satisfaction when actually hired for the job (e.g., Gilliland, 1993). As such, applicant reactions to selection procedures have become an important topic to consider for companies when designing hiring processes (e.g., Hausknecht, Day, & Thomas, 2004; König, Klehe, Berchtold, & Kleinmann, 2010; Truxillo & Bauer, 2011).

Research has revealed that specific testing characteristics can actually dissuade candidates from accepting jobs (e.g., low job relevance, low perceived ability to express one’s abilities, not providing testing information, employer response time of performance feedback; Becker, Connolly, & Slaughter, 2010; Gilliland & Steiner, 2001; Rosse, Miller, & Stecher, 1994; Schleicher, Venkataramani, Morgeson, & Campion, 2006; Truxillo, Bauer, Campion, & Paronto, 2002). Although many attributes of the testing process have been examined in relation to candidate reactions, an area that has received very little interest is the actual testing time, or test length, of the selection process.

Length for commercial assessments can vary greatly, and although a good deal of attention has been given to the 1

topic of applicant reactions, close to none has examined the effect of length or testing time. For entry-level jobs, companies are often concerned using tests of 45 min or longer, seeking to reduce the applicant burden wherever possible.1 As longer tests are capable of assessing a greater number of competencies and increases in the number of items improve measurement reliability, practitioners are put in a difficult spot trying to balance creating the best possible assessment while at the same time appeasing the hiring organization over assessment length. With knowledge regarding the effect of test length, HR practitioners should be better able to ensure positive applicant reactions while at the same time utilize a selection procedure of sufficient length and breadth. As such, the current endeavor tested whether increased test length enhances or worsens applicant reactions after individuals actually take an assessment.

Applicant Reactions and the Role of Procedural Justice Applicant reactions are the general attitudes and evaluative responses applicants experience to hiring procedures.

Gilliland (1993) provided the first substantial model regarding applicant reactions and this theory was largely based

Although the majority of testing has moved to non-proctored online settings, in cases where testing is proctored, concerns may extend beyond applicant reactions to include the costs and time associated with conducting such testing.

Ó 2016 Hogrefe Publishing

Journal of Personnel Psychology (2016), 15(1), 15–24 DOI: 10.1027/1866-5888/a000145

16

upon organizational justice theory (see Greenberg, 1990).

According to Gilliland’s model, applicants hold specific expectations regarding the hiring process (e.g., job relatedness of test, opportunity to perform, consistency of scoring, explanation rules, interpersonal treatment), and meeting these expectations leads to favorable perceptions of procedural justice, in turn affecting reactions during hiring (i.e., job acceptance decisions, recommendations, test motivation, intent of legal actions), reactions after hiring (i.e., performance, organizational citizenship behaviors, job satisfaction, organizational climate), and self-perceptions (i.e., self-esteem, self-efficacy, future job search intentions).

Research has generally supported Gilliland’s model (cf.

Truxillo & Bauer, 2011).

Of particular concern to the present study are how characteristics of selection tests affect applicant reactions, and of the propositions laid out by Gilliland (1993), job relatedness and the opportunity to perform have been shown to be most consistently related (Hausknecht et al., 2004). “Job relatedness refers to the extent to which a test either appears to measure content relevant to the job situation or appears to be valid” (Gilliland, 1993, pp. 703). The first aspect of this definition deals essentially with the content validity of the test, or in terms of applicant perceptions, the face validity and whether a test appears to measure components related to the job. The second aspect deals with whether or not applicants perceive a test to be predictive of job performance. In a comprehensive meta-analysis of the applicant reactions literature, Hausknecht and colleagues (2004) found strong predictive relationships for face validity and perceived predictive validity with a variety of justice outcomes and test perceptions. In turn, perceptions of organizational justice were strongly related to more distal applicant outcomes such as organizational attractiveness and intent to accept a job offer. Opportunity to perform, or the perceived opportunity to adequately express one’s knowledge, skills, and abilities, also demonstrated strong relationships with procedural justice, organizational attractiveness, and intent to accept a job offer.

Does Test Length Matter?

A longer test means more required time to complete the assessment. For a given type of assessment, increases in the number of test questions (i.e., length) will ultimately increase average testing time. While these two test features are highly related, the former is fixed across applicants,2 whereas testing time will vary by applicant. Due to this variation, our focus is on test length, which is likely to affect a number of testing decisions. These include the number of 2

A. B. Speer et al., Applicant Reactions and Test Length

required test administrations per applicant (i.e., one sitting or two sittings), the reliability of the test’s assessed constructs (in general, the more items the higher a measure’s reliability), and the number of constructs capable of being reliably assessed. It may also affect applicant reactions.

We propose that the ways in which test length affect applicant reactions are twofold and operate in two unique and opposing ways.

Negative Effects on Applicant Reactions It is possible that longer tests negatively impact applicant reactions, as boredom and fatigue resulting from increased effort should occur more intensely for longer assessments, conceivably resulting in unfavorable reactions to the test and organization. Longer test lengths may result in unfavorable applicant reactions because of reductions in motivation (Ryan & Ployhart, 2000), and longer tests have been associated with greater subjective fatigue and boredom (Ackerman & Kanfer, 2009). If applicants experience these negative states while taking an assessment they may in turn perceive the test less favorably and tie those negative evaluations to the company itself.

Positive Effects on Applicant Reactions Operating in a simultaneous and yet opposite manner than boredom and fatigue, longer assessments might actually enhance applicant reactions too. First, due to increases in total test content the assessment should presumably become more reliable and comprehensive, and therefore applicants may perceive the test to be more job-related and more predictive of job performance, with each having been shown to be positively related to favorable applicant reactions (Anderson, Salgado, & Hülsheger, 2010; Hausknecht et al., 2004). Second, it is likely that longer tests will be perceived as allowing better opportunities to express knowledge, skills, and abilities, which are also strongly related to perceptions of procedural justice and other applicant reaction outcomes (Hausknecht et al., 2004; Schleicher et al., 2006). Indeed, within Bauer and colleagues’ (Bauer, McCarthy, Anderson, Truxillo, & Salgado, 2012) best practices suggestions, they state that applicants will perceive they have had a better opportunity to perform if a test is of adequate length. If a test is too short applicants are likely to feel they have not fully expressed their qualifications for the job. For these reasons, we believe job relevance and opportunity of expression exert themselves as primary influences on any positive effect found for increasing test length. However, a third proposition has also been suggested, and that is, longer assessments might favorably affect applicant reactions through cognitive dissonance (Rafaeli, 1999).

Assuming, of course, a non-adaptive and non-timed test.

Journal of Personnel Psychology (2016), 15(1), 15–24

Ó 2016 Hogrefe Publishing

A. B. Speer et al., Applicant Reactions and Test Length

In the only known study to explicitly examine the effect of test time or test length on applicant reactions, Rafaeli (1999) adopted principles of cognitive dissonance (Festinger & Carlsmith, 1959) and predicted that the more an applicant invests in an assessment process the more likely they would be to view the testing process and organization favorably. According to Rafaeli, actually taking the time to respond to assessment questions is an investment on the part of the applicant and through cognitive dissonance, results in commitment to the company. Research has long established that once committed toward an endeavor people will subconsciously experience a drive to maintain behavioral or attitudinal consistency with that commitment (Aronson, 1992; Cialdini, 2009; Cialdini, Trost, & Newsom, 1995). In the selection scenario, voluntarily spending more time taking an assessment is equivalent to making a greater commitment to the application process, and in turn should drive tension to justify that behavior by more favorably perceiving the company in question.

Rafaeli (1999) found that students at an Israeli university were more inclined to indicate that they would continue with a selection process and accept a job offer when the assessment was longer. However, the chief limitation of this study was the participants did not actually take the assessments but instead were informed of how long the testing process would last if they had (either 2 or 10 hr). Thus, this previous study could not determine what effect test length would have when individuals actually sit down and respond to a real assessment.

The current study examined test length’s effect on reactions, building off Rafaeli’s (1999) research by examining the effect of test length on applicant reactions using individuals who actually respond to tests of various lengths.

Although longer tests are likely to be associated with feelings of fatigue and boredom, the several mechanisms outlined here may enhance applicant reactions as test length increases. Given the lack of past research on this topic and the conflicting influences on reactions, we offer the following research question.

Research Question 1: What effect does test length have on applicant reactions?

Test Length and Different Types of Assessments It is possible that the effect of test length operates uniquely for specific types of selection assessments. For the current study, we investigated reactions to general mental ability (GMA, i.e., cognitive) tests and personality tests. Because personality inventories are often viewed as unrelated to jobs Ó 2016 Hogrefe Publishing

17

and intrusive, cognitive assessments are usually favored over personality tests (Anderson et al., 2010; Hausknecht et al., 2004). However, many of the studies included within applicant reactions meta-analyses have been based on participants reading brief test descriptions rather than when respondents actually experience a test. To complicate matters further, with there being limited research regarding test length, there is simply little theory to guide predictions as to how length might interact with test type to affect applicant reactions. Given this, we offer the following research question.

Research Question 2: Does test length differentially affect applicant reactions for specific types of assessments?

Methods Participants and Procedure A between-subjects design was utilized with 206 participants from two large universities in the United States. Participants underwent a hiring simulation, and following this they then provided their reactions to that hiring procedure and of a hypothetical organization being applied to. Participants received extra credit in psychology courses in exchange for their time. Additionally, as a way to incentivize applicants to perform well and to attempt to simulate the motivations that come along with a hiring situation, participants were led to believe their performance was contingent upon earning $20, such that the top 20% of performers would be rewarded. Upon debriefing, participants were informed that the incentive involved intentional deception and while the assessments would not be scored, they would be entered into drawings for a total of $250.

Applicant Scenario Participants were tested in groups in a large room containing two dozen computers. Participants were randomly assigned to one of four conditions (short personality, short cognitive, long personality, long cognitive), and to allow for approximately equivalent completion times within a given session, all participants in the same timeslot were assigned to the same experimental condition. After arriving participants were informed that the study’s purpose was to evaluate company hiring procedures and that they would be applying for a hypothetical job. In order to simulate an environment as similar to an applicant context as possible, a company website was created. The website required participants to navigate through several pages of information prior to taking the assigned preemployment test. Included Journal of Personnel Psychology (2016), 15(1), 15–24

18

on the website were pages providing a detailed history of the organization, the organization’s mission statement, and a detailed description of the job they were applying to, which was an office manager position. The job description created for this position was largely based on information obtained from O*NET and designed such that personality traits and cognitive ability were equally important to the job. In other words, people were “applying” for a job that required an equal mix of interpersonal qualities, motivational traits, problem-solving skills, and general intelligence.

After reading and being verbally led through the job description by the study proctor, participants navigated through the website as if they were a prospective job applicant having just graduated from college. Participants then took a link from the company website to a preemployment assessment, where they responded to one of four assessments, depending on the experimental condition. Upon finishing the assessment, participants navigated to an online survey that contained measures for the study’s dependent variables and a check to determine if they were motivated when taking the assessment. After completing this, participants were instructed to wait until all other participants completed the study and then were debriefed as a group.

Test Conditions Four tests were created for the present study (short personality, long personality, short cognitive, long cognitive) and served as the four separate experimental conditions. It was chosen to not use commercial inventories for the assessments to enhance generalizability and to ensure use of a wide range of item content for each type of test.

The short and long cognitive assessments measured three types of narrow abilities: deductive reasoning, verbal reasoning, and quantitative reasoning. Questions were drawn from a pool of items created in an effort to form an IRT-based GMA test. These items were initially formed for a commercial test vendor and underwent extensive item revisions and quality assurance reviews prior to finalization.

From this pool, a set of items were chosen for the present study. One to four items were presented on each webpage, depending on the type of item.

To form the personality assessments, job-relevant personality scales were identified from the International Personality Item Pool (IPIP; Goldberg et al., 2006) and combined into one test. Thus, similar to a personality test fully customized for a particular job, a set of numerous job-related traits were measured within one single assessment. Items were drawn from the following scales taken from the IPIP: the 16 Personality Factor Questionnaire scales of Intellect, Dutifulness, Warmth, Assertiveness, Orderliness, and Reserve; the Broad Big Five Factor Markers of Extraversion, Agreeableness, Conscientiousness, Journal of Personnel Psychology (2016), 15(1), 15–24

A. B. Speer et al., Applicant Reactions and Test Length

Emotional Stability, and Openness to Experience; Emotional Intelligence; the Values in Action scales of Citizenship, Equity, Honesty, Judgment, Leadership, Learning, Wisdom, Prudence, Self-Control, Enthusiasm, and Social Intelligence; and the Oregon Vocational Interests Scales of Leadership, Organization, and Altruism.

Several pilot sessions were conducted to calibrate test lengths. Following these pilots the final tests were created, with 14 questions included in the short cognitive test (43% quantitative, 43% verbal, 14% deductive) and 73 in the long cognitive test (40% quantitative, 40% verbal, 20% deductive). The short personality scale consisted of 111 items, and the long personality scale consisted of 387 items. While this number of items appears high, commercial personality inventories of similar lengths are commonly used (for instance, the California Personality Inventory is 434 items and takes 45–60 min to complete). For the actual study sessions, the short conditions had similar testing times (cognitive: M = 12 min, SD = 3 min; personality: M = 10 min, SD = 2 min), and long conditions were close in time but somewhat more discrepant (cognitive: M = 44 min, SD = 11 min; personality: M = 36 min, SD = 7 min). Due to this difference in completion times across long types of tests, care must be made interpreting findings across test types.

In addition, we sought to examine how much individual variation in testing time was explained by the simple, dichotomous test length variable. The correlation between minutes taken to complete the assessment and the dichotomous length variable (specifying long vs. short conditions) was .89. Thus, test length and testing time were very strongly related, as would be expected. Given the research focus on helping practitioners choose tests and the high degree of multicollinearity between test time and test length, it was chosen to use test condition for all analyses.

Measures All dependent variables were scored using a 7-point ordered response scale with anchors ranging from “strongly disagree” to “strongly agree.” Different item orders were randomly assigned to participants to avoid ordering effects.

There were no significant differences in scores for the different orderings. All items used are listed in Appendix.

Procedural Justice Procedural justice was measured using a five-item scale adapted from Colquitt (2001), with items that targeted attitudes regarding the job relevance of the test and the opportunity for candidates to express their traits. Example items are “The company’s hiring assessment gave me the opportunity to accurately represent myself” and “The company’s hiring assessment is a fair method for selecting employees.” Ó 2016 Hogrefe Publishing

A. B. Speer et al., Applicant Reactions and Test Length

19

Table 1. Correlations among test characteristics and applicant reactions M

SD

1

2

3

4

5

6

7

1. Test length

–

–

–

2. Test type

–

–

.15

–

3. Motivation

3.92

0.87

.08

.14

–

4. Test performance

0.00

1.00

.00

.00

.38

–

5. Perceived effort expended

3.55

1.74

.14

.16

.11

.10

.91

6. Procedural justice

4.58

1.36

.20

.11

.23

.02

.40

.85

7. Organizational attractiveness

4.95

1.36

.20

.11

.26

.03

.39

.60

.88

8. Intentions to accept job offer

5.07

1.57

.18

.15

.17

.00

.23

.45

.79

8

.89

Notes. N = 206. Test length is coded 0 = short, 1 = long. Test type is coded 0 = cognitive, 1 = personality. Values on the diagonal are alpha reliabilities, provided where appropriate. Because there were four test conditions, one single reliability estimate could not be provided.

