mm51n895A即2 Squeezed in the Middle:The Middle Status Trade Creativity for Focus wahMichcMyDgwidLouts lackhLGacaa h opposed to the top or bo g different n In addition.we f of status loss a of power and s ng山 unlike statu and by ma ve der resufor future problem solving on tasks that require either focusor. Research n the antecedents and outcomes of social status has Classical research on social influence suggests not,purporting and related disciplines (Fiske,2010:Fiske& that ind y,1956:H 961:Kelley Shapiro.1954 003.M ch h who aspires to a t fears dise but udying status 201 zelditch.19)more aentio mple e over lower status some of the studies wereco relational or gr onflated status and power.For example Bartos (1958)simpl when th status,posi and more he early research theorized that middle ure d (Adler,Epel nsecurity (Dittes 5 tatus acere in a straiehtforward linear fashion? eical pn s that was assumed to be triggered by middle hus. midd rchy or do they face a unique set of pr res at the middle tngger a sense of insecurity and a fea ofrejecefiom estigating how middle status imne his article should he addressed to Michell Busine status loss-and by measuring it directly.Elucidating the under-
Squeezed in the Middle: The Middle Status Trade Creativity for Focus Michelle M. Duguid Washington University in Saint Louis Jack A. Goncalo Cornell University Classical research on social influence suggested that people are the most conforming in the middle of a status hierarchy as opposed to the top or bottom. Yet this promising line of research was abandoned before the psychological mechanism behind middle-status conformity had been identified. Moving beyond the early focus on conformity, we propose that the threat of status loss may make those with middle status more wary of advancing creative solutions in fear that they will be evaluated negatively. Using different manipulations of status and measures of creativity, we found that when being evaluated, middle-status individuals were less creative than either high-status or low-status individuals (Studies 1 and 2). In addition, we found that anxiety at the prospect of status loss also caused individuals with middle status to narrow their focus of attention and to think more convergently (Study 3). We delineate the consequences of power and status both theoretically and empirically by showing that, unlike status, the relationship between power and creativity is positive and linear (Study 4). By both measuring status (Studies 2 and 3) and by manipulating it directly (Study 5), we demonstrate that the threat of status loss explains the consequences of middle status. Finally, we discuss the theoretical implications of our results for future research on status and problem solving on tasks that require either focus or flexibility. Keywords: status, creativity, conformity, convergent thinking Research on the antecedents and outcomes of social status has become one of the most vibrant streams of research in social psychology and related disciplines (Fiske, 2010; Fiske & Berdahl, 2007; Hall, Coats, & LeBeau, 2005; Keltner, Gruenfeld, & Anderson, 2003; Magee & Galinsky, 2008). Status is defined as the prominence, respect, honor, and influence that individuals enjoy in the eyes of others (Anderson, Srivastava, Beer, Spataro, & Chatman, 2006). A reasonable and widely held assumption in the status literature is that most individuals strive to attain status because of the many benefits that accrue as one moves up the status hierarchy (Fan & Gruenfeld, 1998; Podolny, 1993; Ridgeway & Walker, 1995; Sivanathan & Pettit, 2010). Those attaining higher status are given more control over group decisions (Berger, Rosenholtz, & Zelditch, 1980), more attention and influence over lower status group members (Ridgeway & Walker, 1995), more choice over whom to collaborate with (Hardy & Van Vugt, 2006), and more credit when those collaborations result in successful outcomes (Fan & Gruenfeld, 1998; Podolny, 1993). Presumably all of these psychological and material benefits of status should also make individuals more confident and more self-assured (Adler, Epel, Castellazzo, & Ickovics, 2000), but do the benefits of acquiring status accrue in a straightforward, linear fashion? Classical research on social influence suggests not, purporting instead that individuals with middle status are in fact more insecure and more conforming than those with either high or low status (Dittes & Kelley, 1956; Homans, 1961; Kelley & Shapiro, 1954). Homans (1961, p. 357) dubbed this curvilinear effect “middle status conservatism,” which, he argued, “reflects the anxiety experienced by one who aspires to a social station but fears disenfranchisement.” The early findings pointing to middle-status conformity were promising, but the methodological approaches to studying status were imprecise, the results somewhat inconsistent, and the line of research abandoned before the phenomenon was clearly understood (Homans, 1961; Phillips & Zuckerman, 2001). For example, some of the studies were correlational or qualitative (e.g., Blau, 1955), leaving open the possibility that more conforming individuals seek out middle-status positions. Other studies conflated status and power. For example, Bartos (1958) simply used existing leadership positions as proxies for status, positions that may have also included power over others. Moreover, though the early research theorized that middle-status conformity stems from a sense of insecurity (Dittes & Kelley, 1956; Kelley & Shapiro, 1954), there was no direct evidence to corroborate the psychological process that was assumed to be triggered by middle status. Thus, an important question remains— do middle-status individuals gain confidence having escaped the bottom of the hierarchy or do they face a unique set of pressures at the middle that trigger a sense of insecurity and a fear of rejection? In the current research, we attempt to address this question by investigating how middle status impacts creative task performance. We do so with two important objectives in mind. First, we build on the early research on middle-status conformity by specifying the psychological process underlying this phenomenon—the threat of status loss—and by measuring it directly. Elucidating the underMichelle M. Duguid, Department of Organizational Behavior, Olin Business School, Washington University in Saint Louis; Jack A. Goncalo, Department of Organizational Behavior, School of Industrial and Labor Relations, Cornell University. Correspondence concerning this article should be addressed to Michelle M. Duguid, Olin Business School, Washington University in Saint Louis, 1 Brookings Drive, Campus Box 1156, Saint Louis, MO 63130. E-mail: duguid@wustl.edu This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. Journal of Personality and Social Psychology © 2015 American Psychological Association 2015, Vol. 109, No. 9, 000 0022-3514/15/$12.00 http://dx.doi.org/10.1037/a0039569 1
DUGUID AND GONCALO bout the psycho ical and behavioral tatus tha ared with participants in the middle-statu he eis to broad esults o make trated it cleanl suggest that the threat of stat loss may tie together and U-shan Irelationship bet and confe (Ba Middle Status and the Threat of Status Los 958:Blat 955:Dite Kelley Kelley 1954 s of iduals are mor aintain their stat when face e it i in fa and to idlosing fee that they ha cry lit othing left to los ns of an nt of the 105 ern psychological res arch.Indeed,the idea of mid Kelley,1956). ition with onfidence.High n the ical sc g (Ad (En 001).that rk has not yet addre al intera nd feedback,high-s s with mi the The carly yTescarchonstatusandconformiyyieldtedinconsisten why.For exa ple.Kelle and Shapiro atus indiv luals feel thr wh lly r asked to rate,b phery of the high tatu en led to be by the ays th vhich they d just efly feel y iewed them a ork were also th e least likel ing beh d with being female group' do not want to as a mem a group tha ved as highly accepab .200 sum,middle-sta ay be a follo p experiment.Dittes and Kelley (1956)replic solving and task performance ined man ion of status that led some】 Creativity and Status they were highly acceptabl ateeory was missine from the previou deed,one of the attractive features of a status hierarchy is that i erves to reduce destructive forms of conflict and prom ote volu ured their pror ty to conform to the gre Chou Galinskv 2011).n0 mid estatus may be called upon to implement decisions made by
lying psychological mechanism associated with middle status provides a starting point for investigating interesting new questions about the psychological and behavioral consequences of status that were not proposed in the early research, which was focused narrowly on conformity to a group majority. Hence, our second objective is to broaden the focus of existing research to investigate the consequences of middle status for problem solving and task performance. We suggest that the threat of status loss may make those with middle status more wary of advancing creative solutions out of fear that they will be evaluated negatively. Middle Status and the Threat of Status Loss A growing stream of research investigates the consequences of status, but existing research has almost exclusively compared the perceptions, attitudes, and behaviors of those with high and low status. There are in fact many situations in which individuals may find themselves in a middle-status position— knowing that they are not the most respected, influential, and prestigious person in a group, but also that they are more respected, influential, and prestigious than others (Homans, 1961). Those with middle status may be an important but overlooked segment of the social hierarchy in modern psychological research. Indeed, the idea of middlestatus conformity was intriguing but largely forgotten in social psychology by the early 1970s (Phillips & Zuckerman, 2001). And though the concept of middle-status conformity has attracted some attention in the sociological