the necessity of others is the mother of invention

7/30/2019 The Necessity of Others is the Mother of Invention

1/25

THE NECESSITY OF OTHERS IS THE MOTHEROF INVENTION: INTRINSIC AND PROSOCIAL MOTIVATIONS,

PERSPECTIVE TAKING, AND CREATIVITY

ADAM M. GRANT

University of Pennsylvania

JAMES W. BERRYUniversity of North Carolina at Chapel Hill

Although many scholars believe that intrinsic motivation fuels creativity, research hasreturned equivocal results. Drawing on motivated information processing theory, wepropose that the relationship between intrinsic motivation and creativity is enhanced

by other-focused psychological processes. Perspective taking, as generated by proso-cial motivation, encourages employees to develop ideas that are useful as well asnovel. In three studies, using both field and lab data, we found that prosocial motiva-tion strengthened the association between intrinsic motivation and independent cre-ativity ratings. In our second and third studies, perspective taking mediated this

moderating effect. We discuss theoretical implications for creativity and motivation.

As work becomes increasingly dynamic, uncer-tain, and knowledge-based, organizations dependon creative ideas from employees (George, 2007).Scholars and practitioners share a strong interest inunderstanding the psychological forces that moti-vate creativitythe production of ideas that are

both novel and useful (Amabile, 1996). For severaldecades, researchers have believed that intrinsicmotivation is an important driver of creativity (Els-

bach & Hargadon, 2006). When intrinsically moti-vated, employees expend effort based on interest,curiosity, and a desire to learn (Ryan & Deci, 2000).Intrinsic motivation is thought to enhance creativ-ity by increasing positive affect, cognitive flexibil-ity, risk taking, and persistence (Shalley, Zhou, &Oldham, 2004).

However, the empirical evidence linking intrin-sic motivation to creativity is equivocal (George,2007; Shalley et al., 2004). Some studies have dem-onstrated that intrinsic motivation is associatedwith higher levels of creativity (e.g., Amabile, 1985;Amabile, Hill, Hennessey, & Tighe, 1994), whereas

others have shown weak or nonsignificant associa-

tions (e.g., Dewett, 2007; Perry-Smith, 2006; Shal-ley & Perry-Smith, 2001). In light of these conflict-ing findings, George (2007: 445) observed thatrather than assume that intrinsic motivation un-derlies creativity, researchers need to tackle thistheorized linkage more directly and in moredepth. Organizational scholars need new theoret-ical perspectives and empirical investigations todeepen knowledge of the motivational processes

that drive creativity (Shalley et al., 2004).Our objective in this article is to explain andresolve the inconsistent relationship between in-trinsic motivation and creativity. To do so, we drawon motivated information processing theory fromsocial psychology, which offers a promising con-ceptual framework for both explaining and resolv-ing the inconsistency. The core premise of moti-vated information processing theory is thatmotivations shape cognitive processing: employeesselectively notice, encode, and retain informationthat is consistent with their desires (Kunda, 1990;Nickerson, 1998). Thus, when employees are in-trinsically motivated, their desires to learn, exploretheir interests, and engage their curiosity will leadthem to focus on novel ideas. However, to producecreative ideas, employees also need to attend tousefulness. We propose that since ideas are ulti-mately most useful when they solve problems forother people inside and outside an organization, afocus on usefulness can be engendered by perspec-tive taking. When employees take others perspec-tives, they are more likely to develop ideas that areuseful to others (Mohrman, Gibson, & Mohrman,

For novel and useful feedback, we thank ElizabethMorrison, David Mayer, Ron Piccolo, and Manuela Pri-esemuth. For enlightening discussions, we thank TeresaAmabile, Sigal Barsade, and Jen Mueller. For assistancewith data collection and editing, we thank FrancescaGino, Ted Henifin, Chris Myers, Maggie OBrien, andMarty Whelan.

This article was accepted before the first author be-came an associate editor of this journal.

Academy of Management Journal

2011, Vol. 54, No. 1, 7396.

73Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holders express

written permission. Users may print, download or email articles for individual use only.


2/25

2001). Recent developments in motivated informa-tion processing theory suggest that employees arelikely to attend to usefulness when they experienceprosocial motivationthe desire to benefit oth-erswhich encourages them to consider othersperspectives (De Dreu, Weingart, & Kwon, 2000).We thus hypothesize that when guided by proso-

cial motivation to take others perspectives, em-ployees will channel their intrinsic motivation to-ward producing ideas that are not only novel, butalso useful, thereby achieving higher creativity. Wetested these hypotheses in two field studies and alaboratory experiment.

Our theoretical perspective and empirical find-ings offer important contributions to knowledgeabout creativity and motivation in work organiza-tions. Our research answers calls to identify mod-erators of the relationship between intrinsic moti-vation and creativity (George, 2007; Shalley et al.,2004) by accentuating the importance of other-fo-cused psychological processes in creativity. Ourmotivated information processing viewpoint re-veals how perspective taking, fueled by prosocialmotivation, leads employees to channel their in-trinsic motivation toward ideas that are useful aswell as novel. Our research also has two centraltheoretical implications beyond those of existingresearch on the interaction of intrinsic and proso-cial motivations (Grant, 2008). First, we identifyperspective taking as a key mechanism throughwhich prosocial motivation strengthens the impactof intrinsic motivation on creativity, addressing

Grants (2008) call to build and test theory aboutthe psychological processes that explain this inter-action. Second, in doing so, we introduce perspec-tive taking as an important influence on creativity,showing how a focus on others can encourage em-ployees to direct their intrinsic motivation towardthe generation of creative ideas. Our researchshows how perspective taking interacts with intrin-sic motivation to enhance creativity.

MOTIVATION AND CREATIVITY

Our emphasis in this article is on understandingthe conditions under which intrinsic motivationpromotes creativity. Intrinsic motivation refers tothe desire to expend effort based on interest in andenjoyment of the work that is being performed(Amabile, 1996; Ryan & Deci, 2000). As notedabove, psychologists and organizational scholarshave long believed that intrinsic motivation is animportant enabler of creativity. Researchers haveidentified three interrelated psychological mecha-nisms through which intrinsic motivation maystimulate creativity. First, emotion theorists have

proposed that when employees are intrinsicallymotivated, they experience positive affect (e.g., Sil-via, 2008). This stimulates creativity by broadeningthe range of cognitive information available, ex-panding the scope of attention toward assimilatinga wider set of ideas, and encouraging cognitiveflexibility for identifying patterns and associations

between ideas (e.g., Amabile, Barsade, Mueller, &Staw, 2005; Fredrickson, 1998). Second, self-deter-mination theorists have proposed that when em-ployees are intrinsically motivated, their curiosityand interest in learning will enhance their cogni-tive flexibility, willingness to take risks, and open-ness to complexity, which in turn will expand theiraccess to ideas and potential solutions (Gagne &Deci, 2005; see also Amabile, 1979, 1996).

Third, both emotion and self-determination the-orists suggest that intrinsic motivation promotescreativity by encouraging persistence. From thestandpoint of emotion theories, by fostering posi-tive affect, intrinsic motivation enhances psycho-logical engagement and builds energy for sustain-ing effort, increasing the amount of time thatemployees are willing and able to work on theirtasks (e.g., Fredrickson, 1998). From the standpointof self-determination theory, by fostering confi-dence and interest, intrinsic motivation encouragesemployees to persist with challenging, complex,unfamiliar tasks (Gagne & Deci, 2005), as well as toconcentrate their attention more effectively onthese tasks (e.g., Amabile, 1996).

Conflicting Results

However, empirical research has yielded mixedand often confusing results about whether intrinsicmotivation enhances creativity (for reviews, seeAmabile and Mueller [2007], George [2007], andShalley et al. [2004]). Most of this research hasinvolved laboratory experiments with children andcollege students developing artistic products, inwhich it is possible to manipulate intrinsic moti-vation and obtain independent expert ratings ofcreative outcomes. A number of laboratory experi-

ments have shown that when participants are in-duced to experience high levels of intrinsic moti-vation, their products are rated as more creative(e.g., Amabile, 1979; Koestner, Ryan, Bernieri, &Holt, 1984). However, other laboratory experimentshave shown weak, mixed, or no benefits of intrinsicmotivation for creativity (e.g., Amabile, 1985; Am-abile, Hennessey, & Grossman, 1986; Eisenberger &Aselage, 2009; Shalley & Perry-Smith, 2001). Forexample, Amabile et al. (1986: 21) stated that sev-eral laboratory studies do not allow definitive con-clusions . . . only some of the various intrinsic in-

74 FebruaryAcademy of Management Journal


3/25

terest measures showed correlations with creativity.They further noted these studies are vulnerable tothe possibility of reverse causality. Rather thancausing creativity, self-reports of intrinsic motiva-tion might result from a greater enjoyment andsatisfaction experienced in expressing creativity(Amabile et al., 1986: 21). Table 1 summarizes key

details of the pertinent laboratory studies, showingthat intrinsic motivation has been more consis-tently linked to creativity in artwork and writingtasks than to creativity in producing ideas and so-lutions relevant to business problems.

The results of field studies have also been equiv-ocal. Studies with college students have shown thatintrinsic motivation is correlated with higher ob-server ratings of creativity (Amabile et al., 1994),

but studies with working adults have returned con-flicting results, with intrinsically motivated em-ployees being rated as more creative in some sam-ples and tasks, but not in others (Amabile et al.,1994; Dewett, 2007; Eisenberger & Aselage, 2009;Eisenberger & Rhoades, 2001; Janssen & vanYperen, 2004; Perry-Smith, 2006; Shin & Zhou,2003; Tierney, Farmer, & Graen, 1999). Table 1provides further information on these field studies,suggesting that intrinsic motivation more consis-tently predicts self-reports of creativity than ob-server ratings or archival measures of creativity.