Internal consistency for the scale was .85 and represents a holistic measure of procedural justice.

Organizational Attractiveness Organizational attractiveness was measured using an adapted four-item version of Lievens, Van Hoye, and Schreurs’ (2005) scale. Example items are “A job at The Loading Zone is appealing to me” and “This company would be a good place to work.” Internal consistency for the scale was .88.

Willingness to Accept a Job Offer Willingness to accept a job offer was measured using two items – “If offered the position at this company, I would accept the job,” and “I would not accept a job at The Loading Zone” (r). Internal consistency was .89.

Perceived Effort Expended Perceived effort expended represents an individual’s perception of the degree of effort the test required. It was measured using three items – “I feel like I spent a lot of energy on the company’s assessment,” “The company’s hiring assessment wore me out,” and “The company’s hiring process was not at all exhausting (r).” Internal consistency was .91.

Test-Taking Motivation A single item was used as a check to ensure participants had adequate motivation to perform well on these assessments, where on a 1–5 scale respondents indicated how motivated they were to perform.

Test Performance Within each condition, test performance scores were formed by averaging item scores into composites.

Because the tests varied by condition, composite scores were z-scored within condition and then combined to compare relationships across all conditions. Note that by doing this, test performance became unrelated to test length and test type, although it still allows for comparisons with other Ó 2016 Hogrefe Publishing

variables and the use of test performance as a control variable.

Results Before examining results, participant motivation scores were analyzed to determine whether participants were motivated to perform well on the assessment. The mean motivation score was 3.92 (SD = 0.87, Table 1) on a 1–5 scale. The moderately high scores found here suggest that most participants took these tests seriously.

Descriptive statistics for each condition can be seen in Table 2. Overall, participants felt the selection assessments were moderately fair, as the mean procedural justice value was 4.58 (SD = 1.36) on a 1–7 scale. Organizational attractiveness (M = 4.95, SD = 1.36) and likelihood to accept a job offer (M = 5.07, SD = 1.57) had slightly higher mean scores. For perceived effort expended, variance was noticeably higher compared to the other measures, indicating individual differences in the propensity to perceive assessments as exhausting (M = 3.55, SD = 1.74).

Applicant Reactions by Test Length A MANOVA was performed using all dependent and independent variables. Overall, a significant effect of test length was found for a linear composite of dependent variables, Wilks’ Λ = .96, F(3, 200) = 2.82, p < .05. Independent samples t-tests revealed a significant effect of test length on perceptions of procedural justice such that those individuals who took longer selection assessments perceived those assessments to be fairer (d = .40, p < .01). Longer tests were also positively related to organizational attractiveness (d = .41, p < .01) and intentions to accept a job offer (d = .36, p < .01). Increased test length did not negatively impact applicant reactions, and in fact increased the favorability of such reactions overall.

Journal of Personnel Psychology (2016), 15(1), 15–24

20

A. B. Speer et al., Applicant Reactions and Test Length

Table 2. Applicant reaction descriptive statistics by condition Perceived effort expended Condition

Procedural justice

Organizational attractiveness

Intentions to accept job offer

M

SD

M

SD

M

SD

M

SD

Long

3.80

1.73

4.86

1.40

5.23

1.29

5.36

1.47

Short

3.31

1.72

4.31

1.26

4.69

1.37

4.79

1.62

GMA

3.29

1.71

4.45

1.32

4.81

1.41

4.86

1.68

Personality

3.86

1.72

4.73

1.39

5.11

1.27

5.32

1.39

Long GMA

3.36

1.79

5.05

1.30

5.41

1.31

5.39

1.64

Long personality

4.18

1.60

4.69

1.48

5.08

1.27

5.33

1.31

Short GMA

3.24

1.67

4.02

1.16

4.39

1.34

4.48

1.62

Short personality

3.43

1.80

4.79

1.28

5.16

1.29

5.29

1.50

Notes. N = 206. For long conditions, n = 100; for short conditions n = 106; for GMA conditions, n = 111; for personality conditions, n = 95; for the long GMA condition, n = 46; for the long personality condition, n = 54; for the short GMA condition, n = 65; for the short personality condition, n = 41. GMA: General mental ability. Values are on a 1–7 scale.

Applicant Reactions by Test Type There was a nonsignificant effect of test type, Wilks’ Λ = .98, F(3, 200) = 1.16, p = ns. Even though the MANOVA was nonsignificant, it was chosen to examine specific relationships to more fully understand the nature of this independent variable. Upon doing this, applicant reactions were more favorable for personality assessments, although the effects were small. There were weak, nonsignificant effects for perceptions of procedural justice (d = .21, p = ns) and ratings of organizational attractiveness (d = .22, p = ns), and test type was significantly related to likelihood to accept a job offer (d = .29, p < .05), such that perceptions were more favorable for those taking personality assessments.

Figure 1. Test length and procedural justice relationship by test type.

Test Length – Applicant Reaction Relationships by Test Type The test length by test type interaction term was significant, Wilks’ Λ = .95, F(3, 200) = 3.88, p < .01. ANOVAs were performed to examine how test length affected the three operationalizations of applicant reactions for cognitive and personality tests. A significant interaction between test length and test type was found for procedural justice, F(1, 202) = 9.39, p < .01. As depicted in Figure 1, test length had no effect on perceptions of procedural justice for those taking the personality assessment (d = .06, p = ns). However, for those taking the cognitive test, the increase in test length resulted in pronounced favorability of procedural justice ratings (d = .85, p < .01). This finding is quite noteworthy considering that overall, cognitive tests resulted in lower perceptions of procedural justice, and yet as seen in Table 2, those individuals taking the long GMA test had the highest perceptions of procedural justice.

Journal of Personnel Psychology (2016), 15(1), 15–24

Test type moderated the relationship between test length and organizational attractiveness, F(1, 202) = 8.86, p < .01 (Figure 2), and test length and intentions to accept job offer, F(1, 202) = 4.09, p < .05, in similar ways (Figure 3). Individuals taking the GMA test were less likely to view the hypothetical organization as attractive when the GMA test was very short but when the GMA test was longer, ratings of organizational attractiveness greatly increased (d = .77, p < .01). Length did not affect how individuals perceived the company when taking a personality test (d = .06, p = ns). Additionally, the likelihood that a person would accept a job offer from a hypothetical company significantly improved when test length increased for the cognitive assessment (d = .56, p < .01), whereas it did not change for those taking the personality assessment (d = .03, p = ns). Overall then, test length was related to applicant reactions but only for individuals taking cognitive assessments.

Ó 2016 Hogrefe Publishing

A. B. Speer et al., Applicant Reactions and Test Length

21

Table 3. Tests of main and interaction effects of test length and test score on applicant reactions Test length

Test score

Interaction

β

p

β

p

β

p

Procedural justice

.21

< .01

.02

.81

.19

< .01

Organizational attractiveness

.21

< .01

.03

.68

.11

.12

Intentions to accept job offer

.20

< .01

.00

.98

.07

.30

Note. N = 206. Coefficients are standardized.

Table 4. Tests of simple slopes for interaction of test length and test score on applicant reactions

Figure 2. Test length and organizational attractiveness relationship by test type.

Low test scorers

High test scorers

Procedural justice

.02

.40**

Organizational attractiveness

.10

.31**

Intentions to accept job offer

.12

.36**

Notes. All coefficients are standardized. Low test score defined as 1 standard deviation below the mean, high test score defined as 1 standard deviation above the mean. **p < .01.

mean) and low test-scorers (1 SD below the mean), a clear differentiation is seen such that the relationship between test length and procedural justice was much stronger for high scorers (β = .40) than for low scorers (β = .02, Table 4).

Moderation was nonsignificant when predicting organizational attractiveness (β = .11, ΔR2 = .01, p = ns) and intentions to accept a job offer (β = .07, ΔR2 = .01, p = ns), although a similar trend was seen between low and high-scorers (for organizational attractiveness β’s = .10 and .31, and for intentions to accept β’s = .12 and .36).

Figure 3. Test length and intentions to accept job offer relationship by test type.

Additional Post Hoc Analyses Table 1 displays correlations across all study variables.

Across conditions, test performance had no effect on applicant reactions. Additionally, we found no change in relationships after rerunning the MANOVAs with test performance included as a covariate. Without test performance feedback, this is hardly surprising. However, it is possible more qualified applicants, defined here as those with higher test scores, prefer tests that better allow them to display their qualities. To test this, we used moderated regression to examine the conditional effects of test length on applicant reactions via Hayes’ PROCESS program (Hayes, 2013). The results of these analyses are presented in Table 3. We found that the test-length by test score interaction term was significant for procedural justice (β = .19, ΔR2 = .04, p < .01). Looking at these results further after splitting groups into high test-scorers (1 SD above the Ó 2016 Hogrefe Publishing

Discussion Test vendors offer organizations a wide variety of selection assessments that differ in terms of content and length.

Despite the existence of testing packages that vary from minutes to several hours, no research has explicitly examined the effect of test length on applicant reactions. This was the first known study to examine this after individuals actually responded to a selection assessment. We found that longer tests did not adversely affect applicant reactions, which in of itself is a notable finding given that practitioners and organizations often operate under the notion that longer tests will always lead to negative applicant perceptions. To the contrary of this assumption, this study found a null effect of test length for personality assessments and that longer cognitive assessments (roughly 45 min in length) resulted in more favorable ratings of procedural justice, organizational attractiveness, and intentions to accept a job offer than very short cognitive assessments (roughly 10 min in length).

Journal of Personnel Psychology (2016), 15(1), 15–24

22

Practical Implications These findings present interesting guidelines when developing hiring assessments. There are many lengthy commercial assessments available, and when additional measures are added to a selection battery, testing time can reach very high levels and require a great investment on the part of the applicant. In most situations, concessions are made between creating comprehensive and reliable tests versus maintaining a reasonable length of assessment. Findings from this study suggest this compromise may not always be necessary - that increases in test length do not appear to worsen applicant reactions by default. In fact, in some cases and for some respondents, it may improve reactions. As such, it is possible to obtain reliable measures and to incorporate multiple constructs within a selection battery without necessarily sacrificing how applicants perceive the test and company.

This is not to say that carelessly adding additional items to an inventory is recommended, but rather applicant reactions will not necessarily be adversely affected by utilizing a longer, more job-relevant test.3 Given that employers nearly always assume longer tests will result in negative reactions, this is a particularly important and practical finding.

Why Length Might be Related to Applicant Reactions While we found that longer tests were indeed more fatiguing (i.e., perceived effort expended) and this exhaustion adversely impacted applicant reactions (see Table 1), longer cognitive tests had an overall positive effect. This favorable effect was theorized as a result of enhanced perceptions of job relatedness and opportunity for expression. Indeed, results showed that test length was related to perceptions of procedural justice, which incorporates fairness attitudes similar to relevance and opportunity. Of course, without explicitly measuring these two attitudes, it is difficult to make strong inferences in this regard. Interestingly though, we also found that those with higher test scores were more reactive to longer tests than less qualified respondents. As longer assessments allow candidates more of an opportunity to express their traits, it seems that those with favorable trait levels may prefer tests that permit this level of expression.

Cognitive dissonance has also been proposed as an explanation for the relationship between test length and applicant reactions (Rafaeli, 1999). However, there are several reasons to discount cognitive dissonance as a strong causal factor. First and foremost, applicant reactions did not improve as test length increased for personality assessments. Because longer tests require more behavioral effort 3

A. B. Speer et al., Applicant Reactions and Test Length

by the applicant and therefore should result in a greater tension to justify that behavior, if cognitive dissonance does indeed act as a causal mechanism, test length should be positively related to applicant reactions for all tests. Second, and even though dissonance likely does not operate on such a conscious level, those who felt they exhibited a good deal of energy taking the test actually evaluated the test and organizational more negatively. Thus, while it is possible cognitive dissonance explains some portion of the test length-applicant reaction relationship, findings here do not suggest it is a strong influencing factor.

Length by Test Type Interaction Test length did not hinder applicant reactions, but the positive influencing effect was only found for cognitive assessments. Applicant reactions did not significantly differ across short and long personality tests. This may seem perplexing, especially given how cognitive assessments should be more taxing on applicants and therefore more likely to result in fatigue. However, we believe the observed interaction is a function of differences in perceived incremental opportunity of expression.

Cognitive items usually vary in terms of question format and content, and correct answers are not readily discernible to applicants. As such, increasing test length should have a consistently positive effect on opportunity for expression.

On the other hand, correct responses to personality questions are much more apparent to applicants and given similarly structured questions, at some point increases in test length may reach a point of diminishing returns and be simply viewed as unnecessary repetition. Thus, the incremental opportunity to perform likely declines for personality inventories when similar content is added, whereas the incremental opportunity to perform likely increases at constant levels with every new item added to a cognitive test. At what point diminishing returns might be seen for personality tests is unclear, but given these differences, care should be made when implementing longer personality assessments and best practices should be followed by only including job-relevant personality traits within assessments. If indeed diminished returns occur, fatigue as a function of increased test length might potentially worsen applicant reactions overall for very long assessments.

Study Limitations and Areas for Future Research There are several limitations to the current study and areas for additional research. First, we simulated an applicant

It should be noted though that the current study examined test lengths that on average did not exceed 45 min, and thus it is difficult to extrapolate findings to tests that exceed this testing time.

Journal of Personnel Psychology (2016), 15(1), 15â&#x20AC;&#x201C;24

Ă&#x201C; 2016 Hogrefe Publishing

A. B. Speer et al., Applicant Reactions and Test Length

situation as best as possible. However, some may still argue that the situation was not similar enough to actual applicant scenarios to warrant generalization. With that said, care was made to recreate an applicant context by offering participants incentives for their performance and by utilizing a realistic “organizational” website for participants to navigate before taking the hiring tests. Manipulation checks revealed that participants did possess adequate motivation, and anecdotally, participants found the fake organizational website and the testing materials to be professional and engaging. Thus, we believe our findings will apply to applicant populations. To expand upon this further, there are benefits to a lab study within this avenue of research.

The obvious advantage is the ability to easily manipulate independent variables, which is difficult to achieve in field settings. In addition though, lab designs may be advantageous because real applicants are likely to distort survey responses to appear favorable to prospective employers.

A lab experiment allows assurance that response bias is limited. Finally, research shows that laboratory findings do generalize well to field settings (Mitchell, 2012). Taking all these factors together, we believe the lab design was appropriate to address this research question.