literature (Phillips & Zuckerman, 2001), that work has not yet addressed or identified the underlying psychological process that causes individuals with middle status to more readily conform to a group majority. The early research on status and conformity yielded inconsistent results, and when middle-status conformity did emerge, it was not clear why. For example, Kelley and Shapiro (1954) conducted a seminal experiment in which participants were shown to a room in small groups and asked to introduce themselves to each other. They were asked to rate, based on this brief initial interaction, how acceptable they found each of the other participants as a potential coworker. Each participant was then led to believe that they were viewed by the others as either not at all acceptable (low status) or highly acceptable (high status). Finally, the participants were asked to complete a task that measured their willingness to conform to the group with which they had just briefly interacted. The results showed that the participants who thought their group viewed them as unacceptable coworkers were also the least likely to conform to the group’s opinion. There was much more unexplained variance, however, among participants who believed that they were viewed as highly acceptable coworkers—some readily conformed, whereas others were more likely to resist majority pressure. In a follow-up experiment, Dittes and Kelley (1956) replicated the procedure of the previous experiment but introduced a more fine-grained manipulation of status that led some participants to believe they were highly acceptable to the group, not at all acceptable to the group, or about average. This latter middle-status category was missing from the previous experiment. After each participant clearly understood their relative standing in the group, they were then asked to complete a decision-making task that measured their propensity to conform to the group’s decision—a decision that was clearly incorrect given the available evidence. The results showed that the participants with the highest status were less likely to conform and more likely to dissent from the group judgment compared with participants in the middle-status condition. Unlike the previous experiment, however, conformity in the low-status condition was not as low as conformity in the high-status condition. In other words, taken together, the results of the two studies pointed toward middle-status conformity, though neither study demonstrated it cleanly. We suggest that the threat of status loss may tie together and explain these inconsistent early results pointing to an inverted U-shaped relationship between status and conformity (Bartos, 1958; Blau, 1955; Dittes & Kelley, 1956; Kelley & Shapiro, 1954). The threat of status loss is a particularly potent threat because individuals are more motivated to maintain their status when faced with the prospect of losing it than to gain status in situations in which there is an opportunity to acquire it. Indeed, individuals are willing to pay more money and to exert more effort to avoid losing status than to gain it (Pettit, Yong, & Spataro, 2010). Individuals at the very bottom of the status hierarchy may feel that they have very little or nothing left to lose in terms of esteem that had not already been taken from them. Thus, they are less likely to be threatened by further status loss (Blau, 1955; Dittes & Kelley, 1956). Conversely, being at the top of the social hierarchy may mitigate the threat of status loss by infusing individuals in this position with confidence. High-status actors enjoy greater psychological well-being (Adler et al., 2000), ego satisfaction (Barkow, 1989), and self-esteem (Emerson, 1962). Given this stream of positive social interaction and feedback, high-status individuals are also more likely to trust others (Lount & Pettit, 2012). In contrast to both the low- and high-status individuals, the threat of status loss may be most salient in the middle of the status hierarchy. Research on social mobility lends support to the notion that middle-status individuals feel threatened in situations in which status loss is possible. Socially mobile individuals are those who manage to gain more status than their lower status counterparts but still remain on the periphery of the high-status group (Chattopadhyay, Tluchowska, & George, 2004). These individuals strive to maintain their middle-status position or increase their status, and are very cautious not to act in ways that might interfere with this goal (Davis & Watson, 1982; Duguid, 2011; Ibarra, 1995; Phinney, 1990). For example, there is evidence that women who make it into the top tiers of their organizations but still feel like they are perceived as peripheral or marginal group members make a point of emphasizing behaviors not associated with being female, because they do not want to be classified as a member of a group that generally has lower status in most organizational contexts (Ellemers, 2001; Ely, 1994). In sum, middle-status individuals may be particularly threatened by the prospect of losing status. This concern with status loss may have significant implications for problem solving and task performance. Creativity and Status Middle-status conformity is not necessarily dysfunctional—indeed, one of the attractive features of a status hierarchy is that it serves to reduce destructive forms of conflict and promote voluntary cooperation and coordination between group members (Halevy, Chou & Galinsky, 2011). In organizations, individuals with middle status may be called upon to implement decisions made by This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 2 DUGUID AND GONCALO
STATUS AND CREATIVITY ing middle 2001) of the eva 5 ond, thus disting able from others (Clydesdale n.&a r could in the nrove their statu a.2006 readily conform to convention,then they may be at a sig diverge too far from existing solutions,they could be crit nand creative I to a lowe and abo than the opportunity for status gain (Pettit et al e in vel.they often run counter to existi articularly compared with moving from an valuation sohtiosmyakoceneomtecrnmorebteresianeiomewah nd fa more practi cal solutions (Mueller. 01 an who express The Present Research nform to e we hy e that the threat of status loss should mak the他mid Is at t if individua are asked part to g deas will be riticized (Diehl 18 to the threat of status loss when ther was the mposed of disp would therefore generate the least crative id individuals (Ca whe ility of this ffe re dividual with middle rospect that their ideas will be ev d negatively by other 05 e underlying psychological pro priate given that ide pert rman liver from ma ntaining their or ma ring threat of (M te ins is that the arm glow of soci cceptan e that high tus rovide atus is unigue to s oppos edo than d fo reative idea may be particularly salient to middl ed by tance of others (oun Chigh middle and low and xpected evaluation Magee&Galinsky.2008). high-status nay be ally ly,they will not d b ted that individuals with middle status ould c ideas Compa with individuals with midd status further down the hierarchy hierarchy.and therefore not likely to make a significant move search on middle-status conformity to the realm of problem solv
those higher up—thus, a rebellious and nonconforming middle manager would not necessarily serve the needs of the group (Huy, 2001). Yet particularly when the majority is in error, it is sometimes important for individuals to defy convention to pursue novel ideas that depart from the status quo (Galinsky, Magee, Gruenfeld, Whitson, & Liljenquist, 2008; Goncalo & Duguid, 2012; Goncalo & Staw, 2006). If the threat of status loss causes those with middle status to more readily conform to convention, then they may be at a significant disadvantage in situations that demand creative solutions as opposed to those that demand obedience or cooperation. The distinguishing characteristic of a creative idea over and above ideas that are merely practical is that creative ideas diverge in a novel direction from what is known (Amabile, 1983). Because creative solutions are novel, they often run counter to existing knowledge (Ward, 1994) and are thus likely to be controversial, at least initially (Mueller, Melwani, & Goncalo, 2012). Creative solutions may also encounter more subtle resistance from evaluators who may explicitly state that they desire creativity, but in the end favor more practical solutions (Mueller, Goncalo, & Kamdar, 2011). In other words, individuals who express a creative idea must be willing to risk criticism and resist substantial pressure to conform to existing solutions (Förster, Friedman, Butterbach, & Sassenberg, 2005; Kim, Vincent, & Goncalo, 2013; Nemeth & Staw, 1989). The potential for negative evaluation impedes creativity because even if individuals are able to generate creative solutions, they may be reluctant to share them if they fear that their ideas will be criticized (Diehl & Stroebe, 1987). For example, groups composed of dispositionally anxious individuals generated fewer ideas when brainstorming compared with groups composed of less anxious individuals (Camacho & Paulus, 1995). Moreover, this difference was not apparent when individuals brainstormed alone, suggesting that the fear of being evaluated was what caused anxious individuals to withhold their ideas (Camacho & Paulus, 1995). Individuals with middle status may be particularly threatened by the prospect that their ideas will be evaluated negatively by others, and these feelings of threat may inhibit their willingness to share creative ideas. These feelings of threat are appropriate given that individuals who express creative ideas, as opposed to purely practical ideas, risk appearing quirky and unpredictable; impressions that may prevent them from maintaining their position or may cause them to move down the status hierarchy (Mueller et al., 2011). In other words, middle-status individuals operate just outside the