Taken together, these laboratory and field studiessuggest a variable relationship between intrinsicmotivation and creativity, and this variability hasreceived surprisingly little theoretical and empiri-

cal attention. In recent reviews of the creativityliterature, organizational scholars have called fornew conceptual frameworks and studies to investi-gate the conditions under which intrinsic motiva-tion is more and less likely to fuel creativity (Am-abile & Mueller, 2007; George, 2007; Shalley et al.,2004). Our focus is on answering these calls toexamine contingencies that moderate the impact ofintrinsic motivation on creativity.

Existing research provides clues about why in-trinsic motivation does not guarantee that employ-ees will ultimately produce ideas that are novel and

useful (Amabile, 1996). Several studies have shownthat novelty and usefulness are independent, andtypically orthogonal, dimensions of creativity. AsLitchfield (2008: 659) summarized, Creativity isgenerally treated as a composite of novelty andutility . . . but research has yet to carefully examinethe effects of these dual goals . . . novelty and use-fulness are unrelated dimensions of ideas. Fur-thermore, Ford and Gioia (2000: 727) demonstratedthat different factors independently influencenovelty and usefulness. The studies summarized inTable 1 appeared to more consistently link intrinsic

motivation with creativity in tasks that focused pri-marily on novelty and originalitysuch as creatingartwork and writing poemsthan in tasks withstronger usefulness components, as are common inresearch and development jobs. This evidence pro-vides reason to believe that intrinsic motivationdrives the production of novel, but not necessarily

useful, ideas.Why might intrinsic motivation promote a stron-

ger focus on novelty than on usefulness? Theconceptual framework of motivated informationprocessing theory provides a parsimonious expla-nation. As noted previously, motivated informationprocessing is a pervasive human tendency to selec-tively perceive, encode, and retain information thatis congruent with ones desires (for reviews, seeHeath, Larrick, and Klayman [1998]; Kunda [1990];and Nickerson [1998]). When employees are intrin-sically motivated, they experience a desire to learn,

pursue their interests, and explore their curiosities(Ryan & Deci, 2000). Thus, intrinsic motivation islikely to encourage employees to focus primarilyon ideas that are novel, original, and unique, whichprovide the greatest opportunities for learning andexploration. Indeed, research has shown that in-trinsically motivated employees use interest as aguide for determining which ideas to pursue(Amabile et al., 1994; Ryan & Deci, 2000) and thatinterest primarily directs attention to ideas that arenovel (Silvia, 2008). When intrinsically motivated,employees engage in exploration (Ryan & Deci,

2000), becoming psychologically absorbed in theprocess of working on their tasks (Amabile et al.,1994; Gagne & Deci, 2005) and often viewing thedevelopment of novel ideas as an end in and ofitself (Csikszentmihalyi, 1996). Summarizing threedecades of self-determination research, Ryan andDeci (2000: 71) stated that intrinsic motivation fo-cuses attention on activities that have the appealof novelty.

Intrinsic motivation thus promotes a focus onseeking out new discoveries, which is associatedwith developing products that are higher in origi-nality but not in aspects related to usefulness, suchas technical quality (Csikszentmihalyi & Getzels,1971). Further, classic psychological research hasshown that many intrinsically motivated architectshad difficulty producing creative ideas becausethey were focused on the novelty of their designs

but not necessarily concerned with their usefulness(Barron, 1963; for a recent discussion, see Little[2006]). Reviewing a growing body of research oninterest, Silvia (2008: 58) further explained thatinterest attracts people to new, unfamiliar things,and many of these things will turn out to be trivial.

2011 75Grant and Berry


4/25

TABLE 1Summary of Studies Linking Intrinsic Motivation and Creativity

Study and Context Relevant Manipulations and Measures Results

Laboratory study with femalecollege students makingcollages (Amabile, 1979)

Manipulated expected evaluation (yes/no); measuredintrinsic motivation (self-report) and creativity(artist ratings).

Participants who did not expect evaluationreported higher intrinsic motivation andproduced artwork judged as morecreative.

Laboratory study with 1stand 2nd grade childrenpainting (Koestner et al.,1984)

Manipulated external limits; measured intrinsicmotivation (self-report, task persistence) andcreativity (judge ratings).

Controlling limits decreased both measuresof intrinsic motivation and bothmeasures of creativity.

Laboratory study with youngadults who self-identifiedas creative writers asked towrite two poems (Amabile,1985)

First poem served as baseline; manipulated attentionto reasons for writing (intrinsic, extrinsic, none);measured creativity (poet ratings).

Concentrating on extrinsic reasons led tolower creativity ratings on the secondpoem, but intrinsic reasons did notincrease creativity relative to the control(no reasons).

Laboratory study withelementary school childrenmaking collages, writing

stories, and solvingpuzzles (Study 1, Amabileet al., 1986)

Manipulated task rewards; measured intrinsicmotivation (self-report and behavioral choice tospend free time on the task a week later) and

creativity (teacher ratings).

Self-reports of intrinsic motivation werenot significantly correlated withcreativity for collages, stories, or

puzzles; behavioral choice correlatedpositively with creativity in stories butnot collages or puzzles.

Laboratory study withcollege students makingcollages (Study 3, Amabileet al., 1986)

Manipulated task rewards; measured intrinsicmotivation (self-reports of enjoyment, satisfaction,interest, and motivation) and creativity (artistratings).

Creativity was significantly related toenjoyment and satisfaction, but notinterest or motivation.

Laboratory study withcollege students generatingsolutions to two businessproblems (Shalley & Perry-Smith, 2001)

Manipulated expectations of external evaluation(controlling, informational); measured intrinsicmotivation (self-report) and creativity (expertjudges).

Informational evaluation increasedintrinsic motivation and creativity, butintrinsic motivation was notsignificantly associated with creativity.

Laboratory study withcollege students suggestingcreative titles for a shortstory (Study 3, Eisenberger& Aselage, 2009)

Manipulated rewards for creativity; measuredintrinsic interest (self-report) and creativity(research assistant ratings).

Reward increased intrinsic interest andcreativity, but intrinsic interest was notsignificantly associated with creativity.

Field study of employees indiverse jobs (Study 2,Eisenberger & Aselage,2009)

Employees reported intrinsic interest; supervisorsrated creativity.

Significant, positive correlation betweenintrinsic interest and creativity.

Field study of employees ina sales organization (Study4, Eisenberger & Rhoades,2001)

Employees reported intrinsic task interest;supervisors rated creativity.

Significant, positive correlation betweenintrinsic task interest and creativity.

Field study of multiplesamples of students andworking adults (Amabile etal., 1994)

Employees reported intrinsic motivationalorientation and creativity; independent ratings ofcreativity on various tasks.

Intrinsic motivation was significantlyassociated with creativity in some tasksbut not others.

Field study of R&D personnel(Dewett, 2007)

Employees reported intrinsic motivation andcreativity; supervisors rated creativity.

Intrinsic motivation predicted self-reportsbut not all supervisor ratings ofcreativity.

Field study in Dutch energysupplier (Janssen & vanYperen, 2004)

Employees reported intrinsic interest (masteryorientation); supervisors rated creative/innovativeperformance.

Intrinsic interest predicted supervisorratings of creativity, but not after leader-member exchange was controlled for.



5/25

This evidence suggests, as predicted by moti-vated information processing theory, that bydrawing attention to interest, intrinsic motivationcultivates a primary focus on novelty but not nec-essarily on usefulness. Indeed, Amabile (1996: 118)proposed that early stages of the creative process,where the novelty of the outcome is importantlydetermined, may require intrinsic motivation, butin advancing toward stages where the novelty ofthe work (though still important) is less crucial,other motivators may serve to focus and energizethe individual toward getting the job done in anappropriate way. This proposition raises the pos-sibility that psychological processes that draw at-tention to usefulness may enable employees tochannel the novel ideas prompted by intrinsic mo-tivation toward creativity.

The Impact of Other-FocusedPsychological Processes

We propose that other-focused psychologicalprocesses play an important role in guiding em-ployees toward considering ideas that are not onlynovel, but also useful. Ideas that are maximallyuseful are those that are applicable to addressingthe problems or needs of a wide range of coworkers,supervisors, customers, and clients (Mohrman etal., 2001). Thus, we expect that when employeesfocus their attention on others, they will be more

likely to develop ideas that are ultimately useful toothers. To explain how other-focused psychologi-cal processes channel intrinsic motivation towardideas that are useful as well as novel, we build onmotivated information processing theory, accord-ing to which employees desires shape how theyattend to information (e.g., Kunda, 1990). We firstexplain how prosocial motivation encourages em-ployees to attend to information about others per-spectives. We then articulate how this ensuing pro-cess of perspective taking, the central cognitiveprocess in our conceptual model, enables intrinsi-

cally motivated employees to develop ideas that areuseful as well as novel.