Second, only two variations in length were tested. Considering the attenuating effect of dichotomization, one might expect a stronger effect had there been a greater variation in test length. On the other hand, it is also possible a curvilinear effect may be found such that moderate test lengths result in optimal applicant reactions, and at very short (roughly 10 min) or very long test lengths (roughly several hours or more), applicant reactions will suffer. This might happen if at some point the perceived benefits of increased opportunity to express oneself reaches a point of diminishing returns, and excessive testing times might be more likely to result in extreme states of fatigue and boredom. As such, it would be beneficial to examine how test length affects applicant reactions across numerous gradations of testing time.

Third, although the study provides valuable knowledge regarding applicant reactions, it likely created more questions than it answered. As an initial test of the effects of test length, we offered theory as to why these variables should relate (and show that they do), but results are insufficient to fully understand the causal paths relating test length to applicant reactions. Future research should involve more explicit tests of causal mechanisms such as cognitive dissonance and job relatedness. For example, the face validity and length of the test could be manipulated to isolate whether longer tests are more favorably perceived because of a greater amount of job-relevant content or simply because applicants invested more time in the selection process. In addition, procedural justice could be split into subfacets including perceived job relevance and opportunity Ó 2016 Hogrefe Publishing

23

for trait expression. The Selection Procedural Justice Scale

by Bauer and colleagues (2001) would be a good option for this.

Finally, the relationship between assessment length and applicant reactions should be examined for a wider variety of selection tests (e.g., biodata, in-person assessments such as interviews), and future research could investigate the impact of testing context in moderating this relationship.

To this last point, participants in this study arrived to a testing center with the expectation they’d be taking a test. In a different setting, such as for those responding online and at home, it is possible distractions and alterative home activities might alter perceptions during the testing process, ultimately affecting the test length by applicant reactions relationship. Additional research on this topic is warranted.

Conclusions Although applicant reactions may not be a chief concern of HR practitioners, it is a concern that is important nonetheless. This study provides initial evidence that the length and testing time of an assessment may affect the reactions of job applicants. Contrary to typical assumptions, increased test length did not result in aversive applicant reactions, and for cognitive assessments longer tests may actually be perceived more favorably. That said, there is still much to be learned about how test length affects applicant reactions, and we urge that further attention be given to this very practical topic.

References Ackerman, P. L., & Kanfer, R. (2009). Test length and cognitive fatigue: An empirical examination of effects on performance and test-taker reactions. Journal of Experimental Psychology: Applied, 15, 163–181.

Anderson, N., Salgado, J. F., & Hülsheger, U. R. (2010). Applicant reactions in selection: Comprehensive meta-analysis into reaction generalization versus situational specificity. International Journal of Selection and Assessment, 18, 291–304.

Aronson, E. (1992). The return of the repressed: Dissonance theory makes a comeback. Psychological Inquiry, 3, 303–311.

Bauer, T. N., McCarthy, J., Anderson, N., Truxillo, D. M., & Salgado, J. F. (2012). What we know about applicant reactions to selection: Research summary and best practices (SIOP White Paper Series). Retrieved from http://www.siop.org/WhitePapers/ White%20Paper%20Series%2020112012ApplicantReactions. pdf Bauer, T. N., Truxillo, D. M., Sanchez, R. J., Craig, J. M., Ferrara, P., & Campion, M. A. (2001). Applicant reactions to selection: Development of the selection procedural justice scale (SPJS).

Personnel Psychology, 54, 388–420.

Becker, W. J., Connolly, T., & Slaughter, J. E. (2010). The effect of job offer timing on offer acceptance, performance, and turnover. Personnel Psychology, 63, 223–241.

Journal of Personnel Psychology (2016), 15(1), 15–24

24

Cialdini, R. B. (2009). Influence: Science and practice (5th ed.).

New York, NY: Pearson Education.

Cialdini, R. B., Trost, M. R., & Newsom, J. T. (1995). Preference for consistency: The development of a valid measure and the discovery of surprising behavioral implications. Journal of Personality and Social Psychology, 69, 318–328.

Colquitt, J. (2001). On the dimensionality of organizational justice: A construct validation of a measure. Journal of Applied Psychology, 86, 386–400.

Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences of forced compliance. Journal of Abnormal and Social Psychology, 58, 203–210.

Gilliland, S. W. (1993). The perceived fairness of selection systems: An organizational justice perspective. Academy of Management Review, 18, 694–734.

Gilliland, S. W., & Steiner, D. D. (2001). Causes and consequences of applicant perceptions of unfairness. In R. Cropanzano (Ed.), Justice in the workplace: From theory to practice (Vol. 2, pp.

175–195). Mahwah, NJ: Erlbaum.

Goldberg, L. R., Johnson, J. A., Eber, H. W., Hogan, R., Ashton, M. C., Cloninger, C. R., & Gough, H. C. (2006). The International Personality Item Pool and the future of public-domain personality measures. Journal of Research in Personality, 40, 84–96.

Greenberg, J. (1990). Organizational Justice: Yesterday, today, and tomorrow. Journal of Management, 16, 399–432.

Hausknecht, J. P., Day, D. V., & Thomas, S. C. (2004). Applicant reactions to selection procedures: An updated model and meta-analysis. Personnel Psychology, 57, 639–683.

Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach.

New York, NY: Guilford.

König, C. J., Klehe, U. C., Berchtold, M., & Kleinmann, M. (2010).

Reasons for being selective when choosing personnel selection procedures. International Journal of Selection and Assessment, 18, 17–27.

Lievens, F., Van Hoye, G., & Schreurs, B. (2005). Examining the relationship between employer knowledge dimensions and organizational attractiveness: An application in a military context. Journal of Occupational and Organizational Psychology, 78, 553–572.

Mitchell, G. (2012). Revisiting truth or triviality: The external validity of research in the psychological laboratory. Perspectives on Psychological Science, 7, 109–117.

Rafaeli, A. (1999). Pre-employment screening and applicants’ attitudes toward an employment opportunity. The Journal of Social Psychology, 139, 700–712.

Rosse, J. G., Miller, J. L., & Stecher, M. D. (1994). A field study of job applicants’ reactions to personality and cognitive ability testing. Journal of Applied Psychology, 79, 987–992.

Ryan, A. M., & Ployhart, R. E. (2000). Applicants’ perceptions of selection procedures and decisions: A critical review and agenda for the future. Journal of Management, 26, 565–606.

Schleicher, D. J., Venkataramani, V., Morgeson, F. P., & Campion, M. A. (2006). So you didn’t get the job. now what do you think?

Examining opportunity-to-perform fairness perceptions. Personnel Psychology, 59, 559–590.

Truxillo, D. M., & Bauer, T. N. (2011). Applicant reactions to organizations and selection systems. In S. Zedeck (Ed.), APA handbook of industrial and organizational psychology (Vol. 2, pp.

379–397). Washington, DC: American Psychological Association.

Truxillo, D. M., Bauer, T. N., Campion, M. A., & Paronto, M. E. (2002). Selection fairness information and applicant reactions: A longitudinal field study. Journal of Applied Psychology, 87, 1020–1031.

Journal of Personnel Psychology (2016), 15(1), 15–24

A. B. Speer et al., Applicant Reactions and Test Length

Received August 4, 2014 Revision received December 23, 2014 Accepted December 29, 2014 Published online April 12, 2016 Andrew B. Speer American Family Insurance 6000 American Parkway Madison, WI 53703 USA Tel. +1 608 242-4100 E-mail speerworking@gmail.com

Appendix List of Scales and Items Used Procedural Justice The company’s hiring assessment gave me the opportunity to accurately represent myself.

The company’s hiring assessment is a fair method for selecting employees.

The company’s hiring assessment seemed free of bias.

The company’s hiring procedure accurately assessed my ability as a worker.

The company’s hiring procedures are ethical and moral.

Organizational Attractiveness This company would be a good place to work.

This company is attractive to me as a place for employment.

A job at The Loading Zone is appealing to me.

The Loading Zone does not seem to be a very good place to work.

Willingness to Accept a Job Offer If offered the position at this company, I would accept the job.

I would not accept a job at The Loading Zone.

Perceived Effort Expended I feel like I spent a lot of energy on the company’s assessment.

The company’s hiring assessment wore me out.

The company’s hiring process was not at all exhausting.

Test-Taking Motivation How motivated were you to perform well on the assessment?

Ó 2016 Hogrefe Publishing

Understanding and preventing destructive leadership Topics covered include • The causes and consequences of bad leadership • A multiple pathway model linking charismatic leadership attempts and abusive supervision • Coping with destructive leadership: putting forward an integrated theoretical framework for the interaction process between leaders and followers • Inside the minds of narcissists: how narcissistic leaders’ cognitive processes contribute to abusive supervision • The relative effects of constructive, laissez-faire, and tyrannical leadership on subordinate job satisfaction: results from two prospective and representative studies • Absence makes the errors go longer: how leaders inhibit learning from errors

Birgit Schyns / Jan Schilling (Editors)

Destructive Leadership (Series: Zeitschrift für Psychologie – Vol. 222/4) 2014, iv + 62 pp., large format US $49.00 / € 34.95 ISBN 978-0-88937-464-5 Destructive leadership can take various forms, such as abusive supervision, petty tyranny, and negative or aversive leadership, all of which negatively impact both followers and the organizations themselves. The expert contributions in this volume focus on understanding and preventing destructive leadership and the farreaching consequences it can have on individuals and organizations.

Destructive forms of leadership emerge in an interaction between situations, leaders, and followers. Critical aspects

www.hogrefe.com

of this interaction examined here include the ways in which attempts at charismatic leadership can lead to the perception of abusive leadership, or even to abusive leadership behavior, narcissism as an antecedent of abusive leadership, how leaders and followers interact in coping with negative leadership (different coping strategies and their consequences), the impacts of constructive and destructive leadership behavior on subordinates’ job satisfaction, and the role of leaders in learning from workplace mistakes.

Budapest

Istanbul

Hamburg

KĂśln

Moskau

Wien

Stuttgart

50 exhibitors More than 5 tions & 250 presenta | 8 Keynotes discussions ade visitors over 8,0 0 0 tr

26-27 April 2016 Hamburg

10-11 May 2016 Stuttgart ÂŽ

g

y

eo

p

Twitter: #Pnord16

pl

e

per

for m a n c e

h tec

n

o ol

Twitter: #Psued16

www.personal-messe.com For managing directors, HR decision-makers and members of HR departments in general as well as heads of IT departments At the same time in Stuttgart

Join us!

SAVE 4n0li% n e!

Register O

6th Exhibition for Workplace Health Promotion and Demography Main Media Partners

Original Article

Participative Supervisory Behavior and the Importance of Feeling Safe and Competent to Voice Mari Svendsen, Thomas S. Jønsson, and Christine Unterrainer Department of Psychology and Behavioral Sciences, Aarhus University, Denmark

Abstract: In our field study of 147 employees and their supervisors, we tested a moderated mediation model, investigating how participative supervisory behavior relates differently to promotive and prohibitive voice. Overall, we found a significant effect of participative supervisory behavior on promotive and prohibitive voice, and this effect was mediated by psychological safety for prohibitive voice, but not promotive voice.

Unexpectedly, we did not find a direct moderation effect of occupational self-efficacy. However, our results suggest that occupational selfefficacy creates a conditional indirect effect for prohibitive voice. Accordingly, our results shed light on the boundary conditions of participative supervisory behavior and illustrate the usefulness of conceptualizing voice as promotive and prohibitive. Implications for theory and practice are discussed.

Keywords: promotive voice, prohibitive voice, leadership, psychological safety, occupational self-efficacy

Employees in modern organizations often have a crucial knowledge regarding the challenges and opportunities concerning the efficient functioning of the organization (Morrison, 2011). Unfortunately, many employees choose not to speak up about their knowledge due to fear of negative consequences from supervisors or peers (Milliken, Morrison, & Hewlin, 2003). However, supervisors are increasingly becoming aware of the potential resources which might be utilized by receiving the employees’ input on organizational matters. Thus, during the last two decades, we have seen a renewed interest in how supervisory behavior is related to employee voice, which is a concept referring to employees’ discretionary communication of ideas, suggestions, and concerns at work, with the intent to improve organizational functioning (Le Pine & Van Dyne, 1998; Morrison, 2011). Recent empirical (Liang, Farh, & Farh, 2012) and theoretical (Morrison, 2011; Van Dyne, Ang, & Botero, 2003) works suggest that there may be different motivational forces behind promotive voice, referring to the individuals’ expression of ideas and suggestions, compared to prohibitive voice, which refers to the individuals’ communication of concerns regarding factors which may cause harm to the organization. Nevertheless, there has been no research to our knowledge that investigates how supervisory behavior may relate to the two differently. This knowledge is crucial however, in order to explore how organizations can increase both promotive and prohibitive voice. Ó 2016 Hogrefe Publishing

In our study we focus on participative supervisory behaviors (PSB) such as consulting with and listening to the employees, as our key predictor variable. The main body of research finds a positive relationship between PSB and voice (Morrison, 2011). For example, a longitudinal study

by Detert and Burris (2007) finds a positive relationship between leadership openness and improvement-oriented voice. Additionally Tangirala and Ramanujam (2012) find a positive relationship between leaders who show consulting behavior and improvement-oriented voice. Other studies however find a weak or no relationship between the two phenomena. For example, Janssen, de Vries, and Cozijnsen (1998) do not find a significant relationship between leadership openness and the propensity to voice novel ideas.

Furthermore, Dutton, Ashford, Lawrence, and MinerRubino (2002) find no effect of leader openness on the employees’ propensity to speak up with their ideas or solutions. These inconsistent results suggest that there may be important boundary conditions regarding the effect of PSB, which lack specification in prior studies. An exploration of these boundary conditions is of theoretical and practical importance in order to understand when PSB is effective and how to best facilitate the transmission from PSB to voice. Thus, the main research question of our study is: under which conditions do PSB predict promotive as well as prohibitive voice most efficiently?

We intend to explore this question, and overcome some of the shortcomings from prior studies, by testing a model Journal of Personnel Psychology (2016), 15(1), 25–34 DOI: 10.1027/1866-5888/a000146

26

where we simultaneously include both moderators and mediators in the relationship between PSB and promotive as well as prohibitive voice (Ashford, Sutcliffe, & Christianson, 2009). Specifically we test a model proposing that psychological safety, referring to the individuals’ beliefs in how safe it is to speak up in the group, mediates the effect of PSB on promotive and prohibitive voice (Edmondson, 1999). Moreover, we propose that occupational self-efficacy (OSE), which refers to the employee’s general feeling of competence in doing his or her work (Schyns & Von Collani, 2002), moderates the relationship between PSB and promotive and prohibitive voice, both directly and indirectly.