warm glow of social acceptance that high status provides, yet they are high enough in the status hierarchy to fear the loss of status. Thus, the prospect of being criticized and negatively evaluated for suggesting a creative idea may be particularly salient to individuals with middle status. Bolstered by the social acceptance of others (Lount & Pettit, 2012; Magee & Galinsky, 2008), high-status individuals may be more confident that, even if their creative ideas are evaluated negatively, they will not feel threatened by the prospect of status loss. Hence, confidence in the security of their position in the group will embolden high-status individuals to risk suggesting creative ideas. Compared with individuals with middle status, having low status may also be somewhat liberating for two reasons. First, knowing that they are at the very bottom of the status hierarchy, and therefore not likely to make a significant move further down than they already are, may substantial reduce lowstatus individuals’ evaluation concerns (Blau, 1955). Second, a viable way to gain status may be by suggesting ideas that are very novel and thus distinguishable from others (Clydesdale, 2006; Merton, 1968; Sutton & Hargadon, 1996). Though the middle status could, in theory, improve their status position by being creative, doing so would entail much more risk compared with the risks faced by those with low status. If the middle-status individuals diverge too far from existing solutions, they could be criticized, rejected, and relegated to a lower status position—this risk would be particularly salient given that the prospect of status loss looms larger than the opportunity for status gain (Pettit et al., 2010). Low-status group members may risk moving from “low” to “lower,” but doing so would not constitute a meaningful change in status position—particularly compared with moving from “middle” to “low” (Phillips & Zuckerman, 2001). This relative lack of evaluation apprehension, combined with the potential for acquiring status by standing out from the group, could liberate low-status individuals to suggest novel solutions. The Present Research In sum, we hypothesize that the threat of status loss should make individuals with middle status less creative than individuals at the top or the bottom of the status hierarchy. In Study 1, we manipulated status level and then asked participants to generate new ideas. We expected that middle-status individuals would be most vulnerable to the threat of status loss when there was the prospect of evaluation, and would therefore generate the least creative ideas under conditions of expected evaluation but not under the cover of anonymity. Accordingly, we also manipulated whether participants expected that their ideas would be evaluated after the experiment or whether they would generate ideas anonymously. In Study 2, we tested the robustness and replicability of this effect using a different status manipulation and a different measure of creativity. In addition, in Study 2, we measured the threat of status loss directly in order to trace the underlying psychological process. In Study 3, we investigated the possibility that having middle status might narrow rather than broaden attention, and improve performance on a task that demands convergent rather than divergent thought. We again investigated the role of threat of status loss by both manipulating expected evaluation and measuring threat of status loss directly. One implication of our theoretical perspective is that the curvilinear relationship between creativity and status is unique to status as opposed to power. Because power is derived from control over resources rather than conferred by the group, individuals with middle power should feel less susceptible to the threat of status loss (Blader & Chen, 2012). Hence, in Study 4, we manipulated power level (high, middle, and low) and expected evaluation. Finally, in Study 5, we investigated the threat of status loss directly, by manipulating the stability of the status hierarchy. We expected that individuals with middle status would experience less status threat, and thus be just as creative as their counterparts with high or low status when they are assured that they cannot move further down the hierarchy. The current set of studies extends existing research in at least two important ways. First, our findings extend the classical research on middle-status conformity to the realm of problem solvThis document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. STATUS AND CREATIVITY 3
DUGUID AND GONCALO m 1987:Simon n1999 ence a blind to the psychological experience that has yet to be fully explicate Study 1 not at all novel Method p= which consisted of a 3 (status:high vs.middle vs.low)2 Results Manipulation checks. Gruenfeld,Magee, 2010)two coders blind to both the experimenta condition and ou y the n this stu We verified that ipulations did not also influenc power. Participants then read. 2012).Endpoints were )and 9 (a great deal,7 Status ent t which t and look up to 2.2060=062.p618 0001.evaluation.F.206) nd the valuation Panticinants indicated whethe MIDDL back fror ter or that the o the stent with the condi wha ion to which they were rand omly as Thes In the second phase,participants were told. valuation,F( .206=8.49.p=004 16.08.SD 7.69hg your group yo( iom.F2.206 3.25. 041.=0.034.Planned contasts at whe high-saus(M=16.84,SD=8.24),69) -2.53,p=.014,and given min to ask Finall ignificant difference in numbe of idea eenerated by middle 4704 0.29 and high-status(M of idec rated. =18.80,D=841 al is able to generate in a Thereassiniean itterenc innumber ot lsen
ing and creative task performance. Second, by manipulating expected evaluation and measuring threat of status loss directly, we demonstrate mediating evidence of the underlying psychological mechanism that might explain how status might influence a wide range of outcomes yet to be investigated in existing research. In sum, we present evidence suggesting that middle status is a unique psychological experience that has yet to be fully explicated. Study 1 Method Participants and design. Two hundred twelve participants (mean age 20.13 years; females 43%) took part in the study, which consisted of a 3 (status: high vs. middle vs. low) 2 (evaluation: yes vs. no) between-participants design. Participants were undergraduate students who were given course credit for taking part in the study. Procedure. The study involved two phases. In Phase 1, we manipulated status using an adapted version of Bowles and Gelfand’s (2010) psychological experience of status. Specifically, participants were told, In this study, we were interested in how groups interact to solve difficult problems. Before you interact as a group, however, we would first like to learn more about you as an individual. Please read and respond to the following question. Participants then read, Status determines the extent to which people respect and look up to you or defer to your opinion because you have a lot of experience or competence. Please recall a particular incident in which you were part of a group and in that group your status relative to others was “HIGH,” that is at the top of the status hierarchy/ around the “MIDDLE,” that is neither the top nor the bottom of the status hierarchy/“LOW,” that is at the bottom of the status hierarchy. Please describe this situation in which you had high/middle/low status – what happened, how you felt, and so forth. In the second phase, participants were told, Now we would like you to prepare for the group task by generating solutions to a problem. As you complete the task, please keep in mind that (you will get feedback on your ideas from the experimenter, which could affect your role in the group)/(your ideas will remain anonymous and will not affect your role in the group). They were told that this task involved a scenario in which they would be asked to generate ideas. The scenario was as follows: A recent survey suggested that students are overwhelming dissatisfied with the current state of the undergraduate lounge. The Dean has decided to get students to generate ideas about how to solve the problem. Your task is to come up with as many ideas as you can about how to improve the undergraduate lounge. Participants were given 7 min to complete the task. Finally, participants completed a questionnaire, which included the status, power, and evaluation manipulation check items. Creativity. Number of ideas generated. A simple indicator of creativity is the sheer number of ideas an individual is able to generate in a fixed amount of time. The more ideas an individual generates, the more likely he is to arrive at a novel solution (Diehl & Stroebe, 1987; Simonton, 1999). Idea novelty. In addition to the sheer number of ideas, we also coded for novelty directly. Two coders who were blind to the experimental condition and predictions of the study coded each idea for novelty, which was defined as “the extent to which the idea resembles the typical undergraduate lounge.” Each coder was given a scale of 1 to 5, with the following definitions for specific points on the scale: 5 extremely novel, 3 average novelty, 1 not at all novel. The interrater correlation (ICC) was significant (ICC 0.81, p .006), so the scores were averaged together. Results Manipulation checks. Status. Consistent with previous research (e.g., Galinsky, Gruenfeld, & Magee, 2003; Inesi, 2010; Ronay & von Hippel, 2010), two coders blind to both the experimental condition and our hypotheses categorized the priming essays as high, middle, or low status, or whether they could not determine which category the essay belonged to. One hundred percent of the essays were categorized correctly in terms of condition. We verified that our status manipulations did not also influence power. Participants reported to what extent they felt dependent (reverse coded), powerful, and dominant (Duguid & Goncalo, 2012). Endpoints were 1 (very little) and 9 (a great deal, .78. Analysis