The moderating effect of prosocial motivation.Motivated information processing theory suggeststhat to take others perspectives and determinewhat they find useful, employees need to have adesire to do so (Kunda, 1990; see also Caruso, Ep-ley, & Bazerman, 2006). We propose that prosocialmotivation is an other-focused psychological pro-cess that directs employees attention toward oth-ers perspectives on what is useful, enhancing theimpact of intrinsic motivation on creativity. Pro-social motivation is the desire to expend effort

based on a concern for helping or contributing toother people (Grant, 2007). In contrast with tradi-tional assumptions that prosocial and self-inter-ested motivations involve mutually exclusive oropposing desires (e.g., Batson, 1998; Meglino &

Korsgaard, 2004; Schwartz & Bardi, 2001), recentresearch has indicated that prosocial and self-inter-ested motivations are empirically independent andcan even be positively related (De Dreu & Nauta,2009). This is because prosocial motivation canserve multiple goals. For example, employees candesire to help others because they care about them,

because they feel that it is the right thing to do,because they wish to maintain membership in avalued group, and/or because doing so will makethem feel good about themselves (Batson, Ahmad,Powell, & Stocks, 2008). Thus, prosocial motivation

can involve, but should not necessarily be equatedwith, altruism; it refers to a concern for others, nota concern for others at the expense of self-interest(De Dreu, 2006).

Although prosocial motivation can be directedtoward different beneficiaries, such as coworkers,supervisors, clients, or customers, research hasshown that a more general form of prosocial moti-vation directed toward benefiting others in oneswork context is associated with higher job perfor-mance, personal initiative, and organizational citi-zenship behaviors (De Dreu & Nauta, 2009). Pro-

TABLE 1(Continued)

Study and Context Relevant Manipulations and Measures Results

Field study of scientists(Perry-Smith, 2006)

Employees reported intrinsic motivation; supervisorsrated creativity.

Intrinsic motivation was not significantlyassociated with creativity.

Field study in R&D (Shin &

Zhou, 2003)

Employees reported intrinsic motivation; supervisors

rated creativity.

Intrinsic motivation predicted higher

creativity.

Field study of R&Demployees in a chemicalcompany (Tierney et al.,1999)

Employees reported intrinsic motivationalorientation; creativity was measured withsupervisor ratings, invention disclosure forms, andcreative research reports.

Intrinsic motivational orientation predictedsupervisor ratings, but not the number ofcreative reports, and inconsistentlypredicted creativity on invention forms.



6/25

social motivation can thus be understood as apsychological state in which employees are fo-cused on the goal of benefiting other people (Bat-son, 1998; De Dreu, 2006; Grant, 2007). Prosocialmotivation can be distinguished from intrinsic mo-tivation along three dimensions: self-regulation,goal directedness, and temporal focus (Grant,

2008). In terms of self-regulation, intrinsic motiva-tion involves fully autonomous self-regulation,whereas prosocial motivation is based on other-oriented values that can be internalized to varyingdegrees. In terms of goal directedness and temporalfocus, intrinsic motivation involves a primarilytask-focused emphasis on the process of complet-ing work in the present, whereas prosocial motiva-tion involves a primarily other-focused emphasison producing beneficial outcomes in the future.

Why would prosocial motivation enhance theeffect of intrinsic motivation on creativity? Proso-cial motivation provides employees with the mean-ingful outcome goal of helping others (Batson,1998; Grant, 2007). Therefore, in the context of ideageneration, prosocially motivated employees will

be driven to develop ideas that are useful to thecoworkers, supervisors, clients, or customers who

benefit from their efforts. Indeed, psychological re-search suggests that employees with high prosocialmotivation are driven to produce ideas that are

useful to future generations (McAdams & de St.Aubin, 1992). As such, prosocial motivation mayenable employees to channel their intrinsic moti-vation toward producing ideas that are not onlynovel, but also useful to others. Although intrinsicmotivation offers positive affect and cognitive flex-ibility to help employees generate novel ideas,

prosocial motivation offers an important other-fo-cused outcome goal to help employees focus ontheir most useful and relevant ideas. Illustratingthis synergy, Simonton (1989) found that classicalcomposers, who typically report high intrinsic mo-tivation for writing music, tend to produce theirmost creative, aesthetically significant works whenthey are prosocially motivated to leave behind greatfinal pieces for their audiences. Thus, we predictthat prosocial motivation will enhance the impactof intrinsic motivation on creativity.

Hypothesis 1. Prosocial motivation strengthens

the association between intrinsic motivationand creativity.

The role of perspective taking. To shed light onthe core psychological process underlying this hy-pothesis, we propose that perspective taking willmediate the moderating effect of prosocial motiva-tion on the relationship between intrinsic motiva-tion and creativity. As depicted in Figure 1, in our

FIGURE 1Summary of Hypothesesa

a We propose and depict a direct moderating effect of prosocial motivation in keeping withclassic and contemporary models of mediation, which suggest thatspecial cases such as sup-pression excludedmediation begins with a direct effect that researchers then seek to explainthrough one or more intervening variables (Edwards & Lambert, 2007; Mathieu & Taylor, 2006). Weexpect full mediation such that the moderating effect of prosocial motivation is eliminated whenperspective taking is incorporated into the model.



7/25

mediated moderation model perspective takingasdriven by prosocial motivationenhances the ef-fect of intrinsic motivation on creativity. We firstexplain why prosocial motivation encourages per-spective taking and then examine why perspectivetaking strengthens the impact of intrinsic motiva-tion on creativity.

Perspective taking is a cognitive process inwhich individuals adopt others viewpoints in anattempt to understand their preferences, values,and needs (Parker & Axtell, 2001). Although em-ployees can vary in their dispositional tendenciesto take the perspectives of others, research hasshown that efforts to take the perspectives of othersin specific situations and contexts vary as a func-tion of employees motivations (e.g., Batson, Early,& Salvarani, 1997; De Dreu et al., 2000; Galinsky,Magee, Inesi, & Gruenfeld, 2006). We predict thatprosocial motivation will encourage employees toengage in perspective taking. Motivated informa-tion processing theory (De Dreu, 2006) provides theconceptual basis for this relationship. Recent ad-vances in motivated information processing re-search have revealed how prosocial motivationshapes the cues that individuals selectively notice,encode, and retain (De Dreu et al., 2000). Whenemployees are prosocially motivated, their desiresto benefit others lead them to pay heightened atten-tion to others perspectives in order to identifyways to help them effectively (De Dreu et al., 2000).As De Dreu (2006: 1248) explained, prosocial mo-tivation leads employees to consider information

from multiple perspectives to a greater extent . . .[and] stimulates the processing of social informa-tioninformation from and about relevant others.

Indeed, extensive research in both psychologyand management has shown that prosocially moti-vated individuals are more likely to adopt the per-spectives of a range of other people, including co-workers, supervisors, suppliers, customers, andclients (Axtell, Parker, Holman, & Totterdell, 2007;Batson, 1998; De Dreu et al., 2000; Parker & Axtell,2001). Because prosocially motivated employeesare more aware of, and concerned about, other peo-

ples goals and preferences (Meglino & Korsgaard,2004), they ask questions, listen carefully, and ob-serve behaviors in order to understand what othersvalue, obtaining cues about how to provide helpeffectively (De Dreu et al., 2000). Thus, we predictthat prosocially motivated employees will be morelikely to take others perspectives.

Hypothesis 2a. Prosocial motivation is posi-tively associated with perspective taking.

We further propose that perspective taking, inturn, will strengthen the effect of intrinsic motiva-

tion on creativity. We base this hypothesis on the-ory and research on creative cognitive processingand motivated information processing. The litera-ture on creative cognitive processing shows thatafter generating possibilities, employees progresspsychologically through phases of testing, validat-ing, and refining different ideas (Amabile & Muel-

ler, 2007). In these phases, because novel possibil-ities have already been generated, novelty is lessimportant than converging on ideas that are useful,relevant, and appropriate (Amabile, 1996). Totranslate novel possibilities into creative ideas, em-ployees need to filter out those that are least usefuland retain those that are most useful (Campbell,1960; Litchfield, 2008; Simonton, 2003). After all,many novel ideas end up being trivial or impracti-cal (Silvia, 2008).

Focusing attention on the perspectives of otherswill provide employees with a standard for deter-mining which ideas should be selected as usefulversus discarded as less useful. According to moti-vated information processing theory and research,when employees take the perspectives of others,they are more likely to think in an integrative fash-ion to consolidate and align these perspectives (DeDreu et al., 2000). As employees consider morenumerous and diverse perspectives, they gain adeeper understanding of which ideas differentgroups of beneficiaries, constituents, or stakehold-ers are most likely to consensually consider useful(Amabile, 1996). Indeed, Mohrman et al. (2001)found that when academic researchers took the per-

spectives of practitioners, they succeeded in con-ducting research that practitioners judged as moreuseful. Furthermore, taking others perspectivescan enable employees to identify useful applica-tions of novel but otherwise impractical ideas.Studies of product development teams have shownthat when employees adopt the perspectives of co-workers and customers, they are more capable oftranslating their novel ideas into useful products(Dougherty, 1992; Purser, Pasmore, & Tenkasi,1992).

In summary, theories of creative cognitive pro-

cessing and motivated information processingsuggest that by engaging in perspective taking, em-ployees obtain a clearer, more integrative under-standing of what types of ideas will be useful to thecoworkers, supervisors, clients, customers, andother stakeholders who evaluate and benefit fromtheir work. Intrinsic motivation provides employ-ees with access to novel ideas, and perspectivetaking provides them with a filter for determiningwhich of these ideas to develop and how to elabo-rate them in useful ways (Boland & Tenkasi, 1995;Dougherty, 1992). For example, when an intrinsi-



8/25

cally motivated product development team mem-ber generates novel possibilities, taking a custom-ers perspective is likely to focus her/his attentionand energy on further developing the possibilitiesthat are most useful for solving the customersproblems. As Sethi and Nicholson (2001: 159) ex-plained, awareness of customers needs can en-

hance members commitment to strive for superioroutcomes that can better satisfy these needs, in-creasing the probability of creative ideas. We thusexpect that perspective taking will enhance the ef-fect of intrinsic motivation on creativity.