Theory and Hypotheses Voice as a Dual Construct Voice behavior is a part of the broader proactivity literature that emphasizes the importance of self-started activities aimed at fostering change in the employees’ working environment (Frese & Fay, 2001; see Morrison, 2011 for a review of similarities and differences between voice behavior and other proactivity concepts). Voice is also often conceptualized as a dimension of organizational citizenship behavior (OCB), as it involves discretionary behavior exerted by employees who are willing to “go the extra mile” for the organization (Detert & Burris, 2007). However, voice is conceptually distinct from other forms of OCB as it is inherently more challenging than, that is, altruism or sportsmanship (Le Pine & Van Dyne, 1998). Furthermore, as it concerns the employees’ discretionary initiative to get own ideas communicated to the leader, it is also conceptually different from OCB such as civic virtue which to a larger extent concerns the employees’ active participation in the life of the organization.

Although voice is proposed as a challenging form of OCB, most of the literature within the voice area has focused on voice as a construct relating to improvementoriented suggestions, such as making innovative suggestions for change or new products (Morrison, 2011).

However, recent theoretical and conceptual developments have suggested the necessity of also incorporating speaking up with concerns and worries as part of the voice concept (e.g., Morrison, 2011; Van Dyne et al., 2003). In line with this suggestion, Liang et al. (2012) distinguish between promotive voice and prohibitive voice. Promotive voice is future oriented and focuses on “realizing ideals and possibilities” (Liang et al., 2012, p. 75) whereas prohibitive voice is both past- and future oriented and concerns “stopping and preventing harm” in the organization (Liang et al., 2012, p. 75). Consequently, prohibitive voice is potentially more risky for the individual than promotive voice, as calling Journal of Personnel Psychology (2016), 15(1), 25–34

M. Svendsen et al., Participative Supervisory Behavior and Voice

attention to factors harmful to the organization may implicate a failure of someone who is responsible (Burris, 2012; Liang et al., 2012).

Correspondingly, Morrison (2011) argues that the strength of the motivating forces predicting voice may differ depending on its promotive or prohibitive content. Specifically, Morrison (2011) suggests that self-protective motives may be an especially important motivating factor with respect to prohibitive voicing, due to the increased interpersonal risk associated with it. Promotive voice however, is involving less personal risk and a general perception of being able to influence the working environment is thus argued to be more important (Morrison, 2011). The notion that self-protective motives are more important for prohibitive voice is empirically supported by Liang et al. (2012).

Specifically, Liang et al. (2012) find evidence suggesting that psychological safety is significantly stronger related to prohibitive voice, compared to promotive voice. Furthermore, the empirical findings of Tangirala and Ramanujam (2012) lend some support for the notion that a general perception of being able to influence the working environment is crucial for promotive voice. Specifically, they find that the experience of influence is positively related to improvement- oriented voice when controlling for psychological safety.

In order to acknowledge these findings, we formulate our hypotheses regarding promotive and prohibitive voice separately. Thus, we are the first study to our knowledge to explore how PSB may relate to promotive and prohibitive voice differently.

The Total Effect of Participative Supervisory Behavior on Voice Due to the inherent risk involved when employees express themselves, supervisory behaviors are crucial for the employees’ propensity to voice (Milliken et al., 2003; Morrison, 2011). Scholars argue this is because supervisors often have the resources to punish or reward the employees and to react on the employees’ suggestions or concerns (Liu, Zhu, & Yang, 2010). Theoretically, having a participative supervisor, that is, a supervisor who discusses problems and situations with their followers, listens to their ideas and suggestions, and lets the employees influence decisions that are important to them (Vroom & Jago, 1988), may strengthen the employees’ willingness to express themselves. Based on expectancy theory, a participative supervisor increases the employees’ motivation to voice by giving the employees an expectation of being listened to and valued for their input (Vroom & Jago, 1988). We argue that this motivates the employee to voice both promotive and prohibitive. Furthermore, Spreitzer (1995) argues that the Ó 2016 Hogrefe Publishing

M. Svendsen et al., Participative Supervisory Behavior and Voice

opportunity to participate may increase the employees’ feelings of making an impact at work. Thus, they will also feel more responsible for speaking up, something which Liang et al. (2012) empirically confirm is significantly related to promotive and prohibitive voice.

Empirically, the importance of PSB for promotive voice is supported by, for example, Tangirala and Ramanujam (2012), who find that supervisory consulting behaviors are positively related to improvement-oriented voice. The relationship between PSB and prohibitive voice has to our knowledge never been investigated. However, for example, Kassing (2001) finds a positive relationship between management openness and employee dissent, which is a proactivity construct which shares similarities with prohibitive voice (Morrison, 2011). Thus, based on the presented theoretical arguments and prior empirical studies we propose the following hypotheses: Hypothesis 1a/b: PSB, as perceived by the employees, are positively related to promotive voice and prohibitive voice.

The Indirect Effect of Participative Supervisory Behavior Through Psychological Safety Psychological safety is defined as the employee’s perception of his or her working environment as being safe for interpersonal risk-taking, so that expressing oneself will not lead to negative consequences or sanctions such as being ridiculed, criticized, or losing support from peers and supervisors (Detert & Burris, 2007; Edmondson, 1999). Supervisory behavior is crucial for creating a psychologically safe environment because supervisors also control important sanctions or rewards (Liu et al., 2010). Hence, a supervisor showing openness to and appreciation of suggestions or concerns helps the employee in obtaining the perception that expressing oneself will not lead to negative consequences for the individual (Detert & Burris, 2007).

We further argue that psychological safety is positively related to both promotive and prohibitive voice, as the employee’s perception of psychological safety decreases the risks associated with speaking up by signaling to the employee that doing so will not lead to negative consequences (Detert & Burris, 2007). Empirically, the mediating effect of psychological safety for improvement-oriented voice has been established by Detert and Burris (2007).

However, Morrison (2011) argues, and Liang et al. (2012) empirically confirm that self-protective motives such as psychological safety are significantly stronger related to prohibitive voice compared to promotive voice. In the same vein, as a general perception of influence is argued to be Ó 2016 Hogrefe Publishing

27

more strongly related to promotive voice, the direct effect of PSB should be stronger for promotive than prohibitive voice. Accordingly, we propose that the mediating (indirect) effect of psychological safety is stronger for prohibitive voice rather than promotive voice. Thus, we posit the following hypotheses: Hypothesis 2a/b: The employee’s perceived psychological safety mediates the relationship between the employee’s perceived PSB and promotive voice and prohibitive voice.

Hypothesis 2c: The mediating effect of psychological safety will be stronger for prohibitive voice than for promotive voice.

The Differential Effects of Occupational Self-Efficacy Occupational self-efficacy (OSE) refers to the capacities, competences, and abilities, as perceived by an employee, regarding successful performance of the tasks involved in his or her job (Schyns & Von Collani, 2002). Furthermore, OSE contributes to the employee’s general feeling of confidence at work and thus affects the employee’s perception of contextual aspects in the organization (Bandura, 1997; Le Pine & Van Dyne, 1998; Tangirala & Ramanujam, 2012). Based on social cognitive theory (e.g., Bandura, 1997) we argue that the employees’ OSE moderates both the direct and indirect effects of PSB. The Conditional Direct Effect of OSE Bandura (1997) argues that one of the main outcomes of having high efficacy beliefs is agentic individuals who behave proactively, and with more positive outcome expectations. We argue that employees who are more proactive, and with expectations that voicing will lead to positive outcomes (i.e., improved working procedures), will be more motivated to voice than individuals low in efficacy beliefs.

However, Tangirala and Ramanujam (2012) argue that in order for the employee to voice, he or she must also experience a sense of influence. Accordingly, the individual needs access to a supervisor who provides the employee with direct contact and the opportunity to influence decisions. Tangirala and Ramanujam (2012) find empirical evidence suggesting that the combination of high OSE and a participative supervisor is related to a perception of influence, which in turn increases the employees’ propensity to voice, when controlling for psychological safety. Thus, we argue that in combination with participative practices, OSE will moderate the direct effect of PSB on promotive voice and prohibitive voice. Nevertheless, as previously argued a general feeling of influence may be more Journal of Personnel Psychology (2016), 15(1), 25–34

28

important for promotive voice than for prohibitive voice.

Hence, we offer the following hypotheses: Hypothesis 3a/b: OSE positively moderates the direct relationship between PSB and promotive and prohibitive voice, so that the direct relationship will be stronger when the level of OSE is higher.

Hypothesis 3c: The moderating effect of OSE will be stronger for promotive voice than for prohibitive voice.

The Conditional Indirect Effect of OSE Prior research illustrates how a supervisor showing a high degree of PSB, stimulates a psychological safe working climate by signaling that it is safe and worthwhile to speak up (Detert & Burris, 2007). Furthermore, empirical studies show that employees with both high and low self-efficacy will look to their leader in order to establish whether it is safe and worthwhile to speak up (Milliken et al., 2003).

In line with this reasoning, Vroom and Jago (1988) argue that individuals who feel they have the necessarily abilities (high OSE) to respond to participatory practices also feel safer and more comfortable when given influence and autonomy. Thus, the employees’ feeling of having something qualified to say when invited by their supervisor, contributes to an increased feeling of psychological safety for the employee (Conger & Kanungo, 1988). Therefore, we argue that a high level of OSE will further strengthen the effect of PSB on psychological safety, by giving the employee confidence that he or she has something valuable to contribute with, when asked to participate. Correspondingly, we propose that the indirect effect of PSB on promotive and prohibitive voice underlies the effect of OSE, suggesting a conditional indirect effect (Preacher, Rucker, & Hayes, 2007). A conditional indirect effect implies that the mediating effect of psychological safety will be stronger for those individuals who are also high in OSE. We argue however, that this conditional indirect effect will be stronger for prohibitive voice, as a strengthening of self-protective motives (psychological safety) is empirically found to be more important for prohibitive voice than for promotive voice (Liang et al., 2012). Thus, we offer the following hypotheses: Hypothesis 4a/b: OSE moderates the relationship between the employee’s perceived PSB and employee’s perceived psychological safety. Thus, the mediating effect of the employee’s perceived psychological safety on promotive and prohibitive voice underlies the effect of OSE, so that the effect will be stronger when the employee’s OSE is higher.

Hypothesis 4c: The conditional indirect effect OSE is stronger for prohibitive voice than for promotive voice.

Journal of Personnel Psychology (2016), 15(1), 25–34

M. Svendsen et al., Participative Supervisory Behavior and Voice

Figure 1 gives a summary of the hypotheses, together with an illustration of the full conceptual model and coefficients.

Methods Sample and Procedure The survey was conducted among all the white-collar workers from a medical technology firm in Norway. A link to the survey questionnaire was sent to the 261 employees, via their professional e-mail addresses. The participants were told that participation was voluntary and confidential but that their e-mail addresses would be used in order to match the employees with their supervisor.

A total of 147 employees fully completed the survey questionnaire, resulting in a response rate of 56%. The employees were not organized in teams, but in small units that encompassed one supervisor. A total of 29 supervisors/ units were represented by more than one rater, ranging from 2 to 13 raters. The sample consisted of 37% women and 63% males. The age ranged from 23 to 71 years, and the mean age was 42.39 years (SD = 8.78).

Measures Scales that were originally formulated in English were translated to Norwegian, and then translated back to English in order to make sure that the linguistic content was still accurate (Brislin, 1986). If not otherwise noted, the questions were rated on a 5-point Likert scale (ranging from 1 = strongly disagree to 5 = strongly agree). All the scales showed satisfactory reliabilities (see Cronbach’s α in Table 1). The confirmatory factor analyses (CFAs) of the scales are reported in Table 2.

Participative Supervisory Behavior Participative supervisory behavior was measured using a four-item scale developed by Rank (2006). The questions were rated on a 5-point Likert scale (1 = not at all to 5 = often if not always). A sample item is: “My immediate supervisor generally gives me a lot of influence on what goes on in my work unit.”

Promotive and Prohibitive Voice Promotive and prohibitive voice were measured by means of a 10-item scale developed by Liang et al. (2012). Five items measured promotive voice, and five items measured prohibitive voice. A sample item for the promotive voice scale is: “I suggest new projects which may be beneficial to the work unit,” and a sample item from the prohibitive scale is: “I speak up honestly with problems that might Ó 2016 Hogrefe Publishing

M. Svendsen et al., Participative Supervisory Behavior and Voice

29

Figure 1. Hypothesized moderated mediation model and coefficients. The model is tested as one overall model but is presented separately for the dependent variables, in order to improve the visual representation. The stippled lines indicate that the effect is hypothesized to be stronger for this dependent variable.

Control variables Occupational self-efficacy H3a ns.

H4a .14*

Psychological safety H2a ns.

H2a .43**

Participative supervisory behavior

Promotive voice H1a .21* (total effect) .15* (direct effect)

Control variables Occupational self-efficacy H3b ns.

H4b .14*

Psychological safety H2b .28**

H2b .43**

Participative supervisory behavior

Prohibitive voice H1b .20* (total effect) ns. (direct effect)

Table 1. Means, standard deviations, correlations, and Cronbach’s α among variables Variables

Mean

SD

1

2

3

4

5

6

7

8

9

0.12

0.33

–

42.39

8.78

.18*

–

Job satisfaction

3.60

1.12

.04

.07

–

4.

Gender

0.37

0.49

.06

.06

.02

5.

Education

0.78

0.42

.09

.22**

.14

.22**

–

6.

Participatory supervisory behavior

3.80

0.87

.21**

.02

.32**

.21**

.10

(.90)

7.

Occupational self-efficacy

3.96

0.36

.05

.07

.01

.17*

.15

.05

8.

Psychological safety

4.14

0.74

.21**

.03

.20*

.11

.00

.49**

.09

(.95)

9.

Promotive voice behavior

3.90

0.52

.24**

.08

.16*

.20*

.29**

.25**

.34**

.26**

(.83)

10.

Prohibitive voice behavior

3.64

0.52

.20*

.22**

.01

.03

.05

.25**

.12

.36**

.42**

1.

Leadership responsibility

2.

Age

3.

10

–

(.79)

(.71)

Note. N = 147. Cronbach reliabilities appear in parenthesis along the diagonal. *p < .05, **p < .01.

Table 2. Confirmatory factor analysis of scales Scale

w2

df

p

RMSEA

90% CI

CFI

TLI

SRMR

Participative supervisory behavior

0.65

2

.72

.00

[.00, .12]

1.00

1.02

.01

Promotive voice

4.63

4

.33

.03

[.00, .13]

1.00

0.99

.03

Prohibitive voice

2.44

5

.79

.00

[.00, .08]

1.00

1.05

.02

26.42

18

.09

.06

[.00, .10]

0.96

0.94

.05

Occupational self-efficacy

Note. For occupational self-efficacy and promotive voice, residual correlations were added on.

Ó 2016 Hogrefe Publishing

Journal of Personnel Psychology (2016), 15(1), 25–34

30

cause serious loss to the work unit; even when/though dissenting opinions exist.”

Psychological Safety Psychological safety was measured by a 3-item scale adopted from Detert and Burris (2007) as these items “better tap the individual-level assessment of psychological safety” (Detert & Burris, 2007, p. 872). A sample item is: “I think it is safe for me to make suggestions in the work unit.” The model was saturated and a CFA model was not applicable, as the use of the overall goodness-of-fit test is not possible.