of variance (ANOVA) revealed no main effect of status, F(2, 206) 0.62, p .618, p 0.001, evaluation, F(1, 206) 0.36, p .782, p 0.001, or the interaction between status and evaluation, F(2, 206) 2.01, p .128, p 0.010. Evaluation. Participants indicated whether they would get feedback from the experimenter or that their ideas would remain anonymous. All but two participants reported information that was consistent with the condition to which they were randomly assigned. These two individuals were excluded from the final analyses; however, including their data yielded identical results. Creativity. Number of ideas generated. ANOVA revealed no significant main effect of status, F(2, 206) 0.16, p .849, p 0.002. There was a main effect of evaluation, F(1, 206) 8.49, p .004, p 0.04, such that participants generated fewer ideas when they assumed they would be evaluated (M 16.08, SD 7.69) than when their ideas were anonymous (M 19.62, SD 10.15). There was also a significant interaction between status and evaluation, F(2, 206) 3.25, p .041, p 0.034. Planned contrasts showed that when ideas were being evaluated, middle-status participants generated fewer ideas (M 13.42, SD 5.79) than high-status (M 16.84, SD 8.24), t(69) 2.53, p .014, and low-status (M 17.79, SD 8.19), t(65) 2.04, p .045, participants. There was no significant difference in ideas generated by high- and low-status participants, t(70) 0.63, p .625. In contrast, when ideas were not being evaluated, there was no significant difference in number of ideas generated by middlestatus participants (M 21.47, SD 10.29) and high-status (M 18.56, SD 10.51), t(70) 1.14, p .260, and low-status (M 18.80, SD 8.41), t(69) 1.13, p .263, participants. There was also no significant difference in number of ideas genThis document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 4 DUGUID AND GONCALO
STATUS AND CREATIVITY y high-and low-staus paricipants.(.1. d a stal ANOVA led no significa n effect of ability n.F( vould be used to us and m.F2.2062 038 rship.After th scored.p vel ideas 77D :0.47)than high- (M= 210 istant (low status)po osition The )R6 040.an ould gra high and low-status oower.that is.the and makine fina On th when id e orce their s ()than thoe of 201,SD ed th r to uld pants told to 0g1 then giver Discussion a picture of our hypot sis tha ment all of the articipants in the s sion t he ndividiasEemerac Ity of the ideas generated w hen participants thought their Finally,participant pleted an implicit threat studies (DeMarre Pe 2005) Ve use Has.Katz,Riz ity ure of creativity to der the ity of the able to pe rceiv e it.They vere told that after the word left th te something new (Ward.94). word is kn wn to his task as indi word that had just b een flashed dy ko ble thos is that they fear their ideas will be criticized for bei rd.four words re pres nted and remained on the screen until idual ned by the prospect of evaluation we -relatco sked to ne of the four eoptions wasa threat-of-status-los Study 2 Method Following directly from previous research tud were blind to the study p dicti s:high vs.m ow) ngs and h War'e(1004 chem nts came to the laboratory ir and unusual configuration of the sensory organs (e.g..eyes located
erated by high- and low-status participants, t(69) 0.11, p .914. Idea novelty. ANOVA revealed no significant main effect of status, F(2, 206) 0.99, p .372, p 0.01, or evaluation, F(1, 206) 0.08, p .783, p 0.001. There was, however, a significant interaction between status and evaluation, F(2, 206) 3.32, p .038, p 0.031. Planned contrasts showed that when participants believed their ideas were being evaluated, middle-status participants generated less novel ideas (M 1.77, SD 0.47) than high-status (M 2.10, SD 0.45), t(65) 2.09, p .040, and low-status (M 2.13, SD 0.52), t(65) 2.09, p .040, participants. There was no significant difference in novelty of ideas generated by high- and low-status participants, t(70) 0.20, p .840. On the other hand, when ideas were anonymous, middle-status participants’ ideas were no less novel (M 2.06, SD 0.32) than those of high-status (M 2.01, SD 0.36), t(70) 0.66, p .511, and low-status (M 1.98, SD 0.34), t(69) 1.11, p .272, participants. There was also no significant difference in idea novelty for high- and low-status participants, t(69) 0.40, p .691. Discussion In support of our hypothesis that status would bear a U-shaped relationship to creativity, the results demonstrate that when they expected to be evaluated, middle-status individuals generated fewer and less novel ideas compared with high- and low-status participants. However, there was no difference in the number and novelty of the ideas generated when participants thought their ideas would remain anonymous. In Study 2, we planned to replicate and extend these findings. We used a different status manipulation and investigated threat of status loss as the mechanism underlying the relationship between status and creativity. We also employed a different measure of creativity to demonstrate the robustness and generalizability of the effect to different tasks. An important part of the creative process is the ability to go beyond what is known to generate something new (Ward, 1994). This task is quite difficult, however, as individuals are constrained by what they already know and often generate products or ideas that very closely resemble those that already exist (Ward, 1994). One reason that individuals may find it difficult to break from what is already known is that they fear their ideas will be criticized for being too unusual (Camacho & Paulus, 1995; Mueller et al., 2011). Therefore, in line with our general hypothesis that middle-status individuals are most threatened by the prospect of evaluation, we expect that they will be less creative in a structured imagination task in which they are asked to deliberately generate a novel entity. Study 2 Method Participants and design. One hundred thirty participants (mean age 19.91 years; females 41%) took part in the study, which consisted of a 3 (status: high vs. middle vs. low) 2 (evaluation: yes vs. no) between-participants design. Participants were undergraduate students who were given course credit for taking part in the study. Procedure. Participants came to the laboratory in groups and were told that they would be participating in several studies, the first of which would investigate status and group performance. Participants completed a status assessment consisting of questions that they were told measured ability on the upcoming group decision-making task, and therefore would be used to establish their status position (Pettit et al., 2010). The assessment consisted of a combination of questions related to reasoning, creativity, and leadership. After the assessment was ostensibly scored, participants were randomly assigned to a President (high status), Middle Manager (middle status), or Assistant (low status) position. They were told that their roles differed in how much others would grant them respect and prestige, but that they did not differ in the amount of power, that is, the amount of resources and making final decisions. Moreover, in order to reinforce their status, participants wrote their roles on nametags that they wore and also wrote three behaviors that would lead to individuals being granted their role. Participants were told that before meeting with their group, they would be completing a related individual task. They were given a structured imagination task in which they were told to “imagine going to another galaxy in the universe and visiting a planet very different from Earth” (Ward, 1994). Participants were then given 7 min to draw a picture of an animal that is “local to this other planet.” Participants were either told that at the end of the experiment all of the participants in the session would compare all of the drawings and vote on the drawing that most closely followed the instructions given or they were told that none of the other participants in the session would see their drawing so there would be no evaluation of the drawings. Finally, participants completed an implicit threat measure used in previous studies (DeMarree, Wheeler, & Petty, 2005). We used an implicit measure because individuals are not always forthcoming about the level of threat they experience (Has, Katz, Rizzo, Bailey, & Moore, 1992). Participants were told a word would be flashed on-screen so quickly that only their subconscious would be able to perceive it. They were told that after the word left the screen, a list of words would appear, and they should use their feelings at the moment to select which word they thought was the word that had just been flashed. Trials began with a premask of “X”s serving as an orienting stimulus for 2,000 ms, followed by subliminal presentation (17 ms) of the target words, which was a nonsensical string of letters. A postmask of Xs covered the target word for 1,000 ms. Afterward, four words were presented and remained on the screen until participants made their selection of which word they believed was flashed on-screen. Half of the 12 trials were target trials, in which one of the four response options was a threat-of-status-loss-related word (i.e., “loss,” “demote,” “threat,” “lower, “devalue,” and “downgrade”). The position of the threat-of-status-loss-related words in the response options was randomized, as was the order of the trials. Creativity coding. Following directly from previous research (e.g., Ward, 1994), structured imagination was coded for the atypicality of the space creatures’ sensory organs. Two trained coders who were blind to the study predictions assessed the drawings and accompanying descriptions for evidence of “atypical” sensory organs. In accordance with Ward’s (1994) original coding scheme, space creatures were considered atypical if they (a) lacked a major sensory organ (i.e., eyes, ears, nose), (b) had atypical numbers of a sensory organ (e.g., three eyes), (c) demonstrated an unusual configuration of the sensory organs (e.g., eyes located This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. STATUS AND CREATIVITY 5