Hypothesis 2b. Perspective taking strengthensthe association between intrinsic motivationand creativity.

Our preceding two hypotheses propose thatprosocial motivation increases perspective taking(Hypothesis 2a) and that perspective taking en-

hances the effect of intrinsic motivation on creativ-ity (Hypothesis 2b). Together, these two hypothesespredict that perspective taking mediates the mod-erating effect of prosocial motivation on the rela-tionship between intrinsic motivation and creativ-ity (Hypothesis 1), constituting a case of mediatedmoderation (Edwards & Lambert, 2007). Althoughmediated moderation can take multiple forms, thetype of mediated moderation that we expect ispresent when (1) a variable (prosocial motivation)moderates the relationship between an indepen-dent variable (intrinsic motivation) and a depen-

dent variable (creativity), as in Hypothesis 1; (2)the moderating variable (prosocial motivation)causes a mediating variable (perspective taking),as in Hypothesis 2a; and (3) the mediating variable(perspective taking) moderates the relationship be-tween an independent variable (intrinsic motiva-tion) and a dependent variable (creativity), as inHypothesis 2b, thereby transmittingand eliminat-ingthe moderating effect of the original modera-tor (prosocial motivation). Having already pro-posed these relationships, we present our formalhypothesis for mediated moderation: prosocial mo-tivation strengthens the association between intrin-

sic motivation and creativity by encouraging per-spective taking.

Hypothesis 2c. Perspective taking mediates themoderating effect of prosocial motivation onthe association between intrinsic motivationand creativity.

We tested these hypotheses in three studies. InStudy 1, we tested Hypothesis 1 by examining therole of prosocial motivation in moderating the re-lationship between intrinsic motivation and cre-

ativity. In Studies 2 and 3, we tested the full theo-retical model depicted in Figure 1.

STUDY 1: METHODS

Sample and Procedures

We collected motivation data from 90 securityforce officers and collected lagged creativity ratingsfrom their supervisors at a military base in thenorthwestern United States. Of the officers, 77.8percent were male, and average job tenure for thesample was 3.25 years (s.d. 4.13 years). Theywere responsible for protecting physical securityand preventing theft and sabotage. A human re-sources professional sent an electronic message toall 269 security force officers on the base announc-ing a collaboration with a research team interestedin conducting an academic study of their experi-ences. The officers were invited to complete a sur-

vey online either during or outside of work hours.To protect confidentiality, the survey was hostedon a university server, and participants had theopportunity to identify themselves by code names,which were later matched to their supervisors rat-ings by a neutral third party. We received com-pleted surveys from 90 officers, for a response rateof 33.5 percent. Nine months later, we asked theirsupervisors to evaluate the creativity that they hadexhibited since the surveys were completed. Wereceived supervisor ratings for all 90 officers, a 100percent response rate. With the exception of five

supervisors who rated multiple employees, eachsupervisor rated a single unique employee.

The employees were responsible for monitoringand repairing equipment, coordinating and con-ducting surveillance activities and patrols, devel-oping contingency plans and disaster protocols,preparing for inspections, and assessing, respond-ing to, and neutralizing security threats. Examplesof creative ideas included developing protocols forunforeseen but serious threats, looking for ways toutilize limited staff and resources to cover severalthousand square miles of ground, suggesting newcontingencies and sources of variability in trainingprocedures and contingency plans, generating dy-namic interview protocols for questioning sus-pects, and finding faster ways to repair equipment.

Measures

Unless otherwise indicated, all items used a Lik-ert-type scale anchored at 1 (disagree strongly)and 7 (agree strongly).

Intrinsic and prosocial motivations. The officerscompleted the intrinsic and prosocial motivation



9/25

scales developed by Grant (2008). The scales openwith the question, Why are you motivated to doyour work? and then allow respondents to ratetheir intrinsic and prosocial motivations. The in-trinsic motivation scale is composed of four items,including Because I enjoy the work itself andBecause its fun ( .94). The prosocial motiva-tion scale is also composed of four items, includingBecause I want to help others through my workand Because I care about benefiting others throughmy work ( .91).1

Creativity. At the time that the officers com-pleted their surveys, we asked their supervisors topay attention to each officers creativity over thesubsequent months. Nine months later, we sent asurvey to the supervisors asking them to reflect onthe preceding nine months and to rate each officerscreativity. The supervisors were blind to the study

hypotheses and to officers survey responses. Weselected a nine-month time lag to provide supervi-sors with adequate opportunity to observe eachofficers creative ideas and contributions. The su-pervisors provided the ratings using the nine-itemcreativity scale developed by Tierney et al. (1999),which includes items such as Generates novel, butoperable work-related ideas and Serves as a goodrole model for creativity ( .97).

Control variables. We controlled for contextualand individual factors that could be expected toinfluence both motivation and creativity. Since au-

tonomy can facilitate both intrinsic motivation and

creativity (e.g., Elsbach & Hargadon, 2006), we con-trolled for this job characteristic, measured withthe scale developed by Morgeson and Humphrey(2006). Since conscientiousness and openness mayrelate to both creativity and intrinsic motivation(e.g., Shalley et al., 2004), we controlled for thesetwo traits, measured with the four-item scales de-veloped by Donnellan, Oswald, Baird, and Lucas(2006).

STUDY 1: RESULTS AND DISCUSSION

Means, standard deviation, and correlations forthe key study variables appear in Table 2. We began

by examining the factor structure of the three focalvariables: intrinsic motivation, prosocial motiva-tion, and creativity. We conducted a confirmatoryfactor analysis using EQS software version 6.1 with

maximum-likelihood estimation procedures (e.g.,Kline, 1998). The expected three-factor solutiondisplayed excellent fit with the data (2[116] 196.12, CFI .95, SRMR .04). All factor loadingswere statistically significant and ranged from .87 to.94 for the intrinsic motivation items, .82 to .90 forthe prosocial motivation items, and .84 to .95 forthe creativity items. We tested all alternative nestedmodels to examine whether a more parsimoniousmodel achieved equivalent fit (for intrinsic andprosocial motivations on the same factor, 2[118] 346.29, CFI .86, SRMR .09; for intrinsic moti-

vation and creativity on the same factor, 2

[118]

514.11, CFI .76, SRMR .19; for prosocial mo-tivation and creativity on the same factor, 2[118]447.92, CFI .80, SRMR .19; and for a one-factormodel, 2[119] 733.71, CFI .62, SRMR .23).Chi-square difference tests showed that our modelachieved significantly better fit.

We then conducted hierarchical ordinary leastsquares (OLS) regression analyses to test our hy-pothesis that prosocial motivation would strengthenthe association between intrinsic motivation andcreativity. We followed the moderated regression

1 We introduced this section of the survey by statingthat we were interested in employees workplace rela-tionships, requesting that responses focus on the peopledirectly affected by their jobs. We used the term othersto allow employees to focus on the direct beneficiaries oftheir own jobs. At the end of the survey, we included anopen-ended question asking participants to indicatewhich beneficiaries they had in mind. Ninety percent ofparticipants mentioned coworkers and civilians, and theremaining 10 percent mentioned supervisors.

TABLE 2Study 1: Means, Standard Deviations, and Correlationsa

Variable Mean s.d. 1 2 3 4 5 6

1. Creativity 5.52 1.13 (.97)2. Intrinsic motivation 3.84 1.89 .32** (.94)3. Prosocial motivation 5.13 1.43 .28** .55*** (.91)4. Autonomy 4.22 1.81 .07 .41*** .42*** (.97)5. Conscientiousness 5.32 1.08 .06 .13 .18 .15 (.75)6. Openness 4.87 1.03 .07 .14 .23** .09 .18 (.77)

a n 90. Coefficient alphas are on the diagonal in parentheses.** p .01

*** p .001



10/25

procedures recommended by Aiken and West(1991), entering the control variables in step 1,intrinsic and prosocial motivations in step 2, andtheir interaction in step 3. Table 3 depicts the re-sults of our moderated regression analyses. Withrespect to our core hypotheses about the relation-ship of intrinsic and prosocial motivations with

creativity, it is worth noting that intrinsic motiva-tion was a significant, independent predictor ofcreativity, but prosocial motivation was not (step2). However, as predicted in Hypothesis 1, the par-tialed product of intrinsic and prosocial motivationwas a significant, positive predictor of creativity(step 3). To facilitate the interpretation of the inter-action, as Aiken and West (1991) recommended,we plotted the simple slopes for the relationship

between intrinsic motivation and creativity at onestandard deviation above and below the mean ofprosocial motivation. The results, which are plot-ted in Figure 2, suggest that in keeping with Hy-pothesis 1, prosocial motivation strengthened theassociation between intrinsic motivation and cre-ativity. To test this interpretation, we statisticallycompared the two slopes to zero. As expected,when prosocial motivation was high, intrinsic mo-tivation significantly predicted higher levels of cre-ativity (b .52, s.e. .18, .46, t 2.84, p .01). When prosocial motivation was low, intrinsicmotivation did not predict creativity, as the slopedid not differ significantly from zero (b .01, s.e..18, .01, t 0.04, p .97).