Occupational Self-Efficacy Occupational self-efficacy was measured using an eightitem scale developed by Schyns and Von Collani (2002).

A sample item is: “No matter what comes my way in my job, I’m usually able to handle it.”

Control Variables Based on previous literature, we added age, gender, supervisory responsibility, job satisfaction, and education as control variables (Morrison, 2011). Age was measured by years and months. Gender was dummy coded (0 = female, 1 = male). Supervisory responsibilities were dummy coded based on records from the organization (0 = not supervisor and 1 = supervisor). Job satisfaction was measured by the item “All together, how satisfied are you with your job?”

The response categories were measured on a 5-point Likert Scale, ranging from 1 = very dissatisfied to 5 = very satisfied.

Education was measured by five categories (elementary school, high school, undergraduate studies, bachelor level degree, and master level or postgraduate studies) and further dummy coded to academics (0) and nonacademics (1).

Analytic Approach In the current study, we follow the lines of Tangirala and Ramanujam (2012) by conceptualizing PSB at a dyadic level because supervisors often behave differently in relation to individual employees (Graen & Scandura, 1987). Consequently, we are interested in the employees’ individual perceptions of their supervisor, not their supervisor’s general leadership style. This argument was further supported by our low ICCs, as the ICC(1) for PSB was .06, and the ICC (2) was .18 (Bliese & Halverson, 1998) Nevertheless, because our study consists of employees who are drawn from separate units, represented by one leader, our sample violates the independence assumption and may result in spuriousness due to data clustering (Raudenbush & Bryk, 2002). Thus, we made a cluster robust analysis in Mplus, version 7.11, using manifest variables, in order to adjust for the clustered nature of our data.

Journal of Personnel Psychology (2016), 15(1), 25–34

M. Svendsen et al., Participative Supervisory Behavior and Voice

Results Confirmatory Factor Analyses In order to confirm the measures’ factor structures and to rule out the likelihood of conceptual overlap, we conducted a CFA of all applied items loading on five latent factors and compared it with a single-factor-model (Podsakoff, MacKenzie, Scott, Lee, & Podsakoff, 2003). We used the statistical software, Mplus version 7.11, applying Maximum Likelihood Robust (MLR) estimation of model fit. The measurement model with five latent factors demonstrated a satisfactory fit: w2(262) = 365.37, p < .001, RMSEA = .05, 90% CI [.04, .06], CFI = .93, TLI = .92, SRMR = .08. In contrast, the model fits of the single-factor-model were not acceptable: w2(273) = 1,132.49, p < .001, RMSEA = .15, 90% CI [.14, .16], CFI = .41, TLI = .35, SRMR = .16. A SatorraBentler scaling corrected w2 difference test (Satorra & Bentler, 1999) showed that the 5-factor-model was significantly better than the 1-factor-model (Δw2(11) = 648.07, p < .001). In order to assess whether or not the two forms of voice were separate constructs, we also tested a measurement model, in which all voice items only formed one factor.

The w2 difference test (Satorra & Bentler, 1999) showed that the model that distinguished between the two forms of voice was significantly better than a model with only one factor that measured voice (Δw2(1) = 59.17, p < .001).

Descriptive Statistics and Control Variables The means, standard deviations, correlations, and Cronbach’s α for the variables are presented in Table 1.

As the inclusion of all control variables did not change the significance levels of any of our dependent variables, we excluded insignificant controls from our final analyses for the sake of parsimony.

Test of the Total and Indirect Effects Table 3 illustrates the total and indirect effects of our model. As can be seen in Table 3, we gained support for Hypothesis 1a/b as the total effect of PSB was significantly related to promotive voice β = .21, SE = .09, t = 2.29, p = .02, and prohibitive voice β = .20, SE = .07, t = 2.93, p < .001.

Furthermore, we did not find support for Hypothesis 2a, as we did not find a significant mediational effect of psychological safety for promotive voice. However, we found support for Hypothesis 2b, as psychological safety mediated the relationship between PSB and prohibitive voice (indirect effect: β = .12, SE = .06, t = 2.20, p = .03). This was a full mediation as the direct effect of PSB became insignificant. Ó 2016 Hogrefe Publishing

M. Svendsen et al., Participative Supervisory Behavior and Voice

31

Table 3. Regression results Predictor

β

SE

t

4.9

p

4.7

PSB

.43

.06

7.18

.00

Occupational self-efficacy (OSE)

.09

.08

1.13

.24

PSB OSE

.14

.07

2.12

.03

Leadership responsibility

.11

.05

2.08

.04

Promotive voice Total PSB

.21

.09

2.29

.02

Direct

.15

.07

2.12

.03

Indirect PSB

.06

.03

1.77

.08

OSE

.34

.09

3.81

.00

PSB OSE

.16

.09

1.77

.08

Psychological safety

.14

.07

1.86

.06

Leadership responsibility

.12

.05

2.28

.02

Job satisfaction

.19

.08

2.49

.01

Education

.17

.06

2.81

.01

Total PSB

.20

Prohibitive voice .07

2.93

Psychological safety

Psychological safety

4.5 4.3 4.1 3.9 Low OSE 3.7

High OSE

3.5 Low PSB

High PSB

Figure 2. Interactional effect of PSB and OSE on psychological safety (PSB = participatory supervisory behavior; OSE = occupational selfefficacy).

.00

Table 4. Conditional indirect effect of occupational self-efficacy (mediated by psychological safety)

Direct PSB

.08

.05

1.43

.15

Occupational self-efficacy

Indirect PSB

.12

.06

2.20

.03

Level

OSE

.12

.09

1.30

.19

1 SD

.045

.027

.094

PSB OSE

.16

.10

1.51

.13

+1 SD

.080

.033

.017

Psychological safety

.28

.11

2.68

.01

Age

.19

.08

2.40

.02

Note. N = 147. PSB = participative supervisory behavior; OSE = occupational self-efficacy.

The fact that psychological safety only had a mediational effect for prohibitive voice supports Hypothesis 2c, suggesting that the mediational effect will be stronger for prohibitive voice.

Conditional Direct Effect As can be seen in Table 3, Hypothesis 3a/b/c, was not supported as we did not find any significant interaction effect of OSE and PSB on promotive and prohibitive voice. We only found main effects of PSB and OSE on promotive voice.

Conditional Indirect Effect As there was no significant mediational effect of psychological safety on promotive voice we only tested the conditional indirect effect of OSE on prohibitive voice (Hayes, 2013). We first tested for a significant interaction effect, showing that OSE significantly moderated the relationship between PSB and psychological safety β = .14, SE = .07, t = 2.12, p = .03. The moderation effect is illustrated in Figure 2.

To further test Hypothesis 4b, we followed a procedure recommended by Preacher et al. (2007). We therefore operationalized high or low levels of OSE as one standard Ó 2016 Hogrefe Publishing

B

SE

p

Note. N = 147.

deviation above and below the mean score of OSE. Accordingly, we examined to which degree the indirect effect of PSB on prohibitive voice, via psychological safety, differs in strength across employees high and low in OSE. As can be seen in Table 4, Hypothesis 4b was supported as the conditional indirect effect of PSB on prohibitive voice was stronger and significant for the individuals high in OSE; B = .08, SE = .03, p = .02, and insignificant for the individuals low in OSE. The fact that a conditional indirect effect existed for prohibitive voice, but not for promotive voice, supports Hypothesis 4c, suggesting that the conditional indirect effect of OSE is stronger for promotive rather than prohibitive voice.

Discussion Summary of Results We gained support for Hypotheses 1a/b, 2b, and 2c, confirming that the total effect of PSB was significantly related to both promotive and prohibitive voice and that this effect was mediated by psychological safety for prohibitive voice.

Contrary to our expectations, we did not find a significant mediating effect of psychological safety on promotive voice (Hypothesis 2a), something which supports Hypothesis 2c, Journal of Personnel Psychology (2016), 15(1), 25–34

32

which proposes that the mediating effect of psychological safety is stronger for prohibitive rather than promotive voice. Regarding Hypothesis 3a, b, and c, we only found significant main effects of PSB and OSE, but no significant direct interaction effect between PSB and OSE on promotive and prohibitive voice. Thus Hypothesis 3a, b, and c was not supported. As there was no mediating effect of psychological safety on promotive voice, a conditional indirect effect could not be tested (Hypothesis 4a). However, Hypothesis 4b and c was supported as we found that the effect of PSB was contingent on OSE, such that the mediating effect of psychological safety on prohibitive voice became stronger for individuals high in OSE, but insignificant for individuals low in OSE.

Theoretical Implications The findings of our study contribute to the literature in several ways. Firstly, we did not find any direct interaction effect of OSE on either promotive or prohibitive voice, although we did find a significant main effect of OSE on promotive voice. The latter suggests that the perception of oneself as having something qualified to say, and the proactive tendency which results from high OSE beliefs, may in itself be enough to increase the employees’ promotive voice but not prohibitive voice, where self-protective motives is found to be more important (Morrison, 2011; Liang et al., 2012).

Secondly, we did not find any mediational effect of psychological safety on promotive voice, something which contradicts the results of Detert and Burris (2007), who find that psychological safety fully mediates the effect of leadership openness on improvement-oriented voice. One explanation for these divergent results however, may be that our sample consisted of highly educated white-collar workers (the mean employee had a bachelor degree), compared to the sample of Detert and Burris (2007) which to a large extent consisted of production workers in a restaurant chain, where the education level may be lower. Thus, the high educational level in our sample, combined with the Norwegian culture characterized by high participation and influence sharing, might have reduced the effect of psychological safety on promotive voice.

On the other hand, we found that for prohibitive voice, PSB works fully through the effect of psychological safety.

Furthermore, we elaborate this understanding by showing that this indirect effect underlies the employees’ level of OSE. This knowledge is crucial as it suggests that in order for PSB to be effective in predicting prohibitive voice, individuals must have a certain level of belief in their own competences at work. Thus, our results shed light on the boundary conditions under which PSB operates. This is Journal of Personnel Psychology (2016), 15(1), 25–34

M. Svendsen et al., Participative Supervisory Behavior and Voice

especially prominent for prohibitive voice as the effect of PSB was fully mediated by psychological safety, and this mediational effect disappeared for individuals low in OSE. Consequently, individuals with a low level of OSE will not speak up with their prohibitive concerns, regardless of their supervisors’ PSB. Hence, our study underlines Ashford et al.’s (2009) recommendation to execute research which takes into account how mediators and moderators operate simultaneously, as this may have important consequences for our understanding of the boundary conditions of the variables assumed to predict voice behavior.

Lastly, the fact that the effect of PSB on prohibitive voice was fully mediated by psychological safety, compared to promotive voice where there was no mediational effect, suggests that supervisory behaviors may have a different impact on the two types of voice. One possible interpretation is that PSB had no direct effect on prohibitive voice, due to the inherently larger personal risk at stake when it comes to prohibitive voice (Liang et al., 2012). Consequently, having a supervisor who shows that he or she is open to suggestions and gives influence will yield little effect if not accompanied by a heightened sense of psychological safety. On the contrary, promotive voice is associated with less interpersonal risk, and thus PSB may in itself be enough to increase the employee’s promotive voice. Thus, our results are consistent with and contribute to the arguments and empirical work of Van Dyne et al. (2003), Morrison (2011), and Liang et al. (2012) who highlight the importance of looking at the different motivational processes that may be involved regarding promotive and prohibitive voice.

Practical Implications The results from the present study show that an organization wanting to have their employees’ input should pay attention to and empower their managers to be approachable and willing to give suggestions or concerns some serious thought. This may increase the employees’ promotive voicing directly, but also prohibitive voice indirectly, by creating higher levels of psychological safety. However, having a participative supervisor may not always be a panacea for increasing the employees’ prohibitive voice behavior, as this effect is also dependent on the employees’ belief in their own competences at work. Therefore, organizations should also pay attention to their employees’ OSE beliefs and find other mechanisms to increase prohibitive voice for individuals who are low in OSE.

Limitations and Future Research The current study has some limitations worth noting. First, the design of the study is cross-sectional, which precludes Ó 2016 Hogrefe Publishing

M. Svendsen et al., Participative Supervisory Behavior and Voice

drawing causal interferences. Second, common method bias may have influenced the results because all measures are derived from the employees (Podsakoff et al., 2003).

However, Spector (2006) among others argue that common method bias may generally be an overestimated problem. Furthermore, Podsakoff, MacKenzie, and Podsakoff (2012) show that common method bias only deflates, but cannot inflate interaction effects something which suggests that the conditional indirect effect should be robust. Moreover, the divergent statistical results between promotive and prohibitive voice are not likely to be attributed to common method bias (Podsakoff et al., 2003). Additionally, recent meta-analytical results point to the fact that reports from supervisors or peers regarding proactivity constructs such as voice behavior may be distorted by egocentric bias and observational bias (Tornau & Frese, 2013). Future research should address these limitations by conducting longitudinal research based on combined responses from different sources or diary studies accounting for actual voice.

Third, as previously mentioned our study was conducted on white-collar workers in a Norwegian context. Thus, one must be careful to generalize the results to other cultural contexts and other types of workers (i.e., blue-collar workers.). As the cultural context of participation and influence sharing in itself is conveying the appropriateness voice, this may have lowered the effect of both PSB and psychological safety. Future studies, should therefore look at the influence of culture for the effect of PSB. A final limitation is how we have measured and analyzed our hypotheses at an individual level. Although based on theory as well as prior research, PSB may also exist as a group-level construct (Tangirala & Ramanujam, 2012).

Thus, PSB may have different effects when examined at group versus individual level. Hence, future research should look at the different effects of PSB as an individual and group level construct in order to sort out the potential differences between these two levels.

References Ashford, H., Sutcliffe, K. E., & Christianson, M. K. (2009). Speaking up and speaking out: The leadership dynamics of voice in organizations. In J. Greenberg & M. S. Edwards (Eds.), Voice and silence in organizations (pp. 175–202). Bingley, UK: Emerald.

Bandura, A. (1997). Self-efficacy: The exercise of control.

New York, NY: Freeman.

Bliese, P. D., & Halverson, R. R. (1998). Group size and measures of group-level properties: An examination of eta-squared and ICC values. Journal of Management, 24, 157–172.

Brislin, R. W. (1986). The wording and translation of research instruments. In W. J. Lonner & J. W. Berry (Eds.), Field methods in cross-cultural research (pp. 137–164). Thousand Oaks, CA: Sage.

Ó 2016 Hogrefe Publishing

33

Burris, E. R. (2012). The risks and rewards of speaking up: Responses to voice in organizations. Academy of Management Journal, 55, 851–875.

Conger, J., & Kanungo, R. (1988). The empowerment process: Integrating theory and practice. Academy of Management Review, 13, 471–482.

Detert, R., & Burris, R. (2007). Leadership behavior and employee voice: Is the door really open? Academy of Management Journal, 40, 869–884.