6 DUGUID AND GONCALO typical function (e.g.,ears for protection).The total number of typical features v and so their r ratings were averaged together to create an overal 144)identifie n overall measure of threat of 83 with high or low status,(85)=0.17.p= 867 There was also F1,124 which included 9.01. was anonymous (M Results an inte ion checks The status manipulatio status (M=2.85.Sp 1.35)partic and re (0.0 -401p2001 points were 1 (very or low-status (M 1.23.5D 1.11)40) 4.27 2.124 i0.1,p2001,ng =O.139.Pl 06)th Presid ere not bein 135.84)=2 38.p=004.or Assistant role (M=3.56 evaluate the number of h lated SD high-status (M 1.20,50 100.4 .F(1. 1240= sD=09444) -0.15p=882 evaluation F(2 124)=2 20 ence in number of threat-r late g the 24)= 1.00.p ction betweer .F2 632 All that was h they were ran 9 ANOVA reveale no main effect of dra 402. 0.006.There was,however. tatus and evaluatio mid and lo s indivi 013 status par ativity nigh -status (M =3.86.SD 14).140)04 when threa and lo taus(M=3.91,SD=2.00.40)=2.06,p=.046 of status los was entered into the regression,B 0.34,t 1.4 caivhn re not significantly less creative than high-status (M=3.53 not in the no- aluation condition.138 166 status n the bia 95% confidence inter
below the nose), (d) had an exaggerated or unusual ability (e.g., eyes that had laser beams), or (e) had sensory organs that served an atypical function (e.g., ears for protection). The total number of atypical features was tallied for each participant. The ratings of the two coders reached significant agreement (ICC 0.77, p .008), and so their ratings were averaged together to create an overall measure of creative performance. Threat of status loss. To create an overall measure of threat of status loss, we computed a composite measure by summing the number of threat-related words selected. Scores ranged from 0 to 6; higher scores indicated a stronger feeling of threat of status loss (DeMarree et al., 2005). Finally, participants completed a short survey, which included questions related to demographic information and the manipulation checks. Results Manipulation checks. Status. The status manipulation was checked using four items adapted from Anderson, John, Keltner, and Kring (2001). This included the question, “How much status, prominence and respect would others grant you?” Endpoints were 1 (very little) and 9 (a great deal), .80. ANOVA revealed a main effect of status, F(2, 124) 10.11, p .001, p 0.139. Participants randomly assigned to the President role felt higher status (M 5.25, SD 1.06) than those assigned to the Middle Manager role (M 4.33, SD 1.35), t(84) 2.38, p .004, or Assistant role (M 3.56, SD 1.37), t(86) 4.59, p .001. Additionally, Middle Managers felt higher status than Assistants, t(88) 2.00, p .046. There was no significant main effect of evaluation, F(1, 124) 0.32, p .571, p 0.003, nor an interaction between status and evaluation, F(2, 124) 2.20, p .108, p 0.03. Individuals’ perceptions of power were also measured using the scale from Study 1 ( .91). ANOVA revealed no main effect of status, F(2, 124) 1.00, p .370, evaluation, F(1, 124) 1.99, p .160, or the interaction between status and evaluation, F(2, 124) 0.46, p .632. Evaluation. All participants reported information that was consistent with the evaluation condition to which they were randomly assigned. Creativity. ANOVA revealed no main effect of status, F(2, 124) 0.19, p .825, p 0.003, nor evaluation, F(1, 124) 0.71, p .402, p 0.006. There was, however, a significant interaction between status and evaluation, F(2, 124) 4.18, p .017, p 0.063. Planned contrasts showed that when they thought their drawings were being evaluated, middle-status participants generated less creative drawings (M 2.55, SD 2.27) than high-status (M 3.86, SD 1.74), t(40) 2.11, p .041, and low-status (M 3.91, SD 2.00), t(40) 2.06, p .046, participants. There was no significant difference in the creativity of high- and low-status participants, t(42) 0.08, p .936. In contrast, when they did not think their drawings were being evaluated, middle-status participants’ (M 4.33, SD 2.20) drawings were not significantly less creative than high-status (M 3.53, SD 2.23), t(44) 1.22, p .230, and low-status (M 3.38, SD 1.75), t(41) 1.55, p .128, participants. There was also no significant difference in the creativity of high- and low-status participants, t(41) 0.25 p .802. Threat of status loss. We hypothesized that being middle status would elicit the strongest threat of status loss when individuals expect to be evaluated. ANOVA on the composite measure of threat of status loss revealed a main effect of status, F(2, 124) 7.73, p .001, p 0.111, such that middle-status participants (M 1.98, SD 1.44) identified more threat-of-status-lossrelated words than high-status (M 1.24, SD 1.10), t(83) 2.65, p .010, or low-status (M 1.20, SD 1.01), t(86) 2.94, p .004, participants. There was no significant difference in threat-of-status-loss-related words chosen by those with high or low status, t(85) 0.17, p .867. There was also a main effect of evaluation F(1, 124) 9.01, p .003, p 0.07. When participants assumed they were being evaluated, they identified more threat-related words (M 1.75, SD 1.41) than when they thought their performance was anonymous (M 1.20, SD 0.98). Importantly, there was also an interaction between status and evaluation, F(2, 124) 7.15, p .001, p 0.103. Planned contrasts showed that when they thought their drawings were being evaluated, middle-status (M 2.85, SD 1.35) participants identified more threat-of-status-loss-related words than high-status (M 1.27, SD 1.20), t(40) 4.01, p .001, or low-status (M 1.23, SD 1.11), t(40) 4.27, p .001, participants. There was no significant difference in threat-ofstatus-loss-related words identified by high- and low-status participants, t(42) 0.13, p .897. Conversely, when they thought their drawings were not being evaluated, the differences in the number of threat-related words identified by middle-status participants (M 1.22, SD 1.04) and high-status (M 1.20, SD 1.00), t(41) 0.06, p .956, and low-status (M 1.17, SD 0.94), t(44) 0.15, p .882, participants were not significant. There was also no significant difference in number of threat-related words identified by highand low-status participants, t(41) 0.09, p .930. Mediation analysis. To test the hypothesis that threat of status loss accounts for the relationship between status and evaluation on creativity, we conducted a mediation analysis. There was no main effect of status, 0.03, t 0.38, p .704, or evaluation, 0.07, t 0.81, p .418. There was also an interaction between status and evaluation on creativity, 0.65, t 2.59, p .011, such that when individuals thought their drawings were being evaluated, middle-status participants generated less creative drawings than high- and low-status participants, but there was no significant difference in the creativity of the drawings generated by high-, middle-, and low-status individuals when they did not think their drawings were not being evaluated. There was also a significant relationship between threat of status loss and creativity, 0.44, t 5.48, p .001. The interaction between status and evaluation on creativity was no longer significant when threat of status loss was entered into the regression, 0.34, t 1.41, p .161. To further interpret these findings, we examined the conditional indirect effects at the levels of status. These effects showed that threat of status loss mediated the effect of evaluation on creativity in the evaluation condition, z 1.95, p .035, but not in the no-evaluation condition, z 1.38, p .166. To assess whether threat of status loss mediated the interactive effects of status and evaluation on creativity, we utilized bootstrap estimates to generate bias-corrected 95% confidence intervals (CIs). Based on 1,000 samples, the 95% CI ranged between 0.19 This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 6 DUGUID AND GONCALO
STATUS AND CREATIVITY and98.If the CI does not include ero.the indirect effect is word word and the color r of the word would be different.In both Discussion olor of the ompared with high-and low atus individuals. eveal ev one's final sco e on the task to the entire o in th the drawings eing eval ted,there were told hat u ce o r.we measured and direct O he After the S pleted a short survey.which incud Study3 graphic infe manipulation check The t Results ions.The cor or task pe nce are no Manipuation checks ible that the thre e fror of at 0.196 3) to th anager role (M 8.p ticinants signed to the Middle Manger k 1950 Wells Matthews 100d)Thus s we predict that the evaluation.F(1.136)=0.18.p=.671. 0.001 and ne improve cov mddl rather than diver 136 =0.18.p cal cor ary of ou e aso measured using the thi e ction us valuation.F(136)0.8 之 on which All but one participant ted information tha f the luding the data yielded the identical patter of resuts. rinted in vellow ink ked t the coo of ne with previous r we compu d th the ink whe the Stroop e as an hen it did match the color.Perforn ce on this task improve .200 Baumcister. 200% Nemeth et al.1992) atencies e ater thn deviations above the men (.e Method 000ms)￥ RTodedas2.00 c=20.25 s:fer took part in the stud vs.mddle OW) 136)=0.91.p=342.=0.007.However.as expected,the tin the study in sude 2 The s was m same participants as a study of information processing and coor per-