These results provide initial support for our hy-

pothesis that prosocial motivation strengthens the

association between intrinsic motivation and cre-ativity. Officers with high levels of intrinsic moti-vation were more likely to earn higher supervisorcreativity ratings when they also had high levels ofprosocial motivation. Although these findings areencouraging in providing initial support for ourtheoretical prediction, to strengthen our confidence

in their validity and generalizability, it was impor-tant to conduct a constructive replication with dif-ferent samples and measures. In addition, we hadyet to test our mediating hypotheses about perspec-tive taking as an explanatory mechanism for themoderating role of prosocial motivation, as well asto rule out alternative explanations.

STUDY 2: METHODS

To address these limitations, our second studyused new measures of prosocial motivation andcreativity, included perspective taking, and con-trolled for several alternative explanations.

Sample and Procedures

We collected data from 111 employees and theirdirect supervisors at a water treatment plant in thesoutheastern U.S. We sent e-mails to all 796 em-ployees on staff, asking them to participate in aconfidential survey about work motivation. We re-ceived responses from 209 employees, for a re-sponse rate of 26.3 percent. We asked participantsto list their supervisors names and e-mail ad-

dresses, and we sent a creativity survey to their

TABLE 3Study 1: Moderated Regression Analyses Predicting Creativitya, b

Variable

Step 1 Step 2 Step 3

b s.e. t b s.e. t b s.e. t

Autonomy .04 .07 .07 0.58 .08 .07 .12 1.01 .06 .07 .10 0.85Conscientiousness .01 .13 .01 0.05 .05 .12 .04 0.37 .07 .12 .07 0.61Openness .01 .11 .01 0.12 .06 .10 .06 0.57 .06 .10 .06 0.57Intrinsic motivation .30 .14 .26 2.09* .30 .14 .26 2.11*Prosocial motivation .23 .15 .21 1.56 .36 .16 .32 2.28*

Intrinsic prosocial motivation .26 .13 .24 2.08*

R2 .01 .13* .18*F (df) 0.13 (3, 84) 5.95 (2, 82) 4.33 (1, 81)R2 .01 .12** .05*

a n 90. Values in bold are relevant to tests of hypotheses.b Since there were five supervisors who rated multiple employees, we conducted additional analyses using fixed effects, inserting

dummy variables for each supervisor to adjust for dependencies in the data. The interaction between intrinsic and prosocial motivationswas still significant (b .36, s.e. .13, .33, t 2.71, p .01). We replicated this pattern of results using random coefficient modelingas well.

* p .05** p .01



11/25

supervisors, asking supervisors to rate only onesubordinate. We received 111 unique supervisorresponses, representing a response rate of 53.1 per-cent. For the final data set, we focused on thematched sample of 111 employees and their directsupervisors. Of these employees, 68.5 percent werefemale, and they averaged 44.1 years of age, 14.0years of experience in the organization, and 6.47years in their current jobs. Of the supervisors, 77.5percent were male, and they averaged 49.1 years ofage, 19.0 years of experience in the organization,and 5.1 years supervising the employees. We askedemployees to provide ratings of their intrinsic and

prosocial motivations, levels of perspective taking,and several control variables and asked supervisorsto rate employees creativity. The employees wereresponsible for monitoring and repairing equip-ment, responding to customer questions, updatingsafety standards, developing and improving engi-neering procedures, preventing and resolving sys-tem problems, reducing pollution, and implement-ing new testing processes. Examples of creativeideas included developing techniques for prevent-ing equipment failures, proposing new pollutioncontrol methods, and suggesting new work pro-

cesses and safety protocols.

Measures

Unless otherwise indicated, all measures used ascale anchored at 1 (disagree strongly) and 7(agree strongly).

Intrinsic and prosocial motivations. We mea-sured intrinsic motivation with the same four-itemscale (Grant, 2008) as we used in our previousstudy ( .91). We measured prosocial motivationusing Grant and Sumanths (2009) five-item scale,

which includes items such as I get energized byworking on tasks that have the potential to benefitothers and I prefer to work on tasks that allow meto have a positive impact on others ( .90).

Perspective taking. Because perspective takingis an intrapsychic or internal psychological processof adopting anothers viewpoint, it is often not ap-parent to observers or overtly displayed in every-day behaviors. Thus, in field settings, employeesthemselves are often in the strongest position toreport on their own perspective-taking efforts. Re-search has demonstrated the reliability and validityof self-reports of perspective taking (e.g., Davis,Conklin, Smith, & Luce, 1996), which predict su-pervisor ratings of employee behaviors such as co-operation with team members (Parker & Axtell,2001) and helping customers (Axtell et al., 2007).Employees indicated the extent to which they tookothers perspectives at work using a four-item scaleadapted from the Davis et al. (1996) perspectivetaking measure. The items were On the job, I fre-quently try to take other peoples perspectives,At work, I often imagine how other people arefeeling, On the job, I make an effort to see theworld through others eyes, and At work, I regu-larly seek to understand others viewpoints ( .80).2

2 As in Study 1, we asked participants to respond tothe prosocial motivation and perspective taking ques-tions in reference to the direct beneficiaries of their jobs.In an open-ended question at the end of the survey, 89percent of participants mentioned coworkers and com-munity members, and the remaining 11 percent men-tioned supervisors.

FIGURE 2Study 1: Simple Slopes

3.5

4

4.5

5

5.5

6

6.5

Low intrinsic motivation

Creativity

Low prosocial

motivation

High prosocial

motivation

High intrinsic motivation



12/25

Creativity. Supervisors provided ratings of em-ployees creativity using the 13-item creativityscale developed by George and Zhou (2001), whichincludes Comes up with new and practical ideasto improve performance and Is a good source ofcreative ideas ( .97).

Control variables. From a demographic stand-

point, we controlled for sex, job tenure, and maritalstatus. For job characteristics, we controlled forautonomy with the same scale used in Study 1(Morgeson & Humphrey, 2006). We also used theMorgeson and Humphrey (2006) scale to measureskill variety, which has been linked to both in-trinsic motivation and creativity (e.g., Elsbach &Hargadon, 2006). For personality traits, we measuredconscientiousness and openness using the samescales as in Study 1 (Donnellan et al., 2006). We alsocontrolled for psychological safety, which might en-hance both intrinsic motivation and creativity, usingEdmondsons (1999) seven-item scale.

To provide a more robust test of the unique mod-erating role of prosocial motivation, we also con-trolled for extrinsic motivations. Doing so had thepotential to strengthen our results by allowing us totest whether extrinsic motivations, which can alsocultivate a focus on outcome goals, may serve thesame function as prosocial motivation in enhanc-ing the relationship between intrinsic motivationand creativity. We measured three extrinsic moti-vationsexternal, introjected, and identifiedus-ing four items for each adapted from Ryan andConnell (1989). As with the intrinsic and prosocial

motivation items, we opened with the question,Why are you motivated to do your work? Theexternal items were Because I need to pay my

bills, Because I need to earn money, Because Ihave to, and Because I need the income ( .92). The introjected items were Because I want toavoid feeling guilty, Because Ill feel bad aboutmyself if I dont, Because I want to avoid looking

bad, and Because Ill feel ashamed if I dont (.90). The identified items were Because I think its

important, Because I dont want to cause harm,Because its satisfying, and Because I want to doa good job ( .80).


Means, standard deviation, and correlations aredisplayed in Table 4. We conducted a confirmatoryfactor analysis to assess the factor structure of thefour key variables: intrinsic motivation, prosocialmotivation, perspective taking, and creativity. Thepredicted four-factor solution achieved adequate fitwith the data (2[293] 522.54, CFI .92, SRMR.06). All factor loadings were statistically signifi-cant, ranging from .82 to .93 for the intrinsic moti-vation items, .75 to .90 for the prosocial motivationitems, .66 to .83 for the perspective taking items,and .65 to .95 for the creativity items. Plausiblealternative models displayed significantly poorerfit. However, the CFI value fell below .95, whichmight be due to the fact that the creativity scaleincluded a large number of items, constituting anoveridentified variable. This can cause parameterinstability related to the presence of multiple solu-tions, correlated residuals and cross-loadings, and

increased standard errors, especially in a smallsample such as ours (Little, Cunningham, Shahar, &Widaman, 2002). To adjust for these issues by re-

TABLE 4Study 2: Means, Standard Deviations, and Correlationsa

Variableb Mean s.d. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

1. Creativity 5.49 0.98 (.97)2. Intrinsic motivation 5.63 1.03 .21* (.91)3. Prosocial motivation 5.74 0.78 .10 .27** (.90)4. Perspective taking 5.49 0 .76 .02 .08 .28** (.80)5. External motivation 6.25 0.85 .06 .17 .02 .15 (.92)6. Introjected motivation 5.60 0.97 .04 .11 .24* .30** .11 (.90)

7. Identified motivation 4.14 1.15 .13 .09 .04 .05 .20* .10 (.80)8. Sex 0.32 0.47 .09 .03 .17 .13 .08 .03 .12 9. Job tenure 6.47 7.52 .05 .06 .07 .15 .05 .04 .06 .16

10. Marital status 0.77 0.42 .04 .11 .05 .01 .08 .02 .00 .15 .18 11. Autonomy 5.66 0.88 .19* .19* .22* .13 .06 .16 .10 .11 .07 .08 (.87)12. Skill variety 6.29 0.62 .15 .32** .22* .22* .03 .22* .02 .10 .03 .14 .43*** (.91)13. Psychological safety 3.99 0.52 .18 .09 .12 .10 .11 .03 .09 .04 .00 .09 .21* .16 (.68)14. Conscientiousness 5.45 1.01 .17 .07 .19* .09 .01 .03 .07 .09 .08 .06 .22* .06 .12 (.79)15. Openness 5.29 1.04 .13 .15 .11 .17 .09 .06 .27** .18 .12 .06 .13 .33*** .19* .11 (.78)

a n 111. Coefficient alphas are on the diagonal in parentheses.b For sex, 0 female, 1 male, For marital status, 0 single, 1 married.