Dutton, J., Ashford, S., Lawrence, K., & Miner-Rubino, K. (2002).

Red light, green light: Making sense of the organizational context for issue selling. Organization Science, 13, 355–369.

Edmondson, A. C. (1999). Psychological safety and learning behaviors in work teams. Administrative Science Quarterly, 44, 350–383.

Frese, M., & Fay, D. (2001). Personal initiative: An active performance concept for work in the 21st century. Research in Organizational Behavior, 23, 133–188.

Graen, G. B., & Scandura, T. A. (1987). Toward a psychology of dyadic organizing. In L. L. Cummings & B. M. Staw (Eds.), Research in organizational behavior (pp. 175–208). Greenwich, CT: JAI Press.

Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis. New York, NY: Guilford.

Janssen, O., de Vries, T., & Cozijnsen, A. J. (1998). Voicing by adapting and innovating employees: An empirical study on how environment and personality interact to affect voice behavior.

Human Relations, 51, 945–967.

Kassing, J. W. (2001). From the look of things: Assessing perceptions of organizational dissenters. Management Communication Quarterly, 14, 442–470.

Le Pine, J. A., & Van Dyne, L. (1998). Predicting voice behavior in work groups. Journal of Applied Psychology, 83, 853–868.

Liang, J., Farh, I. C., & Farh, J. L. (2012). Psychological antecedents of promotive and prohibitive voice: A two-wave examination. Academy of Management Journal, 55, 71–92.

Liu, W., Zhu, R., & Yang, Y. (2010). I warn you because I like you: Voice behavior, employee identifications, and transformational leadership. The Leadership Quarterly, 21, 189–202.

Milliken, F. J., Morrison, E. W., & Hewlin, P. F. (2003). An exploratory study of employee silence: Issues that employees don’t communicate upward and why. Journal of Management Studies, 40, 1453–1476.

Morrison, E. W. (2011). Employee voice behavior: Integration and directions for future research. The Academy of Management Annals, 5, 373–412.

Podsakoff, P. M., MacKenzie, S. B., & Podsakoff, N. P. (2012).

Sources of method bias in social science research and recommendations on how to control it. Annual Review of Psychology, 65, 539–569.

Podsakoff, P. M., MacKenzie, S. B., Scott, B., Lee, J. Y., & Podsakoff, N. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879–903.

Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing moderated mediation hypotheses: Theory, methods and prescriptions. Multivariate Behavioral Research, 42, 185–227.

Rank, J. (2006). Leadership predictors of proactive organizational behavior: Facilitating personal initiative, voice behavior, and exceptional service performance. Graduate School Theses and Dissertations. Retrieved from http://scholarcommons.usf.edu/ etd/2669 Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical Linear Models (2nd ed.). Thousand Oaks, CA: Sage.

Satorra, S., & Bentler, P. M. (1999). A scaled difference chi-square test statistic for moment structure analysis Unpublished Technical Report, University Pompeu Fabra, Barcelona, Spain.

Journal of Personnel Psychology (2016), 15(1), 25–34

34

Schyns, B., & Von Collani, G. (2002). A new occupational selfefficacy scale and its relation to personality constructs and organizational variables. European Journal of Work and Organizational Psychology, 11, 219–241.

Spector, P. E. (2006). Method variance in organizational research: Truth or urban legend? Organizational Research Methods, 9, 221–232.

Spreitzer, G. M. (1995). Psychological empowerment in the workplace: Dimensions, measurement and validation. Academy of Management Journal, 38, 1442–1465.

Tangirala, S., & Ramanujam, R. (2012). Ask and you shall hear (but not always): Examining the relationship between manager consultation and employee voice. Personnel Psychology, 65, 251–282.

Tornau, K., & Frese, M. (2013). Construct clean-up in proactivity research: A meta-analysis on the nomological net of workrelated proactivity concepts and their incremental validities.

Applied Psychology, 62, 44–96.

Van Dyne, L., Ang, S., & Botero, I. C. (2003). Conceptualizing employee silence and employee voice as multidimensional constructs. Journal of Management Studies, 40, 1359–1392.

Journal of Personnel Psychology (2016), 15(1), 25–34

M. Svendsen et al., Participative Supervisory Behavior and Voice

Vroom, V. H., & Jago, A. G. (1988). The new leadership: Managing participation in organizations. Englewood Cliffs, NJ: PrenticeHall.

Received June 10, 2014 Revision received December 11, 2014 Accepted December 29, 2014 Published online April 12, 2016 Mari Svendsen Department of Psychology and Behavioral Sciences Aarhus University Bartholins Alle 9 8000 Aarhus C Denmark Tel. +47 46449109 E-mail mari@psy.au.dk

Ó 2016 Hogrefe Publishing

Original Article

Spaces That Signal Identity Improve Workplace Productivity Katharine H. Greenaway,1 Hannibal A. Thai,1 S. Alexander Haslam,1,2 and Sean C. Murphy1 1

School of Psychology, University of Queensland, Brisbane, Australia

2

Canadian Institute for Advanced Research, Toronto, ON, Canada

Abstract: The physical spaces we inhabit have a profound impact on psychological functioning. People generally experience positive outcomes in spaces that support important identities and negative outcomes in spaces that threaten those identities. We investigated the effects of working in an ingroup or outgroup space on organizational performance. Participants completed exercises in a simulated work environment as a member of a research education development (RED) work team. The office space was designed to be identity affirming (decorated by a RED team), identity threatening (decorated by a rival business legacy usability and engineering [BLUE] team), or undecorated. Work teams performed better in both ingroup spaces and outgroup spaces than in undecorated spaces. The findings highlight the importance of considering the impact of physical space on psychological functioning in the workplace and beyond.

Keywords: social identity, space, work, productivity, communication

“To be human is to live in a world that is filled with significant places: To be human is to have and know your place.”

Relph (1976, p. 1) As Edward Relph (1976) noted in his seminal book Place and Placelessness, the physical spaces people inhabit have a profound impact on their psychological experience. Yet in psychological research, the physical space people inhabit is typically neglected as a factor that might influence the way they think and act (Hopkins & Dixon, 2006). Where research has considered this question, it has typically focused on the way in which the physical properties or layout of a space affect psychological outcomes (e.g., Gosling, Ko, Mannarelli, & Morris, 2002; Schaller, Park, & Mueller, 2003). Yet, as Relph (1976) notes, it is not simply their physical elements that make spaces psychologically impactful; often space is meaningful as a result of the way it supports, communicates, and channels group identity.

Research shows that the identity-related meaning of spaces has an impact on well-being and performance. For instance, sports teams win more games at home than away (Allen & Jones, 2014), Christians are happier in a room with a Christmas tree than one without (Schmitt, Davies, Hung, & Wright, 2010), and show better self-reported and physiological health when immersed in a cathedral rather than a Mosque (Ysseldyk, Haslam, & Morton, 2014). Yet, just as spaces can affirm the identities of some people, they Ó 2016 Hogrefe Publishing

can also challenge and threaten the identities of others.

For instance, Sikhs and Buddhists are less happy in a room with a Christmas tree than one without (Schmitt et al., 2010); atheists show worse health when immersed in a cathedral rather than a museum (Ysseldyk et al., 2014); women report less interest in computer science in a room containing objects considered stereotypical (a Star Trek poster) rather than non-stereotypical of that field (a nature poster; Cheryan, Plaut, Davies, & Steele, 2009). Extending this work, the present research investigates the effects of feeling “at home” compared to “out of place” in one’s physical space in an organizational context – considering the consequences of this experience for individual and team performance.

Space in the Workplace One popular approach to organizational space management emerges from the work of Taylor (1911), who advocated tight managerial control over workspace. In the latter part of the 20th century, this functional approach to workspace design spawned the “lean” philosophy of space management, which minimizes opportunity for personalization in the office (Harris & Harris, 2006). The lean approach is designed to eliminate waste and improve efficiency (Skinner, 2005), ensuring employees focus on work tasks without distractions in the form of personal belongings or decorations. Indeed, lean approaches involve stripping a workspace and workflow back to the minimum materials Journal of Personnel Psychology (2016), 15(1), 35–43 DOI: 10.1027/1866-5888/a000148

36

required to perform the job functions at hand (Durmusoglu & Kulak, 2008). Although in the present research we focus largely on the identity-related implications of workspace management, this approach thus denies employees both instrumental (i.e., function-related) and symbolic (i.e., identity-related) control over their workspace.

Yet despite having considerable appeal in managerial circles, much of the evidence that is invoked to support a lean approach to space management is anecdotal. Moreover, the approach is challenged by recent experimental research that has found “leaning” one’s workspace can have detrimental effects (e.g., Nieuwenhuis, Knight, Postmes, & Haslam, 2014). In particular, Knight and Haslam (2010a) conducted two experiments in which participants performed a productivity task in one of four conditions: (1) a lean office with minimal decoration, (2) an enriched office decorated by the experimenter, (3) an empowered office decorated by the participant, or (4) a disempowered office decorated by the participant and then redecorated by the experimenter. In both studies the enriched office increased productivity relative to a lean office. Interestingly, productivity was even higher in the empowered office when participants were responsible for decoration. The authors argued that this was because an empowered office allowed participants the opportunity to express and realize their own identity and therefore enabled them to imbue the space with personal meaning.

As Knight and colleagues indicate, one theoretical framework that can be used to make sense of such findings is the social identity approach (e.g., Haslam, 2004, after Tajfel & Turner, 1979; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). This asserts that individuals internalize group memberships as part of their self-concept in a way that fundamentally shapes their perceptions, thoughts, and behavior.

As a result, in a range of social contexts, people view the world through a social lens – interpreting their own and others’ actions with reference to social identities that they do and do not share. More specifically, the social identity approach to organizational life points to the importance of recognizing social identities in the workplace, and to the way in which the interaction of multiple identities serves to structure individual and group behavior (Haslam, 2004, 2014). Among other things, this means that when people feel their workplace identities are respected by employers and valued within an organization, they are typically more willing to work toward organizational goals because here the organization’s interests are more likely to be seen to be aligned with “our interests” (or even “my interests”). As a result, being embedded in a supportive team environment generally provides a motivational basis for improved performance (Haslam, Postmes, & Ellemers, 2003).

Journal of Personnel Psychology (2016), 15(1), 35–43

K. H. Greenaway et al., Identity and Productivity

Identity in the Workplace In organizational contexts, identification with an organization, department, or team tends to promote positive organizational outcomes in the form of job satisfaction, well-being, and productivity (Knight & Haslam, 2010b; Millward, Haslam, & Postmes, 2007; Van Dick, Grojean, Christ, & Wieseke, 2006). As a result, workspaces that provide opportunities for the realization of important work identities are typically found to promote positive organizational outcomes – such as enhanced performance – relative to lean workspaces that do not afford their occupants the same opportunities (Knight & Haslam, 2010a; Nieuwenhuis et al., 2014). In the present research we extend this work by examining whether these same benefits can be gained vicariously through shared (and non-shared) group memberships.

In this regard, a core tenet of the social identity approach is that people are motivated to achieve or maintain positive social identities. One consequence of this is that they often engage in intergroup competition that involves striving to compare ingroups (e.g., the work team to which they belong) favorably with outgroups (e.g., a rival work team).

Among other things, this can mean that they are highly attuned to cues of outgroup presence and may find themselves distracted or intimidated by such cues. For example, Brown and Baer (2011) found that people who negotiated in another person’s office performed worse than they did if they negotiated in their own office or a neutral room. Alternatively, though, being in an “outgroup space” may also galvanize individuals toward better performance. That is, people can engage in social competition to try to assert the superiority of the ingroup when confronted with cues of outgroup presence.

In parallel to the social identity literature, such patterns are also documented in the literature on workplace territoriality, which has studied the effects of feelings of psychological ownership associated with organizational spaces (e.g., Brown, Lawrence, & Robinson, 2005). Here, a sense of territoriality has sometimes been observed to confer organizational advantages (Wells, 2000). For example, workspace territoriality can motivate individuals to aggressively defend their space, which can manifest itself in the form of greater motivation on work tasks (Brown, 2009).

Again, though, territoriality can also have detrimental effects. For example, workspace territoriality can preoccupy employees, leaving them with less time and energy to work on other tasks (Hall & Richter, 1989). Indeed, employees can become disengaged and discomforted when other people (e.g., managers) encroach upon or attempt to control their workspace (Haslam, 2004; Knight & Haslam, 2010b). As a result, workspaces that threaten important

Ó 2016 Hogrefe Publishing

K. H. Greenaway et al., Identity and Productivity

identities – for example, by reflecting the identity of rival work teams – may lead to negative organizational outcomes.

The Present Research The goal of the present study was to investigate the organizational effects of being in a space that contained identity cues relative to a lean space in which identity markers were absent. For this purpose, participants formed work teams and completed exercises in a space that appeared to be decorated by an ingroup, an outgroup, or was undecorated. We measured organizational performance in terms of effective group communication and productivity. While these represent distinct forms of organizational performance, effective group communication is nevertheless necessary for groups to be productive. Given that initial differences in group communication ability may influence later measures of communication and productivity, we controlled for baseline group communication when assessing the effects of the space manipulation on later outcome variables. In addition, productivity was assessed both at the individual level and at the group level. This allows us to establish whether teams function better together, or individuals put more effort into team productivity, or both, when in different types of workspaces.

Whereas prior work has investigated the effects of an office space manipulation at intrapersonal (Knight & Haslam, 2010a) or intragroup levels (Nieuwenhuis et al., 2014), in the present research we focus on an intergroup manipulation that explores the consequences of being in “ingroup territory” compared to “outgroup territory” for individual and group performance. We investigate this research question in the social identity tradition of minimal groups, in which participants are assigned to ad hoc groups (e.g., red vs. blue group). These groups have no prior meaning for participants, and thus allow researchers to establish the minimal conditions necessary to produce relevant forms of group behavior (e.g., discrimination; Tajfel, Flament, Billig, & Bundy, 1971).

On the basis of social identity theorizing we hypothesized that ingroup space would increase performance (operationalized as communication and productivity) relative to lean space (H1). In contrast, we countenanced competing hypotheses about being in an outgroup space, predicting that this might either decrease performance relative to lean space (H2a) or increase performance relative to lean space (H2b). Thus, we expected the social meaning of the space to impact organizational performance – positively, when the space affirmed team identity and indeterminately when the space threatened team identity.

We also investigated potential identity-related mechanisms through which physical space might influence performance. In line with the social identity approach, Ó 2016 Hogrefe Publishing

37

we anticipated that participants would feel more identified with ingroup team members when in an ingroup space.

However, we anticipated that participants would feel more competitive toward outgroup team members when in an outgroup space. We also tested exploratory hypotheses that increased identification would mediate the effect of ingroup space on performance and that increased competition would mediate the effect of outgroup space on performance.