and 0.98. If the CI does not include zero, the indirect effect is deemed significant, and mediation can be said to be present (Hayes, 2009; Preacher & Hayes, 2004). Discussion Consistent with the pattern of results in Study 1, we found that when being evaluated, middle-status individuals generated less creative drawings compared with high- and low-status individuals. When their drawings were not being evaluated, there was no difference in creativity for middle-status, high-status, and lowstatus individuals. Moreover, we measured and directly demonstrated the threat of status loss as the mechanism underlying the relationship between status and evaluation and creativity. Study 3 The results of Studies 1 and 2 suggest that individuals with middle status will be at a disadvantage on tasks that demand creative solutions. The consequences of middle status for task performance are not necessarily negative, however. It is possible that the threat of status loss might cause individuals with middle status to narrow their focus of attention, filter out irrelevant stimuli, and think more convergently on only a subset of relevant information (Gable & Harmon-Jones, 2010). It is well known that when individuals experience relatively intense negative emotions like anxiety (as opposed to low-intensity negative emotions like sadness), their attentional focus narrows (Easterbrook, 1959; Wells & Matthews, 1994). Thus, we predict that the perceived threat of status loss may actually boost performance on tasks that demand narrowed rather than broadened attention, and convergent rather than divergent thought. Our prediction that middle status should improve cognitive control is a logical corollary of our prediction that middle status should diminish creativity: Performance on both kinds of tasks is impacted by a narrowed focus of attention, only in opposite ways. We test this prediction using the Stroop test (Stroop, 1935)—a task on which convergent thinking actually facilitates performance (Friedman & Förster, 2005; Nemeth, Mosier, & Chiles, 1992; Peterson & Nemeth, 1996). On this task, participants view the name of a color printed in ink of the same color (e.g., the word “Red” printed in red ink) or a different color (e.g., the word “Red” printed in yellow ink) and are asked to name the color of ink. Stroop (1935) reported that participants took significantly more time to name the ink when the word did not match the color than when it did match the color. Performance on this task improves when participants are able to focus their attention more convergently on the color of ink while filtering out the word itself (Nemeth et al., 1992). Method Participants and design. One hundred forty-two participants (mean age 20.25 years; females 45%) took part in the study, which consisted of a 3 (status: high vs. middle vs. low) 2 (evaluation: yes vs. no) design. Participants were undergraduate students who were given course credit for taking part in the study. Procedure. Status was manipulated using the same procedure as in Study 2. The next part of the study was presented to participants as a study of information processing and color perception. Participants were told that they would be presented with congruent words, for which the color word and the color of the word would match, and incongruent words, for which the color word and the color of the word would be different. In both conditions, they were told that their task was to respond to the color of the word as quickly and accurately as they could. Participants were also either told that their performance on the task was not anonymous, and at the end of the study the experimenter would reveal everyone’s final score on the task to the entire group in the session; or they were told that their performance on the task was completely anonymous, and at the end of the study the experimenter would dismiss them without revealing their final score on the task. After the Stroop test, all participants completed the threat of status loss measure used in Study 2. Finally, participants completed a short survey, which included questions related to demographic information and the manipulation checks. Results Manipulation checks. Status. Using the scale from Study 2 ( .80), ANOVA revealed a main effect of status, F(2, 136) 16.31, p .001, p 0.196. Participants randomly assigned to the President role felt higher status (M 4.83, SD 1.13) than those assigned to the Middle Manager role (M 4.03, SD 1.10), t(84) 3.48, p .001, or Assistant role (M 3.51, SD 1.13), t(86) 2.22, p .029. Additionally, participants assigned to the Middle Manger role felt higher status than those assigned to the Assistant role, t(88) 5.65, p .001. There was no significant main effect of evaluation, F(1, 136) 0.18, p .671, p 0.001, and no interaction between status and evaluation, F(2, 136) 0.18, p .838, p 0.003. Individuals’ perceptions of power were also measured using the scale from Study 1, .79. ANOVA revealed no main effect of status, F(2, 136) 0.43, p .654, evaluation, F(1, 136) 0.88, p .351, or interaction between status and evaluation, F(2, 136) 0.79, p .456. Evaluation. All but one participant reported information that was consistent with the condition to which they were randomly assigned. This individual was excluded from the final analyses, but including the data yielded the identical pattern of results. Stroop test. In line with previous research, we computed the difference in time between the incongruent and congruent trials in the Stroop task, and used this difference as an assessment of convergent thinking (DeWall, Baumeister, & Vohs, 2008; Richeson & Trawalter, 2005).1 Moreover, consistent with the procedures detailed in Richeson and Shelton (2003) and Richeson and Trawalter (2005), all Stroop latencies greater than 2.5 standard deviations above the mean (i.e., times 2,000 ms) were recoded as 2,000 ms. Lower scores reflect faster reaction times (RTs) and indicate better information processing. Tests of between-subjects effects yielded no main effect of status, F(2, 136) 1.49, p .230, p 0.02, or evaluation, F(2, 136) 0.91, p .342, p 0.007. However, as expected, the 1 Accuracy was not included in the final analysis, because the overall error rate was very low (1%) and including incorrect trials did not impact the results. This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. STATUS AND CREATIVITY 7
DUGUID AND GONCALO we examined the conditional indirect effectsat the 47069,D= d low tatus (M 451.32.SD =280.40,44) 2.40.p condition. 3.39.p<.001 but not in the no-evaluation ss whether loss mediated the inter anc wasnot being d ther and high-status (M=437.09.SD=230.33).(6) -0.34.p 2009 Preacher&Hayes.004).Based mples the ow-status (M41.68. 24.91.F1.45) 95%CI ranged between 16.93 and 59.98.excluding zero. Discussion ANOVA O hat individu th mid and and ically.RT 2.03. -statu no significant diffe e in th indivTeatedwords sen by those with high or 9u such that when y thought the g evaluate anding of why middl when they thoughhe 0.94,SD=0.94 eon tasks that demand creative solutions,occupying the middle ted,middl-statusnd ve tctendtcdourpweiousfining.byinvcstigatiagaakin words thar atus (M =0.92.SD=1.04 rienced by middle be a off i =0.79.SD ns of tas On the on han -430 nce in th relate n their creativity ing 2011:Camach and,this focus may y actually improve perf on tasks,lik 279 72 individua d status part nts,45)=183,p .08 Study 4 est the hypothe s th the lassical 63n 376 However,ther vas an inter status and examining the impact of being in the middle versus the top an evaluated middle-status individuals aster than the RTs of those with high middlle.r betw status ted.the differences in the rTs of individuals with middle high. low status ere of stat nlike status,which is social etween status and evaluation on RT was no longer significant
interaction between status and evaluation was significant, F(2, 136) 3.15, p .046, p 0.04. Planned contrasts showed that when performance was being evaluated, middle-status participants’ RTs (M 270.63, SD 223.89) were faster than those of high-status (M 470.69, SD 234.59), t(45) 2.98, p .005, and low-status (M 451.32, SD 280.40), t(44) 2.40, p .021, participants. There was no significant difference in RTs of high- and low-status participants, t(47) 0.26, p .794. When performance was not being evaluated, the differences in the RTs of middle-status participants (M 463.73, SD 312.50) and high-status (M 437.09, SD 230.33), t(46) 0.34, p .738, and low-status (M 413.68, SD 224.91), F(1, 45) 0.63, p .533, participants were not significant. There was no significant difference in the RTs of high- and low-status participants, t(45) 0.35, p .726. Threat of status loss. ANOVA on the composite threat of status loss measure revealed a main effect of status, F(2, 136) 8.15, p .001. Middle-status individuals (M 1.67, SD 1.74) identified more threat-of-status-loss-related words than high-status (M 1.08, SD 1.04), t(93) 2.03, p .045, or low-status (M 0.79, SD 0.75), t(86) 3.20, p .002, individuals. There was no significant difference in threat-of-status-loss-related words chosen by those with high or low status, t(94) 1.59, p .116. There was also a main effect of being evaluated, F(1, 136) 7.18, p .008, such that when they thought they were being evaluated (M 1.41, SD 1.53), individuals identified more threat-related words than when they thought they were not being evaluated (M 0.94, SD 0.94). Importantly, there was also a significant interaction of status and evaluation, F(2, 136) 12.57, p .001. Planned contrasts showed that when performance was being evaluated, middle-status individuals (M 2.64, SD 1.87) identified more threat-of-statusloss-related words than high-status (M 0.92, SD 1.04), t(45) 3.96, p .001, or low-status (M 0.79, SD 0.83), t(44) 4.39, p .001, individuals. There was no significant difference in threat-of-status-loss-related words identified by highand low-status individuals, t(47) 0.48, p .636. When there was no evaluation, the differences on the threatrelated words identified by middle-status individuals (M 1.25, SD 1.03) and high-status (M 0.79, SD 1.02), t(46) 1.55, p .129, and low-status (M 0.78, SD 0.67), t(45) 0.04, p .972, individuals were not significant. There was also no significant difference in threat-related words identified by highand low-status participants, t(45) 1.83, p .080. Mediation analysis. To test the hypothesis that threat of status loss accounts for the relationship between status and evaluation on convergent thinking, we conducted a mediation