* p .05** p .01

*** p .001



13/25

ducing the number of indicators to the more parsi-monious three per factor, we aggregated the creativ-ity scale items using parceling techniques. Afterreducing the 13-item creativity scale to three par-cels, the four-factor solution displayed excellent fit(2[98] 146.53, CFI .96, SRMR .06). Thissuggested that the relatively low CFI value might be

an artifact of sample size and scale length, indicat-ing that the psychometric properties of our vari-ables were acceptable for further investigation.

To test Hypothesis 1, we used the same moder-ated regression procedures as in Study 1 (Aiken &West, 1991). Table 5 reports results of these analy-ses. Even after we included all of the control vari-ables and the interactions of external, introjected,and identified motivations with intrinsic motiva-tion, the results showed that intrinsic and prosocialmotivations interacted significantly to predict su-pervisors ratings of creativity, accounting for 3 per-cent additional variance in creativity (see Table 5,Step 1). Simple slopes suggested that as in the prior

study, intrinsic motivation was positively relatedto creativity when prosocial motivation was high,

but not when it was low (see Figure 3). Comparingthe slopes to zero substantiated this interpretation,as intrinsic motivation significantly predictedhigher levels of creativity when prosocial motiva-tion was high (b .45, s.e. .14, .43, t 3.13,

p .01) but not when it was low (b .00, s.e. .13, .00, t 0.01, p .99). Thus, prosocial moti-vation once again strengthened the association be-tween intrinsic motivation and creativity.

We then tested Hypotheses 2a and 2b. In supportof Hypothesis 2a, prosocial motivation was signif-icantly associated with perspective taking (see Ta-

ble 5, Perspective Taking). In support of Hypothesis2b, a moderated regression analysis showed thatperspective taking and intrinsic motivation inter-acted to predict creativity (see Table 5, Step 2).Simple slopes showed that the form of the moder-ating effect of perspective taking mirrored the mod-erating effect for prosocial motivation: intrinsic

TABLE 5Study 2: Regression Analysesa,b

Variable

Perspective Taking Creativity, Step 1 Creativity, Step 2


Sex .19 .16 .12 1.21 .00 .20 .00 0.02 .04 .20 .02 0.22Job tenure .01 .01 .13 1.39 .01 .01 .08 0.88 .01 .01 .11 1.18Marital status .05 .17 .03 0.29 .09 .22 .04 0.41 .07 .22 .03 0.31Autonomy .05 .09 .06 0.55 .28 .11 .25 2.43* .26 .12 .24 2.21*

Skill variety .12 .15 .10 0.84 .18 .18 .11 1.00 .27 .18 .17 1.47Psychological safety .02 .14 .01 0.12 .37 .17 .20 2.12* .30 .17 .17 1.73Conscientiousness .03 .07 .04 0.43 .31 .09 .32 3.41** .34 .09 .35 3.66***Openness .07 .08 .09 0.84 .13 .10 .14 1.30 .07 .10 .07 0.69External motivation .08 .08 .10 1.03 .03 .09 .03 0.32 .02 .09 .02 0.24Introjected motivation .19 .07 .26 2.64* .03 .09 .03 0.34 .00 .09 .00 0.02Identified motivation .02 .08 .03 0.26 .17 .10 .17 1.72 .12 .10 .12 1.18Intrinsic motivation .11 .09 .14 1.20 .37 .11 .36 3.26** .43 .13 .39 3.36**Prosocial motivation .18 .09 .22 2.11* .09 .11 .09 0.90 .00 .11 .00 0.02Intrinsic external motivation .21 .10 .24 2.09* .20 .13 .17 1.55 .36 .13 .30 2.69**Intrinsic introjected motivation .05 .07 .08 0.72 .08 .08 .09 0.94 .05 .08 .07 0.65Intrinsic identified motivation .05 .08 .06 0.59 .17 .09 .18 1.79 .13 .12 .12 1.14Intrinsic prosocial motivation .03 .08 .04 0.40 .20 .09 .19 2.06* .13 .10 .13 1.37Perspective taking .07 .11 .07 0.65Perspective taking intrinsic

motivation

.32 .14 .25 2.20*

R2 .28* .34** .38**F (df) 2.07 (17, 93) 2.56 (17, 93) 2.44 (2, 91)R2 .28* .34** .04*

a n 111. Values in bold are relevant to tests of hypotheses.b We also conducted an analysis in which we entered the interaction of intrinsic and prosocial motivations in a separate step between

steps 2 and 3, and the addition of this interaction term significantly increased variance explained by 3 percent (to R2 .33, F[1, 93] 4.25,p .04).

* p .05** p .01

*** p .001



14/25

motivation was positively associated with creativ-ity when perspective taking was high (b .60, s.e. .19, .54, t 3.23, p .01) but not when itwas low (b .03, s.e. .15, .03, t 0.23,p .82). In this analysis, when we included themoderating effect of perspective taking, the moder-ating effect of prosocial motivation was reduced tononsignificance (see Table 5, Steps 1 and 2).

To complete our test of Hypothesis 2c, whichpredicts that perspective taking mediates the mod-erating effect of prosocial motivation on the rela-tionship between intrinsic motivation and creativ-ity, we followed the moderated path analysis

procedures recommended by Edwards and Lambert(2007). Our previous analyses showed that proso-cial motivation predicted perspective taking andthat perspective taking moderated the association

between intrinsic motivation and creativity, reduc-ing the coefficient for the moderating effect ofprosocial motivation. To examine whether ac-counting for the moderating effect of perspectivetaking significantly reduced the moderating effectof prosocial motivation, we used a bootstrap proce-dure.3 This procedure allowed us to examine the

size of the indirect effect of prosocial motivation (inmoderating the relationship between intrinsic mo-tivation and creativity) through the mediator of per-spective taking. We constructed bias-corrected con-fidence intervals by drawing 1,000 random sampleswith replacement from the full sample. An indirecteffect is significant when the 95% confidence in-terval excludes zero (Edwards & Lambert, 2007;MacKinnon, Fairchild, & Fritz, 2007). In our data,the size of the indirect effect from the full samplewas .092, and the 95% confidence interval from the

bootstrap analysis excluded zero (.021, .224). Theseresults support Hypothesis 2c, demonstrating thatperspective taking mediated the moderating effectof prosocial motivation on the relationship be-tween intrinsic motivation and creativity.

These results constructively replicated and ex-tended our findings from Study 1. We found thatprosocial motivation enhanced the relationship be-tween intrinsic motivation and creativity in a newsample with different measures of prosocial moti-vation and creativity and that perspective takingmediated this moderating effect. Furthermore, themoderating effect of prosocial motivation held evenafter we controlled for external, introjected, andidentified motivations, their interactions with in-trinsic motivation, and several possible commoncauses of intrinsic motivation and creativity. How-ever, these results are subject to at least two keylimitations. First, both of our studies relied on cor-relational designs and used the same measure ofintrinsic motivation, making them vulnerable toalternative explanations. To strengthen causal in-ferences, it was important to use an experimentaldesign with a different operationalization of intrin-sic motivation.

3 Our model differs from Edwards and Lamberts(2007) models in that we started with a moderating effectthat we sought to explain, rather than starting with amediated effect that is then moderated at one or twostages. However, the moderated path analysis proceduresspecified by Edwards and Lambert still applied here; thekey difference was that our model involved computingthe indirect effect by calculating the reduced-form equa-tion for the product of (1) the path from prosocial moti-vation to perspective taking and (2) the path for theinteraction of perspective taking and intrinsic motivationin predicting creativity.

FIGURE 3Study 2: Simple Slopes



15/25

Second, in both studies, we relied on supervisorratings of creativity, which are vulnerable to a num-

ber of weaknesses (e.g., Shalley et al., 2004). For ex-ample, supervisors do not always have the opportu-nity to observe employees creativity in situ;supervisors are not always experts in the product andservice domains in which employees are expected to

be creative; and supervisor ratings are often skewedby halo effects, affective cues, and other information-processing and decision-making biases that influenceappraisals. For example, it is possible that supervi-sors are biased toward seeing employees with highintrinsic and prosocial motivationswho expresshigh task interest and a concern for contributing toother people and the organizationin a more favor-able light. To overcome these limitations and ruleout such alternative explanations, we conducted astudy in which we used independent ratings of actualideas that were separated from the individuals whogenerated them.

STUDY 3: METHODS

To strengthen causal inferences and rule out al-ternative explanations, we conducted a laboratoryexperiment in which participants generated cre-ative ideas to solve a business problem. We inde-pendently manipulated intrinsic and prosocial mo-tivations, and we measured perspective taking as amediator. To prevent supervisor biases from influ-encing creativity ratings, we asked independentraters to evaluate the creative ideas generated by

participants. The raters were blind to all character-istics of the individual participants, as well as tothe studys design, manipulations, and hypotheses.In addition, Study 2 only provided correlationalevidence linking prosocial motivation to perspec-tive taking. Directly manipulating prosocial moti-vation in this study allowed us to rule out thepossibility that omitted knowledge, skill, and abil-ity variablesimportant influences on creativity(Amabile & Mueller, 2007)were responsible forthe results. For example, emotional intelligencemay influence both prosocial motivation (e.g., Cote

& Miners, 2006) and perspective taking (Schutte etal., 2001). By randomly assigning participants toexperience different levels of prosocial motivation,we could test whether prosocial motivation di-rectly increases perspective taking.