Method Participants and Design Fifty-four university students (45 female, Mage = 21.05, SD = 1.80) participated in the experiment for $AU20. The experiment had a between-subjects design in which participants occupied an ingroup space (n = 18), an outgroup space (n = 16), or a lean space (n = 20). Testing sessions were conducted in groups of three to four participants. This resulted in 15 teams, with 5 teams per condition.

Manipulations and Measures Work Team Creation Participants imagined that they worked for a fictional organization with two work teams: The Research Education Development (RED) team and the Business Legacy Usability and Engineering (BLUE) team. All participants were assigned to the RED team and were provided with unique staff numbers and job descriptions. Team members completed a series of get-to-know-you exercises in which they introduced themselves and their job role, answered personal questions in a “fast friends” paradigm (Page-Gould, Mendoza-Denton, & Tropp, 2008), created a team poster, and decorated the room in their team color (red). To enhance motivation, participants were informed that the total productivity scores of all RED teams would be compared with the total productivity scores of all BLUE teams.

Participants were provided with iPads to complete the rest of the procedure.

Baseline Team Communication Participants completed a measure of team communication before the space manipulation to control for baseline differences in group communication ability. Participants played SPACETEAMÓ, an iPad game in which each individual has a unique control panel they must operate collectively to fly a spaceship. During each round, individuals receive simultaneous on-screen instructions to operate switches and buttons on their own and other teammates’ panels.

The team must therefore verbally communicate Journal of Personnel Psychology (2016), 15(1), 35–43

38

instructions to other team members, and comprehend and implement the instructions of others under time pressure in order to progress past a game level. Team communication was measured as the number of levels cleared in the game within a five-min period.

Office Space Manipulation The experimenter exited the room during the SPACETEAMÓ task. On returning, they informed participants that the room was double-booked and that they therefore needed to relocate to an adjacent room. Participant groups were randomly assigned to continue the experiment in one of three conditions in which the room was designed to appear as though another work team had previously occupied the space. The experimenter explained that another team had used the room but there was no time to redecorate before beginning the next task. In all conditions, the space contained two desks and four chairs in the middle of the room. The ingroup space (coded as 1) was decorated as though a previous RED team had occupied the space.

The outgroup space (coded as 1) was decorated as though a previous BLUE team had occupied the space. The lean space (coded as 0) was undecorated.

Team Communication and Productivity After relocating rooms, participants completed another round of the SPACETEAMÓ communication task. They then completed Crown’s (2007) letter-word-sentence task, which has been used as a measure of productivity in previous research (Ronay, Greenaway, Anicich, & Galinsky, 2012). Each team member received a unique matrix of 16 letters and was instructed to use the matrix to create words of more than three letters. Teams then worked to combine their individual words into sentences. Participants were given 10 min to complete the task.

Individual productivity was measured as the total number of words produced by each team member (M = 5.62, SD = 2.82). Team productivity was measured as the total number of words in sentences created by groups divided by the number of participants per group (M = 6.74, SD = 2.95). Individual contribution to the team was measured as the total number of words each individual contributed to team sentences divided by the number of participants per group (M = 1.97, SD = 1.54).

Self-Reported Attitudes Four items created by the authors so as to have face validity in the experimental context measured identification with the room’s previous occupants (“I identify with the group that was in this room before us”; “The group that was in this room before us was part of the RED team”; “I could work with the group that was in this room before us”; and “The group that was in this room before us had our Journal of Personnel Psychology (2016), 15(1), 35–43

K. H. Greenaway et al., Identity and Productivity

best interests at heart”), α = .67. The authors also used face validity as a basis for creating four items that measured competitiveness with the room’s previous occupants (“I feel competitive toward the group that was in this room before us”; “The group that was in this room before us represents a threat”; “I want to win against the group that was in this room before us”; and “I dislike the group that was in this room before us”), α = .78. All items were scored on a scale ranging from 1, strongly disagree to 7, strongly agree.

Results Team Outcomes Participants scored as a team on the measures of group communication and group productivity, and thus only one outcome is available for analysis per team. We therefore report the results below collapsing the group-level communication and productivity measures across teams. It is important to note that, given the small number of teams, an analysis of these outcomes is very underpowered to detect our effects of interest. We therefore include these results primarily to note that the pattern of means mirrors that of the individual-level outcomes, described below.

Baseline Team Communication A one-way ANOVA revealed that baseline communication (prior to the manipulation) differed marginally as a function of the space manipulation, F(2, 12) = 3.56, p = .061, ηp2 = .372. Participants in the ingroup space showed poorer initial team communication (M = 0.68, SD = 0.27) than participants in the outgroup space (M = 1.60, SD = 0.86), p = .029, and marginally poorer communication than participants in the lean space (M = 1.45, SD = 0.45), p = .060. There was no significant difference between the latter two conditions, p = .691. To control for these coincidental preexisting differences, and to account for the fact that better initial team communication is likely to affect later team outcomes, baseline communication was included as a covariate for analyses on post-manipulation communication and productivity variables.

Post-Manipulation Team Communication A one-way ANCOVA (controlling for baseline communication) revealed no significant effect of the space manipulation on post-manipulation communication, F(2, 11) = 1.56, p = .254, ηp2 = .201.

Team Productivity A one-way ANCOVA (controlling for baseline communication) revealed a marginally significant effect of the space manipulation, F(2, 11) = 3.66, p = .060, ηp2 = .313. Ó 2016 Hogrefe Publishing

K. H. Greenaway et al., Identity and Productivity

To account for the nesting of individuals within groups, individual-level outcome measures were tested with mixed-effects models using the “lme4” package in R (Bates, Maechler, Bolker, & Walker, 2014). These are multilevel models that account for the nonindependence of observations that comes with nested data. The ANOVAs and ANCOVAs reported below are analyzed similarly to their non-multilevel counterparts, with the exception that the variance in observations attributable to group membership is first parceled out, yielding results that control for the nesting of observations within these groups. p-values and degrees of freedom for these models are based on the Satterthwaite approximation, which can result in decimal places for the degrees of freedom (Schaalje, McBride, & Fellingham, 2002).

Individual Productivity A one-way mixed-effects ANCOVA (controlling for baseline communication) revealed no significant effect of the office space manipulation, F(2, 11.71) = 0.12, p = .885.

Individual Contribution to Team Productivity A one-way mixed-effects ANCOVA (controlling for baseline communication) revealed a significant effect of the office space manipulation, F(2, 9.88) = 7.05, p = .013. Supporting H1, participants in the ingroup space contributed more words to team productivity (M = 2.43, SD = 1.40) than participants in the lean space (M = 1.05, SD = 0.80), p = .032.

Supporting H2b, participants in the outgroup space also contributed more words (M = 2.72, SD = 1.98) than participants in the lean space, p = .006. There was no difference in contributions between ingroup and outgroup conditions, p = .650; see Figure 1.

Team Identification A one-way mixed-effects ANOVA revealed a marginally significant effect of the office space manipulation, F(2, 52) = 3.11, p = .053. As expected, participants in the ingroup space felt more identified with previous room occupants (M = 3.97, SD = 1.21) than participants in the lean space (M = 2.78, SD = 1.36), p = .020, while participants in the outgroup space did not (M = 2.89, SD = 1.15), p = .230. However, participants in the ingroup space did Ó 2016 Hogrefe Publishing

Individual contribution to group productivity

Individual Outcomes

3.5 3 2.5 2 1.5 1 0.5 0 Outgroup space

Lean space

Ingroup space

Figure 1. Number of words contributed by individual participants to group sentences as a function of the office space manipulation.

7

Attitudes toward previous room occupants

Supporting H1, participants in the ingroup space showed greater team productivity (M = 8.21, SD = 3.34) than participants in the lean space (M = 4.70, SD = 2.35), p = .050. Supporting H2b, participants in the outgroup space also showed greater team productivity (M = 8.65, SD = 2.59) than those in the lean space, p = .045. There was no difference between the ingroup and outgroup conditions, p = .816.

39

6

Identification Competition

5 4 3 2 1 Outgroup space

Lean space

Ingroup space

Figure 2. Identification and competition with previous room occupants as a function of the office space manipulation.

not feel more identified than participants in the outgroup space, p = .260, see Figure 2.

Team Competition A one-way mixed-effects ANOVA revealed a significant effect of the office space manipulation, F(2, 51) = 5.67, p = .006. As expected, participants in the outgroup space felt more competitive toward the room’s previous occupants (M = 4.83, SD = 1.49) than participants in the lean space (M = 3.62, SD = 1.62), p = .007, and the ingroup space (M = 2.96, SD = 1.10), p = .003. There was no difference in perceived competition between the ingroup and lean space conditions, p = .662, see Figure 2.

Testing for Mediation We used the “mediation” package in R (Tingley, Yamamoto, Hirose, Keele, & Imai, 2014) to assess whether the office Journal of Personnel Psychology (2016), 15(1), 35–43

40

K. H. Greenaway et al., Identity and Productivity

Table 1. Effects of ingroup space and outgroup space versus lean space on contribution to group productivity mediated through identification Ingroup vs. lean space Estimate

CI Lower

Outgroup vs. lean space CI Upper

Estimate

CI Lower

CI Upper

Indirect Effect (IE)

0.31

.05

0.88

0.11

.14

0.46

Direct Effect (DE)

1.10

.03

2.20

1.61

.65

2.54

Total Effect (TE)

1.41

.05

0.95

1.71

.73

2.68

space manipulation had indirect effects on the individual productivity variables via team identification or team competition. To account for the nesting of observations, we conducted multilevel mediation. This involved using a standard difference-in-coefficients approach (Zhang, Zyphur, & Preacher, 2009) to compare the effect of the independent variable (IV) on the dependent variable (DV) with and without the mediator in the regression equation, and with the addition that both regression equations were estimated with multilevel models that controlled for the variance attributable to group membership. We created dummy-coded variables that contrasted the ingroup space against the lean space in one mediation, and then in a separate mediation contrasted the outgroup space against the lean space to examine possible differences in causal mechanisms behind their effect on individual productivity. For all analyses we entered baseline communication as a covariate and used bootstrapping with 10,000 resamples to generate confidence intervals.

There were no significant indirect effects via team competition (i.e., all confidence intervals spanned zero). The indirect effects via team identification on individual productivity were also nonsignificant. However, as seen in Table 1, for individual contribution to group productivity, there was a marginally significant indirect effect of the ingroup space via team identification (IE = 0.31, 95% CI = 0.050 to 0.877), while there was no evidence of mediation through identification for the outgroup space (IE = 0.11, 95% CI = 0.140 to 0.462). These analyses suggest that increased productivity in the ingroup space may be partially due to increased team identification, while the same cannot be said for increased productivity in the outgroup space.

However, the relative size of the two indirect effects was not significantly different (IE = .20, 95% CI = .311 to .819).

Discussion The goal of this experiment was to investigate the impact of identity-enriched space on individual and group productivity. Here space that had ostensibly been decorated by an ingroup was found to promote work-related productivity relative to a lean space that contained no identity-related information. This pattern accords with previous work showing that enriched workspaces enhance organizational Journal of Personnel Psychology (2016), 15(1), 35–43

performance (Nieuwenhuis et al., 2014) – particularly when people are able to live out their identity within the space (Knight & Haslam, 2010a). We also found that working in a space decorated in the color of a rival outgroup led to improved productivity relative to a lean space. This finding is consistent with evidence from previous social identity research which has shown that intergroup rivalry can sometimes lead to improved organizational performance because it serves to increase motivation and engagement (Ellemers, De Guilder, & Haslam, 2004).

It is possible that the same mechanism was responsible for the effect of both identity-enriched spaces on productivity— insofar as the decoration of office space per se had a positive effect on team functioning. This explanation would accord with other research that has observed a positive effect of workspace enrichment on organizational productivity (Knight & Haslam, 2010a; Nieuwenhuis et al., 2014). Nevertheless, findings on self-report measures suggest that other processes may also be at work here. For, as one would expect, participants felt more identified with previous occupants when those occupants ostensibly belonged to the same RED team, but felt more competitive with previous occupants when those occupants belonged to the rival BLUE team. Although we did not find statistical evidence for the differential impact of these processes, this pattern suggests that different mechanisms may underpin the effects of working in ingroup rather than outgroup space.

Limitations As with all researches, the present experiment was not without limitations. First, our sample size was relatively small due to the complicated procedure and necessity of recruiting participants in groups. Second, participants were undergraduate students and not actual employees, which compromise the study’s external validity. We attempted to increase external validity by having participants interact for an extended period at the start of the study, allowing them to become acquainted with one another and their work roles, encouraging them to feel ownership in decorating their workspace, and generally giving them an opportunity to function as a team. It is also notable that prior experimental studies of related phenomena by Knight and Haslam (2010a) have found effects produced in the laboratory to be generalizable to real-world office contexts. Ó 2016 Hogrefe Publishing

K. H. Greenaway et al., Identity and Productivity

Nevertheless, future research would clearly benefit from replicating the present study in a real workplace.

Third, as intimated above, it is unclear which specific mechanism is responsible for the effects of physical space on organizational performance. Although there is suggestive evidence that team identification and competition both play a role, our findings here were inconclusive. Understanding the precise mechanisms through which physical space influences psychological outcomes therefore continues to be an important direction for future research. In this process, it may also be useful to compare the benefits gained through crafting a comfortable physical environment to the benefits gained by other organizational interventions, such as increasing job autonomy and the quality of leadership.

Implications for Social and Organizational Psychology Organizational research that has investigated productivity has tended to focus on what job employees perform and how they perform it as key job features for intervention (e.g., Arthur, Bennett, Edens, & Bell, 2003; Combs, Liu, Hall, & Ketchen, 2006). The present research suggests that the relatively neglected question of where employees work also has an important bearing on organizational outcomes.

In particular, our findings speak to an emerging literature demonstrating the dangers of “leaning” office space (Knight & Haslam, 2010b; Nieuwenhuis et al., 2014) and the benefits of adding enrichment and “meaning” to office space for organizational performance (Knight & Haslam, 2010a). Importantly, they do this by demonstrating (to our knowledge for the first time) that such effects can occur vicariously through shared (or in some cases, unshared) group membership. More specifically, people appear to benefit from inhabiting a space that has been decorated by fellow ingroup members. This could be important in large organizations in which individual personalization of collective spaces is unfeasible. The findings also demonstrate that outgroup decorated space can also improve performance, at least in the short term. Nevertheless, we suspect that over time this boost in motivation may fade, leaving individuals disengaged and discomforted in a space that does not reflect their identity (Cheryan et al., 2009; Schmitt et al., 2010; Ysseldyk et al., 2014). Such an effect would be consistent with work by Brown and Baer (2011) in which it was observed that people performed worse on negotiations when in another person’s workspace.

The present work also serves to identify points of contact between the social identity approach and previous work on territoriality. More specifically, our findings help to explain why people become territorial over space – namely, because

Ó 2016 Hogrefe Publishing

41

this reflects an extension of themselves and their social identities. Although speaking to similar issues, the territoriality and social identity approaches have not previously been combined to understand attachment to space. Nevertheless, it would appear that integrating these theoretical approaches has the power to yield insights into people’s connection to place and space that might fruitfully be applied to domains outside the workplace (e.g., Hopkins & Dixon, 2006).