analysis. As established above, there was no main effect of status, 0.14, t 1.63, p .106, or evaluation, 0.07, t 0.89, p .376. However, there was an interaction between status and evaluation on convergent thinking, 0.20, t 2.29, p .024, such that when performance was being evaluated, middle-status individuals’ RTs were faster than the RTs of those with high- and low-status participants. However, when performance was not being evaluated, the differences in the RTs of individuals with middle, high, and low status were not significant. There was a significant relationship between threat of status loss and RT, 0.43, t 5.70, p .001. The interaction between status and evaluation on RT was no longer significant when threat of status loss was entered into the regression, 0.08, t 0.35, p .700. To further interpret these findings, we examined the conditional indirect effects at the levels of status. These effects showed that threat of status loss mediated the effect of evaluation on RT in the evaluation condition, z 3.39, p .001 but not in the no-evaluation condition, z 1.36, p .295. To assess whether threat of status loss mediated the interactive effects of status and evaluation on RT, we utilized bootstrap estimates to generate bias-corrected 95% CIs (Hayes, 2009; Preacher & Hayes, 2004). Based on 1,000 samples, the 95% CI ranged between 16.93 and 59.98, excluding zero. Discussion As predicted, the results demonstrate that individuals with middle status performed better on the Stroop task, which requires convergent thought, than individuals with high and low status when they thought they were being evaluated. Specifically, RTs for correctly identifying the color of the words were significantly shorter for individuals with middle status compared with those with high status and low status. Furthermore, threat of status loss was shown to mediate the relationship between status and evaluation and performance on the Stroop task. These findings provide a deeper understanding of why middle status might stifle creativity. Although we have shown, in Studies 1 and 2, that individuals with middle status will be at a disadvantage on tasks that demand creative solutions, occupying the middle status position in the social hierarchy may not inevitably lead to error, as the early research assumed (Homans, 1961). In Study 3, we extended our previous findings by investigating a task in which performance might be boosted by threat of status loss. The anxiety experienced by middle-status individuals may be a trade-off in terms of task performance. On the one hand, the anxiety middlestatus individuals feel at the prospect of being evaluated negatively and potentially losing status may constrain their creativity by limiting their willingness to explore new solutions (Byron & Khazanchi, 2011; Camacho & Paulus, 1995). Yet, on the other hand, this focus may actually improve performance on tasks, like the Stroop test, that require convergent thought to complete quickly and accurately (Friedman & Förster, 2005; Peterson & Nemeth, 1996). Study 4 Drawing on the classical research on middle-status conformity, we have argued that middle-status individuals should be less creative than individuals with high or low status. Would the same hold true for power? Study 4 builds upon the previous findings by examining the impact of being in the middle versus the top and bottom of the power hierarchy on creativity (Galinsky et al., 2008); previous research focused on high and low power, excluding the middle. Though differences between power and status are rarely delineated, there is recent evidence to suggest that their consequences are not necessarily identical (Blader & Chen, 2012). An important difference between status and power is that, unlike status, which is socially conferred, power is a property of the actor and is less susceptible to the subjective evaluations of This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 8 DUGUID AND GONCALO
STATUS AND CREATIVITY 2 Pow ontrol the ability o a pa you ividuals who may be /"L0 ect to or others (Kell uld Results sing access to valued r should Manipulation checks. Iwo co ders.blind o both the experimental the more power an individual has.the s fications of the essay: rand.2001).The tendency of the powerful to disregard others nd thus m nore adept at generating(Galinsky I,po ant (reve and sub dinate (re -coded 2008 were should infuse )and archy.we do wer.F2.142) 3.6 ha actors to be and anxious than their lov disposal.This reaso ning leads u oredict that the relationshir condition felt mor en power and a different form than v 001 relationship between power and creativity to be positive and tion between status and evaluation,)199. 0.0 nd powe at the tically important.b whether e pow .weus d in St igher up n the po hierarc should boo 0.00 p=977,n =0.001 ting these differential effects of status and po would ain effect of we.F2.14216.67.p<00 16.70 (M=13.24.SD=5.36.96) =234.p owe Method s in the middl 142 6221.001.=0.305.When ideas wer bein part in the stud 060, ant main effec e of power instead of sta sby using an adapted vers G女yae nts in the 0.5 Specifically.participants read. M-2.87.SD=0.551.96=-3.19.p=002.and1ow-powc
others (Blader & Chen, 2012). We expect that this key distinction between status and power would also lead to differential effects on creativity. Power is typically defined as the extent to which an individual controls valued resources (Fiske, 2010; Magee & Galinsky, 2008). Therefore, unlike low-status individuals who may be liberated by having nothing to lose (Blau, 1955; Hollander, 1960), low-power individuals live in a world of risk and looming threats because they lack access to valued resources and are therefore subject to the whims of others (Keltner et al., 2003; Magee & Galinsky, 2008). Gaining power should alleviate this vulnerability, as increasing access to valued resources should also increase feelings of control (Fast, Gruenfeld, Sivanathan, & Galinsky, 2009; Kraus, Chen, & Keltner, 2011). Moreover, the more power an individual has, the less concerned and less aware of others’ needs and opinions he will be (Galinsky, Magee, Inesi, & Gruenfeld, 2006; Lee-Chai, Chen, & Chartrand, 2001). The tendency of the powerful to disregard others’ judgments makes them more resistant to conformity pressure and thus more adept at generating creative solutions (Galinsky et al., 2008). Therefore, although both power and status should infuse individuals with confidence at the top of the hierarchy, we do not expect low-power individuals to be as creative as those with high power (Galinsky et al., 2008), nor do we expect middlepower actors to be more insecure and anxious than their lowpower counterparts, given they have more resources at their disposal. This reasoning leads us to predict that the relationship between power and creativity will take a different form than we found for status. That is, unlike with status, we expect the relationship between power and creativity to be positive and linear. Comparing status and power at the middle of the hierarchy is theoretically important, because doing so sheds light on why middle status (as opposed to middle power) is uniquely constraining. If power involves control over resources, then being higher up in the power hierarchy should boost confidence. However, because status is rooted in the subjective evaluations of others, the threat of status loss should become more acute, as in the middle of the status hierarchy versus in the high and low positions (any deviation risks criticism and rejection). Demonstrating these differential effects of status and power would further strengthen the argument that threat of loss underlies the relationship between status and creativity. Method Participants and design. One hundred forty-eight participants (mean age 19.96 years; females 47%) took part in the study, which consisted of a 3 (power: high vs. middle vs. low) 2 (evaluation: yes vs. no) between-participants design. Participants were undergraduate students who were given course credit for taking part in the study. Procedure. The procedures of the study were the same as Study 1, with one exception. We manipulated the psychological experience of power instead of status by using an adapted version of Galinsky and colleagues’ (2003) power prime. Participants recalled situations in which they had high, middle, or low power. Specifically, participants read, Power determines the extent to which people control the ability of another person or persons to get something they want, or are in a position to evaluate those individuals. Please recall a particular incident in which you were part of a group and in that group your power relative to others was “HIGH,” that is at the top of the power hierarchy/ around the “MIDDLE” that is, neither the top nor the bottom of the power hierarchy/”LOW,” that is at the bottom of the power hierarchy. Please describe this situation in which you had high/middle/low power – what happened, how you felt, and so forth Results Manipulation checks. Power. Two coders, blind to both the experimental condition and hypotheses, categorized the priming essays as high, middle, low power, or whether they could not determine which category the essay belonged to. The coders’ classifications of the essays were 100% consistent with the conditions. As an additional power manipulation check, participants reported to what extent they felt influential, independent, powerful, unimportant (reverse-coded), and subordinate (reverse-coded; Duguid & Goncalo, 2012). Endpoints were 1 (very little) and 9 (a great deal), .87. ANOVA revealed a significant main effect of power, F(2, 142) 56.12, p .001, p 0.007. Participants in the high-power condition (M 5.69, SD 0.55) felt more powerful than those in the middle-power (M 4.06, SD 0.43), t(96) 16.18, p .001, and low-power (M 2.35, SD .53), t(98) 30.96, p .001, conditions. Moreover, participants in the middlepower condition felt more powerful than those in the low-power condition, t(96) 