Sample, Design, and Procedures

We conducted an experiment with 100 under-graduates at a large public U.S. university. Sixtypercent of the participants were female, and theycompleted the study via computer. We recruited

them from a university mailing list in exchange fora $10 gift certificate to Amazon.com. The experi-ment used a two-by-two (intrinsic motivation, low/high, by prosocial motivation, low/high) between-subjects factorial design. We introduced the study

by explaining that we were interested in studyinghow people solve business problems. We informed

participants that they would have a chance to solvea problem that a local organization was facing, andonline software randomly assigned them to one ofthe four experimental conditions using a randomnumber generator. At that point, we introducedour manipulations of intrinsic and prosocialmotivations.

Intrinsic motivation manipulation. Psycholo-gists typically manipulate intrinsic motivation byvarying the interest level of a task and/or theamount of free choice that individuals are providedin completing the task (Deci, Koestner, & Ryan,1999). To ensure that participants in the low andhigh intrinsic motivation conditions would experi-ence different levels of intrinsic motivation, ourmanipulation featured a combination of task inter-est and task choice. We gave participants in thehigh intrinsic motivation condition a choice be-tween two tasks and accepted their choice of thetask described as interesting, supporting the ex-perience of self-determination. For the low intrin-sic motivation condition, we gave participants thesame choice but then rejected their selection andrequired them to perform the task described asboring, undermining the experience of self-deter-

mination. In actuality, all participants performedthe same task; the key difference was that partici-pants in the high intrinsic motivation conditionwere allowed to choose a task framed as interesting,while participants in the low intrinsic motivationcondition were prevented from realizing theirchoices by the requirement to perform a taskframed as boring. To set the stage for the manipu-lations, we provided participants in both condi-tions with the following overview:

We are conducting two different studies with local

bands, and we have received feedback that insteadof assigning you randomly to one of the studies,participants like to have a choice. To inform yourdecision, we have asked past participants to rankhow interesting each study was. Accordingly, pleasechoose one of the two tasks below:

(a) A music study that has been rated by participantsas extremely interesting, with average ratings of 6.73on a 7-point fun scale

(b) A recording study that has been rated by partic-ipants as relatively boring, with average ratings of2.3 on a 7-point fun scale



16/25

We chose these labels of music study and re-cording study because the ensuing task could beinterpreted as either one, and we wanted to guideall participants to select the interesting study.Indeed, all participants selected the first study,which took them to the next screen on their com-puters. At this point, we introduced the intrinsic

motivation manipulation.For those in the high intrinsic motivation condi-tion, the screen read, Thank you for selecting themusic study, and provided the instructions for thetask. This allowed participants to believe that theywould be completing the interesting study that theyhad chosen, signaling high levels of both task in-terest and free choice. In the low intrinsic motiva-tion condition, the screen read, Unfortunately, themusic study that you chose is full. We will needyou to participate in the recording study. This ledparticipants to believe that they would be complet-ing the boring study that they did not choose, sig-naling low levels of both task interest and freechoice. We believe that the free choice manipula-tion has greater ecological validity for organization-al settings, as there is ample evidence that givingemployees choices can achieve long-term increasesin intrinsic motivation (e.g., Hackman, Pearce, &Wolfe, 1978). In contrast, task-framing effects aretypically short-lived because employees actualtask experiences tend to override frames (e.g., Za-lesny & Ford, 1990). Nevertheless, in light of evi-dence that describing a task as interesting can tem-porarily increase intrinsic motivation in the

laboratory (e.g., Glynn, 1994; Zalesny & Ford,1990), the task-framing cues were a potentially ef-fective reinforcement of intrinsic motivation.

Prosocial motivation manipulation. For all con-ditions, we then informed participants that theywould be asked to generate ideas to increase a local

bands revenues. Psychologists typically manipu-late prosocial motivation by varying the level ofneed that a task beneficiary or recipient expresses,which cultivates empathic concern and thus a de-sire to help the beneficiary or recipient (Batson,1998). Accordingly, we provided participants with

different information about the level of need thatthe band members were experiencing. In the highprosocial motivation condition, participants read astatement that the band members were in direstraits:

We have been approached by a local band, the FileDrawers, for help with generating ideas for increas-ing their revenues. In the last three years, the FileDrawers have seen their CD sales drop by 92%. Allsix members of the band have families to feed, andthey are in dire straits; on a weekly basis, they arestruggling to make ends meet. As Bryan Strickland,

the lead singer, told us, It used to be that we madeour profits on CD sales, which have tanked. On ourcurrent profits, I can hardly support myself, letalone my family. Our regional popularity has takenoff, and I think were on the verge of success if wecan only stay profitable for the next four monthswhile were writing and recording new songs. I re-ally need your help in coming up with new ideas for

bringing in revenue and publicizing our music. Illdo whatever it takes.

In the low prosocial motivation condition, theband members were not in need:

A local band, the File Drawers, is looking to generateideas for increasing their revenues. In the last threeyears, the File Drawers have seen their CD salesdrop by 7%. All six members of the band are finan-cially secure with successful careers in business andlaw, but the band is a hobby that they find enjoyableand would like to continue. As Bryan Strickland,the lead singer, told us, It used to be that we sold a

lot of CDs, but now were looking for additionalways to get our music out there.

Participants in all conditions were asked to gen-erate ideas for how the band could increase reve-nues. We informed them that we would send theirsuggestions to the band at the end of the month.After they had generated their ideas, they com-pleted a questionnaire that included a measure ofperspective taking and manipulation checks. Wethen recruited two independent experts to ratethe creativity of the ideas that the participantsgenerated.

Measures

Unless otherwise indicated, all items used thesame scale anchors as in the previous studies (1 disagree strongly, and 7 agree strongly).

Creativity. We recruited two independent ex-perts to rate the creativity of participants ideas.One of the raters had created and sold music CDswhile playing in a band for four years, the other hadgained experience with the music industry whileworking as a business consultant for three years,

and both had completed extensive coursework inorganizational behavior. They also refreshed theirknowledge of the music industry by conductingresearch on strategies that bands and record com-panies use to market acts and sell music. We pro-vided them with each participants ideas in aspreadsheet, stripped of all identifying informa-tion. The only information that we offered aboutthe study was that we had asked participants toprovide ideas for how a local band could generaterevenues and were interested in the raters assess-ments of the creativity of each participants ideas.



17/25

Following Amabiles (1996) consensual assessmenttechnique, we explicitly defined creative ideas asthose that are both novel and useful, and we askedthe raters to evaluate the ideas on a scale anchoredat 1, not at all creative; 4, somewhat creative;and 7 very creative. Since the two raters achievedgood reliability (ICC2 .69, p .001), and agree-

ment (average deviation .63)well within con-ventional guidelines (LeBreton & Senter, 2008)we averaged their ratings into a measure of theoverall creativity of each participants ideas.

Examples of ideas rated as highly creative in-cluded producing free music for video games inexchange for publicity; playing concerts for charityor a college homecoming football game; creating a

behind-the-scenes YouTube video to generate in-terest in the band, along with blog and Twitterentries; hiring business students for volunteer mar-keting internship positions; and building the

bands brand image by offering free music lessons,a day with the lead singer, a chance to visit with the

band backstage or perform with the band onstage,or an opportunity to write a song for the band. Ideasrated as less creative included eliminating CDsales; performing more concerts in bars and restau-rants; selling individual songs on iTunes; selling

band merchandise such as T-shirts; doing radiopromotions; providing free music downloads on awebsite; and advertising in newspapers, on web-sites, and on bulletin boards.

Perspective taking. Participants completed atask-specific measure of perspective taking using

four items adapted from the perspective takingscale developed by Davis et al. (1996). The itemsasked participants to indicate the extent to whichthey tried to see the band members perspectives: Imade an effort to see the world through the bandmembers eyes, I imagined how the band mem-

bers were feeling, I sought to understand theband members viewpoints, and I tried to take theband members perspectives ( .89).

Manipulation checks and control variables. Toensure that our manipulations were effective, weasked participants to complete scales measuringtheir levels of intrinsic and prosocial motivation inthe task. For intrinsic motivation, since our manip-ulation varied both task interest and free choice, wemeasured both constructs. As Deci et al. (1999: 655)

stated, The best way to ensure one is assessingintrinsic motivation is to measure both free-choice. . . and self-reported interest and to consider themintrinsic motivation only when they correlatewithin conditions or studies. We measured taskinterest with the seven-item interest/enjoymentscale developed by Ryan, Koestner, and Deci(1991), which includes items such as I enjoyeddoing this task very much and This task was funto do ( .94). We measured perceptions of freechoice with the Ryan et al. (1991) seven-item per-ceived choice scale, which includes items such asI did this activity because I wanted to and I

believe I had some choice about doing this activity( .91). We measured prosocial motivation withfour items adapted from Grants (2008) prosocialmotivation scale in such a way that they focusedspecifically on the task: I wanted to have a posi-tive impact on the band members, I wanted tohelp the band members, I was focused on bene-fiting the band members, and I was trying tomake the band members better off ( .70). Fi-nally, to rule out the possibility that knowledgerelevant to the problem influenced the results, wecontrolled for two sources of experience: whether

participants had ever worked in the music industryand whether they had ever earned money throughmusic.