The use of minimal groups allows us to conclude that group identity-related enrichment of office space improves productivity, because participants were assigned to ad hoc groups (RED vs. BLUE). Nevertheless, it is possible that the effects observed here may be stronger (though less “pure”) in preexisting groups that have more meaning for participants. Certainly, research on territoriality suggests that individuals can become protective of spaces that hold meaning and express important (preexisting) identities (Brown et al., 2005). To address this possibility, future research might therefore consider investigating these effects in groups that do have prior meaning for participants (e.g., where space communicates work or university identities).

Conclusion The present research supports the conclusion that the way organizational space is managed has important implications for work-related performance. In the simplest terms, our findings suggest that when it comes to space management meaning beats leaning – insofar as spaces that are rich in identity-laden information promote better performance.

More broadly, the study underlines the point on the identity-related dimensions of space matter. Elsewhere, researchers routinely neglect the physical environment as an important contextual factor in their studies, assuming that the spaces participants occupy are neutral backdrops against which psychological processes of scientific interest unfold. Yet the present research shows that the mere presence or absence of identity cues in one’s environment can influence psychological functioning in previously unanticipated, but powerful, ways. As a result, we see that space is not just a stage for performance, but a stimulus.

Acknowledgments Preparation of this paper was facilitated by awards to Katharine H. Greenaway and S. Alexander Haslam from the Canadian Institute for Advanced Research: Social Interactions, Identity, and Well-being Program. The authors

Journal of Personnel Psychology (2016), 15(1), 35–43

42

would like to thank Nicholas Williams and Jason Weiss for their assistance in data collection.

References Allen, M. S., & Jones, M. V. (2014). The “home advantage” in athletic competitions. Current Directions in Psychological Science, 23, 48–53.

Arthur, W. Jr, Bennett, W. Jr, Edens, P. S., & Bell, S. T. (2003).

Effectiveness of training in organizations: A meta-analysis of design and evaluation features. Journal of Applied Psychology, 88, 234–245.

Bates, D., Maechler, M., Bolker, B., & Walker, S. (2014). lme4: Linear mixed-effects models using Eigen and S4. R package version 1.1-7. Retrieved from http://CRAN.R-project.org/ package=lme4 Brown, G. (2009). Claiming a corner at work: Measuring employee territoriality in their workspaces. Journal of Environmental Psychology, 29, 44–52.

Brown, G., & Baer, M. (2011). Location in negotiation: Is there a home field advantage? Organizational Behavior and Human Decision Processes, 114, 190–200.

Brown, G., Lawrence, T. B., & Robinson, S. L. (2005). Territoriality in organizations. Academy of Management Review, 30, 577–594.

Cheryan, S., Plaut, V. C., Davies, P. G., & Steele, C. M. (2009).

Ambient belonging: How stereotypical cues impact gender participation in computer science. Journal of Personality and Social Psychology, 97, 1045–1060.

Combs, J., Liu, Y., Hall, A., & Ketchen, D. (2006). How much do high-performance work practices matter? A meta-analysis of their effects on organizational performance. Personnel Psychology, 59, 501–528.

Crown, D. F. (2007). Effects of structurally competitive multilevel goals for an interdependent task. Small Group Research, 38, 265–288.

Durmusoglu, M. B., & Kulak, O. (2008). A methodology for the design of office cells using axiomatic design principles.

Omega-International Journal of Management Science, 36, 633–652.

Ellemers, N., De Gilder, D., & Haslam, S. A. (2004). Motivating individuals and groups at work: A social identity perspective on leadership and group performance. Academy of Management Review, 29, 459–478.

Gosling, S. D., Ko, S. J., Mannarelli, T., & Morris, M. E. (2002). A room with a cue: Personality judgments based on offices and bedrooms. Journal of Personality and Social Psychology, 82, 379–398.

Hall, D. T., & Richter, J. (1989). Balancing work life and home life: What can organizations do to help? Academy of Management Executive, 20, 213–223.

Harris, C., & Harris, R. (2006). Developing a lean workforce: A guide for human resources, plant managers, and lean coordinators.

New York, NY: Productivity Press.

Haslam, S. A. (2004). Psychology in organizations. London, UK: Sage.

Haslam, S. A. (2014). Making good theory practical: Five lessons for an applied social identity approach to challenges of organizational, health, and clinical psychology. British Journal of Social Psychology, 53, 1–20.

Haslam, S. A., Postmes, T., & Ellemers, N. (2003). More than a metaphor: Organizational identity makes organizational life possible. British Journal of Management, 14, 357–369.

Hopkins, N., & Dixon, J. (2006). Space, place, and identity: Issues for political psychology. Political Psychology, 27, 173–185.

Journal of Personnel Psychology (2016), 15(1), 35–43

K. H. Greenaway et al., Identity and Productivity

Knight, C., & Haslam, S. A. (2010a). The relative merits of lean, enriched, and empowered offices: An experimental examination of the impact of workspace management strategies on wellbeing and productivity. Journal of Experimental Psychology: Applied, 16, 158–172.

Knight, C., & Haslam, S. A. (2010b). Your place or mine? Organizational identification and comfort as mediators of relationships between the managerial control of workspace and employees’ satisfaction and well-being. British Journal of Management, 21, 717–735.

Millward, L. J., Haslam, S. A., & Postmes, T. (2007). Putting employees in their place: The impact of hot desking on organizational and team identification. Organization Science, 18, 547–559.

Nieuwenhuis, M., Knight, C., Postmes, T., & Haslam, S. A. (2014).

The relative benefits of green versus lean office space: Three field experiments. Journal of Experimental Psychology: Applied, 20, 199–214. doi: 10.1037/xap0000024 Page-Gould, E., Mendoza-Denton, R., & Tropp, L. R. (2008). With a little help from my cross-group friend: Reducing anxiety in intergroup contexts through cross-group friendship. Journal of Personality and Social Psychology, 95, 1080–1094.

Relph, E. (1976). Place and placelessness. London, UK: Pion.

Ronay, R., Greenaway, K., Anicich, E. M., & Galinsky, A. D. (2012).

The path to glory is paved with hierarchy when hierarchical differentiation increases group effectiveness. Psychological Science, 23, 669–677.

Schaalje, G. B., McBride, J. B., & Fellingham, G. W. (2002).

Adequacy of approximations to distributions of test statistics in complex mixed linear models. Journal of Agricultural, Biological, and Environmental Statistics, 7, 512–524.

Schaller, M., Park, J. H., & Mueller, A. (2003). Fear of the dark: Interactive effects of beliefs about danger and ambient darkness on ethnic stereotypes. Personality and Social Psychology Bulletin, 29, 637–649.

Schmitt, M. T., Davies, K., Hung, M., & Wright, S. C. (2010). Identity moderates the effects of Christmas displays on mood, selfesteem, and inclusion. Journal of Experimental Social Psychology, 46, 1017–1022.

Skinner, C. (2005). First steps to making your office lean. The lean office: Collected practices and cases (pp. 65–69). New York, NY: Productivity Press.

Tajfel, H., Flament, C., Billig, M. G., & Bundy, R. F. (1971). Social categorization and intergroup behaviour. European Journal of Social Psychology, 1, 149–177.

Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33–47). Monterey, CA: Brooks/Cole.

Taylor, F. W. (1911). Principles of scientific management. New York, NY: Harper.

Tingley, D., Yamamoto, T., Hirose, K., Keele, L., & Imai, K. (2014).

Mediation: Causal mediation analysis. R package version 4.4.3.

Retrieved from http://cran.r-project.org/web/packages/mediation/ index.html Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social group: A selfcategorization theory. Oxford, UK: Blackwell.

Van Dick, R., Grojean, M. W., Christ, O., & Wieseke, J. (2006).

Identity and the extra mile: Relationships between organizational identification and organizational citizenship behaviour.

British Journal of Management, 17, 283–301.

Wells, M. M. (2000). Office clutter or meaningful personal displays: The role of office personalization in employee and organizational well-being. Journal of Environmental Psychology, 20, 239–255.

Ó 2016 Hogrefe Publishing

K. H. Greenaway et al., Identity and Productivity

Ysseldyk, R., Haslam, S. A., & Morton, T. A. (2014). Stairway to heaven? (Ir)religious identity moderates the effects of religious spaces on self-esteem, physical health, and cardiac reactivity.

Manuscript in preparation.

Zhang, Z., Zyphur, M. J., & Preacher, K. J. (2009). Testing multilevel mediation using hierarchical linear models problems and solutions. Organizational Research Methods, 12, 695–719.

Received August 18, 2014 Revision received February 16, 2015 Accepted February 17, 2015 Published online April 12, 2016

Ó 2016 Hogrefe Publishing

43

Katharine Greenaway School of Psychology McElwain Building The University of Queensland St Lucia, QLD 4072 Australia Tel. +61 7 3346-9563 Fax +61 7 3365-4466 E-mail k.greenaway@psy.uq.edu.au

Journal of Personnel Psychology (2016), 15(1), 35–43

Instructions to Authors Aims and Scope of Journal of Personnel Psychology: The journal welcomes excellent empirical and theoretical contributions to basic and applied research in personnel psychology and related methodology. Reviews are also welcome, as are replications of previous research. Articles deal with all fields in personnel psychology, such as personnel selection, performance measurement, motivation, leadership, organizational commitment, personnel development and training, new test developments, and job analysis. As many topics in personnel psychology are closely related to issues in other branches of psychology or, more generally, the social sciences and human resource management, the journal is open to contributions of an interdisciplinary nature.

Journal of Personnel Psychology publishes the following types of articles: Original Articles, Review Articles, Research Notes, Registered Reports, and Hybrid Registered Reports.

Manuscript submisson: All manuscripts should be submitted electronically at http://www.editorialmanager.com/jppsy Detailed instructions to authors are provided at http://www. hogrefe.com/periodicals/journal-of-personnel-psychology/ advice-for-authors/ Copyright Agreement: By submitting an article, the author confirms and guarantees on behalf of him-/herself and any coauthors that the manuscript has not been submitted or published elsewhere, and that he or she holds all copyright in and titles to the submitted contribution, including any figures, photographs, line drawings, plans, maps, sketches, and tables, and that the article and its contents do not infringe in any way on the rights of third parties. The author indemnifies and holds harmless the publisher from any third-party claims.

The author agrees, upon acceptance of the article for publication, to transfer to the publisher the exclusive right to reproduce and distribute the article and its contents, both physically and in nonphysical, electronic, or other form, in the journal to which it has been submitted and in other independent publications, with no limitations on the number of copies or on the form or the extent of

distribution. These rights are transferred for the duration of copyright as defined by international law. Furthermore, the author transfers to the publisher the following exclusive rights to the article and its contents: 1. The rights to produce advance copies, reprints, or offprints of the article, in full or in part, to undertake or allow translations into other languages, to distribute other forms or modified versions of the article, and to produce and distribute summaries or abstracts.

2. The rights to microfilm and microfiche editions or similar, to the use of the article and its contents in videotext, teletext, and similar systems, to recordings or reproduction using other media, digital or analog, including electronic, magnetic, and optical media, and in multimedia form, as well as for public broadcasting in radio, television, or other forms of broadcast.

3. The rights to store the article and its content in machinereadable or electronic form on all media (such as computer disks, compact disks, magnetic tape), to store the article and its contents in online databases belonging to the publisher or third parties for viewing or downloading by third parties, and to present or reproduce the article or its contents on visual display screens, monitors, and similar devices, either directly or via data transmission.

4. The rights to reproduce and distribute the article and its contents by all other means, including photomechanical and similar processes (such as photocopying or facsimile), and as part of so-called document delivery services.

5. The right to transfer any or all rights mentioned in this agreement, as well as rights retained by the relevant copyright clearing centers, including royalty rights to third parties.

Online Rights for Journal Articles: Guidelines on authors’ rights to archive electronic versions of their manuscripts online are given in the Advice for Authors on the journal’s web page at www.hogrefe.com

January 2016

The first structured resource for psychologists that combines mindfulness with character strenghts Ryan M. Niemiec

Mindfulness and Character Strengths A Practical Guide to Flourishing

2014, xx + 274 pp. + CD with meditation exercises US $39.80 / € 27.95 ISBN 978-0-88937-376-1 Also available as eBook At the core of this hands-on resource for psychologists and other practitioners, including educators, coaches, and consultants, is MindfulnessBased Strengths Practice (MBSP), the first structured program to combine mindfulness with the character strengths laid out in the VIA Institute’s classification developed by Drs. Martin E. P. Seligman and Christopher Peterson. This 8-session program systematically boosts awareness and application of character strengths – and so helps people flourish and lead more fulfilling lives.

The author’s vast experience working with both mindfulness and character strengths is revealed in his sensitive and clear presentation of the conceptual, practical, and scientific elements of this unique combined approach. It is not only those who are new to mindfulness or to character strengths who will appreciate the detailed primers on these topics in the first

www.hogrefe.com

section of the book. And the deep discussions about the integration of mindfulness and character strengths in the second section will benefit not just intermediate and advanced practitioners. The third section then leads readers step-by-step through each of the 8 MBSP sessions, including details of session structure and content, suggested homework, 30 practical handouts, as well as inspiring quotes and stories and useful practitioner tips. An additional chapter discusses the adaption of MBSP to different settings and populations (e.g., business, education, individuals, couples). The mindfulness and character strengths meditations on the accompanying CD support growth and development.

This highly accessible book, while primarily conceived for psychologists, educators, coaches, and consultants, is suitable for anyone who is interested in living a flourishing life.

How to be more persuasive and successful in negotiations “Presented in a concise and even entertaining style, this book succeeds in demonstrating how to negotiate successfully and fairly at the same time. A clear recommendation.”

Heinz Schuler, PhD, Hohenheim University, Stuttgart, Germany

Marco Behrmann

Negotiation and Persuasion

The Science and Art of Winning Cooperative Partners 2016, viii + 128 pp.

US $34.80 / € 24.95 ISBN 978-0-88937-467-6 Also available as eBook Scientific research shows that the most successful negotiators analyze the situation thoroughly, self-monitor wisely, are keenly aware of interpersonal processes during the negotiation – and, crucially, enter negotiations with a fair and cooperative attitude. This book is a clear and compact guide on how to succeed by means of such goal-oriented negotiation and cooperative persuasion. Readers learn models to understand and describe what takes place during negotiations, while numerous figures, charts, and checklists clearly summarize effective

www.hogrefe.com

strategies for analyzing context, processes, competencies, and the impact of our own behavior. Real-life case examples vividly illustrate the specific measures individuals and teams can take to systematically improve their powers of persuasion and bargaining strength. The book also describes a modern approach to raising negotiation competencies as part of personnel development, making it suitable for use in training courses as well as for anyone who wants to be a more persuasive and successful negotiator.