17.59, p .001. There was no significant main effect of evaluation, F(1, 142) 0.01, p .927, p 0.007, or interaction between status and evaluation, F(2, 142) 1.99, p .140, p 0.007. In order to determine whether the power manipulation also influenced individuals’ perceptions of status, we used the manipulation check for status used in Study 2, .92. ANOVA revealed no main effect of power, F(2, 142) 0.44, p .644, p 0.006, evaluation, F(1, 142) 0.88, p .350, p 0.006, or interaction between status and evaluation, F(2, 142) 0.02, p .977, p 0.001. Creativity. Number of ideas generated. As expected, ANOVA revealed a significant main effect of power, F(2, 142) 16.67, p .001, p 0.190. Participants in the high-power condition (M 16.76, SD 9.05) generated more ideas than those in the middle-power (M 13.24, SD 5.36), t(96) 2.34, p .022, and low-power (M 9.88, SD 6.83), t(97) 4.27, p .001, conditions. Participants in the middle-power condition also generated more ideas than those in the low-power condition, t(97) 2.73, p .008. There was also a significant main effect of evaluation, F(1, 142) 62.21, p .001, p 0.305. When ideas were being evaluated, participants generated more ideas (M 17.19, SD 8.81) than when ideas were not being evaluated (M 9.45, SD 3.60). There was no significant interaction of power and evaluation, F(2, 142) 2.90, p .060, p 0.039. Idea novelty. ANOVA also revealed a significant main effect of power for idea novelty, F(2, 142) 133.91, p .001, p 0.654. Participants in the high-power condition (M 3.22, SD 0.55) generated more novel ideas than those in the middle-power (M 2.87, SD 0.55), t(96) 3.19, p .002, and low-power This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. STATUS AND CREATIVITY 9
10 DUGUID AND GONCALO Alhough most of the does not take in novel ideas than th se in the low-power con dition,97=12.17 in sor ces.individuals may 048.0027 e that stable and oaluated rticipants ated more novel were not being e 903s and Given that we found that middle-status individuals Discussion The pesults revealed that individuals who exn nced high powe hen the middle ced low t in th others (An nd who are ut their s ding in the stat ecaus moving from nay no ary of g d about and e ing from a middle-status hey bec e less co mps &ckerman.2001). ce,when the status even at the middle of the hier solutions thn the topor whereas increasing powe y the th that the threat of loss.We test these predictions in the following study. Method Study Participants and design. We have theorized that middle- sindividuals are concemed ants (mean age=1.6 years:females45.3%)took part in the n status Tand th at mak es individuals more focus whic high vs.mi c vs.low) this the undergraduate students who were paid $10for stud of tho ith midd th as thos ant role in our theorizing becase upcoming task,they could never )move down nth d.in the real world,status hi dcdqpcstion related to demographic s to Results hether the hierarchy is stable (you can never move down)or Manipulation checks. role in the 4749.p<001. Picsidcntrocrcthihcrtatsw tatus as the xtent to which one perceives that an ate ative stat 123)= 4.79.p<.001.or Assi tant role (M 293 SD position is likely to berealiz In stable 122 08P≤.001 001. nstable hierarchies may believe that changing status rank ain effect of status stability,F(181)=0.45
(M 1.52, SD 0.56), t(97) 15.37, p .001, conditions. Participants in the middle-power condition also generated more novel ideas than those in the low-power condition, t(97) 12.17, p .001. In addition, there was a significant main effect of evaluation, F(1, 142) 3.96, p .048, p 0.027. When ideas were being evaluated, participants generated more novel ideas (M 2.62, SD 0.91) than when ideas were not being evaluated (M 2.44, SD 0.92). However, there was no a significant interaction of power and evaluation, F(2, 142) 2.90, p .916, p 0.001. Discussion The results revealed that individuals who experienced high power generated more ideas and more novel ideas than those who experienced middle or low power. Likewise, individuals who experienced middle power were more creative than those who experienced low power. Given that status is granted by others (Anderson & Kilduff, 2009; Fiske, 2010; Fiske & Berdahl, 2007), individuals with middle status who are concerned about their standing in the status hierarchy may be wary of going against convention for fear that they will be evaluated negatively and lose status. However, as individuals’ power increases, they become less concerned about and less constrained by others’ opinions and judgments (Galinsky et al., 2006; Lee-Chai et al., 2001), and therefore they will generate more creative ideas. Thus, whereas increasing power, even at the middle of the hierarchy, may help mitigate feelings of threat, moving from low to middle status seems to exacerbate feelings of threat. Building upon Studies 2 and 3, these results lend further support to the argument that the threat of status loss may be driving middle-status insecurity. Study 5 We have theorized that middle-status individuals are concerned with status loss and this threat makes individuals more focused but less creative. The results from our previous studies are consistent with this theoretical account—introducing the possibility of evaluation exacerbated perceived threat, which, in turn, stifled the creativity of those with middle status. One limitation of our previous studies is that we did not vary the stability of the status hierarchy. This limitation is important for two reasons. First, the stability of the status hierarchy plays an important role in our theorizing because we suggest that the threat of status loss may occur when individuals can possibly move further down the hierarchy (i.e., when the hierarchy is potentially unstable rather than fixed). Second, in the real world, status hierarchies may vary in the extent to which they are stable or unstable. Thus, it is important to vary this dimension of status to specify more precisely the scope conditions of our theory. In Study 5, we manipulated the threat of status loss directly by systematically varying whether the hierarchy is stable (you can never move down) or unstable (you can move down). The stability of a status hierarchy may play a role in the behaviors of social actors who are in high, middle, and low positions. Tajfel and Turner (1979, 1986) defined the stability of status as the extent to which one perceives that an alternative status position is likely to be realized. In stable hierarchies, individuals may assume that their current status rank is constant, whereas those in unstable hierarchies may believe that changing status rank is possible. Although most of the literature on status does not take into account the stability of the hierarchy but rather conceptualizes status as a static construct, in some instances, individuals may perceive that the status hierarchy is stable and unlikely to be altered, whereas in other instances, they may perceive that there is the possibility to change their status position (Jordan, Sivanathan, & Galinsky, 2011). There is an influential body of literature that shows that individuals’ beliefs about the stability of the status structure affects individuals’ perception, attitudes, decisions, and behaviors (e.g., Ellemers, Wilke, & van Knippenberg, 1993). Given that we found that middle-status individuals were more susceptible to the threat of status loss when being evaluated than high- and low-status individuals, we would expect those with middle status to be much more conservative in the number and novelty of ideas they express when the status hierarchy is unstable and there is the possibility of moving down in rank. We suggest that high-status individuals may be more confident in their social acceptance and, hence, assume they have more leeway to take risk. Low-status individuals, on the other hand, may think they have less to lose because moving from “low” to “lower” may not represent as meaningful or significant a change in status position as moving from a middle-status position to a low-status position (Phillips & Zuckerman, 2001). Hence, when the status hierarchy is unstable, individuals with middle status may be less likely to suggest creative solutions than individuals at the top or the bottom of the status hierarchy. Conversely, when the status hierarchy is stable, the middle status may not be stifled by the threat of status loss. We test these predictions in the following study. Method Participants and design. One hundred eighty-seven participants (mean age 19.06 years; females 45.3%) took part in the study, which consisted of a 3 (status: high vs. middle vs. low) 2 (status stability: stable vs. unstable) between-participants design. Participants were undergraduate students who were paid $10 for taking part in the study. Procedure. The cover story and status manipulation were the same as those used in Study 2. In order to manipulate the stability of the status hierarchy, after the status manipulation, participants were told that (regardless of/depending on) their performance in the upcoming task, they could (never/always) move down in the status hierarchy. The upcoming task was the same creativity task used in Study 2. At the end of the study, participants completed a short survey, which included questions related to demographic information and the manipulation checks. Results Manipulation checks. Status. ANOVA revealed a main effect of status, F(2, 181) 47.49, p .001, p 0.344. Participants randomly assigned to the President role felt higher status (M 4.99, SD 1.30) than those assigned to the Middle Manager role (M 4.00, SD 0.98), t(123) 4.79, p .001, or Assistant role (M 2.93, SD 1.22), t(122) 9.08, p .001. Additionally, Middle Managers felt higher status than Assistants, t(123) 5.40, p .001. There was no significant main effect of status stability, F(2, 181) 0.45, p .503, p 0.002, nor an interaction between status and status This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 10 DUGUID AND GONCALO