Means and standard deviations for each condi-tion appear in Table 6. A multivariate analysis ofvariance (MANOVA) showed the expected main

TABLE 6Study 3: Means and Standard Deviations by Conditiona

Condition CreativityPerspective

TakingPerceived

ChoiceTask

InterestProsocial

Motivation

Low intrinsic motivation, low prosocialmotivation (n 25)

4.04 (1.03) 4.19 (1.24) 5.51 (1.45) 3.43 (1.33) 3.96 (1.04)

Low intrinsic motivation, high prosocialmotivation (n 22)

4.11 (1.55) 4.76 (1.24) 5.13 (1.51) 3.75 (1.15) 4.68 (0.95)

High intrinsic motivation, low prosocialmotivation (n 26)

3.81 (1.11) 4.58 (0.71) 6.01 (0.96) 3.95 (1.00) 4.46 (0.68)

High intrinsic motivation, high prosocialmotivation (n 27)

4.96 (1.23) 4.99 (0.71) 5.90 (1.12) 4.49 (0.88) 4.81 (0.70)

a Standard deviations are in parentheses.



18/25

effects of the intrinsic motivation manipulation onperceived choice (F[1, 96] 6.23, p .01) and taskinterest (F[1, 96] 8.28, p .01) as well as theexpected main effects of the prosocial motivationmanipulation on self-reports of prosocial motiva-tion (F[1, 96] 9.94, p .01). No other effects weresignificant. Turning to creativity as the dependent

variable, in support of Hypothesis 1, theMANOVA also showed a significant interactioneffect of the intrinsic and prosocial motivationmanipulations on creativity (F[1, 96] 4.78, p .03). The interaction was robust even after wecontrolled for the two knowledge variables (F[1,94] 4.50, p .04). Simple effects showed thatintrinsic motivation increased creativity whenprosocial motivation was high (F[1, 96] 6.70, p .01), but not when prosocial motivation waslow (F[1, 96] .50, p .48).

Turning to perspective taking, in support of Hy-pothesis 2a, the MANOVA showed a significanteffect of the prosocial motivation manipulation onperspective taking (F[1, 96] 6.03, p .02). Theeffect was stronger after the two knowledge vari-ables were controlled for (F[1, 94] 6.58, p .01),and no other effects were significant. To test Hy-pothesis 2b, we used the Aiken and West (1991)moderated regression procedures to examinewhether participants reports of perspective takingmoderated the effect of the intrinsic motivationmanipulation on creativity. Table 7 presents theseresults. Supporting Hypothesis 2b, perspective tak-ing interacted positively with the intrinsic motiva-

tion manipulation to predict higher creativity (see

Table 7, column 3). In this analysis, when weadded the moderating effect of perspective taking,the moderating effect of prosocial motivation wasreduced to nonsignificance (see Table 7, columns 2and 3).

These analyses established the moderating effectof prosocial motivation on the relationship be-

tween intrinsic motivation and creativity, the effectof prosocial motivation on perspective taking, andthe moderating effect of perspective taking on therelationship between intrinsic motivation and cre-ativity. We then examined whether perspective tak-ing mediated the moderating effect of prosocial mo-tivation on the relationship between intrinsicmotivation and creativity, using the proceduresrecommended by Edwards and Lambert (2007).Drawing 1,000 random samples using replacementfrom the full sample, we constructed bias-correctedconfidence intervals for the indirect moderatingeffect of prosocial motivation through the mediatorof perspective taking. The indirect effect from thefull sample was .29. In keeping with Hypothesis 2c,the 95% confidence interval from the bootstrapanalysis excluded zero (.052, .730). These resultssupport Hypothesis 2c, showing that perspectivetaking mediated the moderating effect of prosocialmotivation on the association between intrinsicmotivation and creativity. This evidence construc-tively replicates our findings that prosocial motiva-tion enhances the effect of intrinsic motivation oncreativity, and the experimental manipulations fa-cilitate stronger causal inferences than our prior

studies allowed.

TABLE 7Study 3: Mediated Moderation Analysisa, b

Variables

Perspective Taking Creativity (Step 1) Creativity (Step 2)


Knowledge control 1 0.44 .54 .09 0.82 0.59 .67 .10 0.88 0.45 .64 .08 0.71Knowledge control 2 0.31 .69 .05 0.45 0.07 .86 .01 0.08 0.17 .82 .02 0.20Intrinsic motivation 0.40 .28 .20 1.42 0.23 .35 .09 0.67 0.25 .34 .10 0.73Prosocial motivation 0.61 .30 .30 2.05* 0.12 .37 .05 0.33 0.03 .36 .01 0.08

Intrinsic prosocial motivation 0.16 .40 .07 0.41 1.06 .50 .36 2.12* 0.84 .49 .29 1.70Perspective taking 0.15 .14 .12 1.04Perspective taking intrinsic motivation 0.55 .27 .24 2.07*

R2 .09 .13* .22**F (df) 1.90 (5, 94) 2.88 (5, 94) 5.23** (2, 92)R2 .09 .13 .09**

a In predicting perspective-taking, when we entered prosocial motivation in a separate step, variance explained increased significantlyby 4 percent (F[1, 94] 4.22, p .04).

b Values in bold are relevant to tests of hypotheses.* p .05

** p .01



19/25

GENERAL DISCUSSION

In three studies, we tested the hypothesis that therelationship between intrinsic motivation and cre-ativity is moderated by other-focused psychologi-cal processes. The studies provide convergent evi-dence in support of this hypothesis, revealing thatperspective taking, as generated by prosocial moti-

vation, strengthens the association between intrin-sic motivation and creativity. The use of multi-source measures in three distinct sampleswithlagged creativity data in Study 1, multiple moti-vations controlled in Study 2, and experimentaldata with independent expert ratings in Study3strengthens the validity of our conclusions.

Theoretical Contributions

Our research takes a step toward resolving thecontroversy about the link between intrinsic moti-vation and creativity. We proposed and found thatthis relationship is contingent on other-focusedpsychological processes: intrinsic motivation ismost likely to be associated with higher levels ofcreativity when employees are also prosocially mo-tivated to take the perspectives of others. Studieshave begun to identify contextual moderators of therelationship between intrinsic motivation and cre-ativity, such as task complexity (Gagne & Deci,2005) and leader-member exchange relationships(Tierney et al., 1999), both of which are thought tostimulate the creativity of intrinsically motivated

employees by providing them with challenges andfreedom from constraints. However, little researchhas addressed the possibility that other psycholog-ical processes may moderate this relationship.

Our studies demonstrate that perspective takingis an other-focused psychological process thatstrengthens the relationship between intrinsic mo-tivation and creativity. Psychologists have recog-nized that perspective taking can directly enhancecreativity by providing access to new ideas (e.g.,Galinsky, Maddux, Gilin, & White, 2008), and or-ganizational scholars have found that perspective

taking increases the usefulness of ideas (Mohrmanet al., 2001). However, neither group has developedor tested theory on the role of perspective taking inmoderating the association between intrinsic moti-vation and creativity. Our emphasis on perspectivetaking answers recent calls to identify moderatorsof the relationship between intrinsic motivationand creativity (George, 2007; Shalley et al., 2004).We also introduce perspective taking as a newmechanism for explaining the moderating effects ofprosocial motivation on the relationship betweenintrinsic motivation and creativity.

In doing so, our research presents a new rela-tional view of creativity. A number of researchershave studied how structural and behavioral dimen-sions of interpersonal relationships, such as socialnetworks (e.g., Perry-Smith, 2006; Perry-Smith &Shalley, 2003) and communication styles (e.g.,Amabile, 1979; Koestner et al., 1984; Shalley &

Perry-Smith, 2001), influence creativity. Our re-search complements these structural and behav-ioral approaches by documenting the importance ofthe psychologicaldimensions of interpersonal rela-tionships in fueling creativity.

Interestingly, several studies have suggested thatother-focused psychological processes can con-strain creativity by fostering a focus on conformity,which reduces employees capability and motiva-tion to think divergently. For example, Goncaloand Staw (2006) found that other-focused valuesemphasizing collectivism are associated with lowercreativity in groups. In contrast, we found evidencethat different other-focused psychological process-esprosocial motivation and perspective takingare associated with higher creativity. These find-ings may be explained by attending to howcollectivism differs from prosocial motivation andperspective taking. Collectivism is rooted in con-formity values emphasizing the importance ofmeeting others expectations and maintaining har-mony, which can encourage employees to suppresscreative thoughts and unique ideas (Goncalo &Staw, 2006). On the other hand, prosocial motiva-tion is rooted in values of benevolence and univer-

salism emphasizing the importance of benefitingothers (Grant, 2008; Schwartz & Bardi, 2001),which can encourage employees to think creativelyabout others perspectives and identify new strate-gies for helping them (De Dreu et al., 2000; Parker &Axtell, 2001). Our research thereby suggests thatdifferent other-focused psychological processesmay have contrasting effects on creativity.

Finally, our research deepens knowledge aboutthe interaction of intrinsic and prosocial motiva-tions, addressing calls to explore how multiple mo-tivations interact to influence creativity (Amabile,

1996; Eisenberger & Aselage, 2009; George, 2007).Although research has shown that intrinsic andprosocial motivations interact to predict higherpersistence, performance, and productivity (Grant,2008), these behaviors emphasize wo

the necessity of others is the mother of invention

Documents