The impact of competition on intrinsic motivation and creativity: considering gender, gender segregation and gender role orientation

Personality and Individual Differences 30 (2001) 1273±1289

www.elsevier.com/locate/paid

The impact of competition on intrinsic motivation and creativity: considering gender, gender segregation and gender role orientation Regina Conti a,*, Mary Ann Collins b, Martha L. Picariello c a

Department of Psychology, Colgate University, 13 Oak Drive, Hamilton, NY 13346-1398, USA Department of Psychology, Spring Hill College, 4000 Dauphin Street, Mobile, AL 36608, USA c Department of Psychology, Brandeis University, 415 South Street, Waltham, MA 02254-9110, USA b

Received 19 June 2000; received in revised form 12 September 2000; accepted 6 November 2000

Abstract The present research investigated whether competition in¯uences children's artistic creativity and intrinsic motivation toward an art activity. Study 1 tested the hypothesis that boys' creativity would be enhanced by competition, while girls' creativity would be undermined. Fifty children (aged 6±10) made paper collages in one of two conditions; half competed for prizes and half did not. Results supported our hypotheses, and further showed that when children self-segregated by gender, the impact of competition was much more pronounced. Study 2 was designed to clarify the unexpected gender-segregation ®nding from Study 1. The Children's Sex Role Inventory [Boldizar, J.P. (1991). Assessing sex typing and androgyny in children: the Children's Sex Role Inventory. Developmental Psychology, 27, 505±515] was administered to 143 children (aged 6±11). One week later, these children made paper collages in one of four conditions; in addition to manipulating competition, assigned seating ensured that half of children were segregated by gender and half were not. Following the collage activity, an intrinsic and extrinsic motivation questionnaire was administered. Masculine children reported higher levels of intrinsic motivation when competing and when segregated by gender; they also reported higher levels of extrinsic motivation, especially when segregated by gender. These ®ndings demonstrate that gender role is an important factor in determining children's responses to competition. # 2001 Elsevier Science Ltd. All rights reserved. Keywords: Creativity; Competition; Intrinsic motivation; Gender; Gender role; Gender segregation; Masculinity; Children's artwork

* Corresponding author. Tel.: +1-315-228-7661; fax: +1-315-228-7942. E-mail address: [email protected] (R. Conti). 0191-8869/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved. PII: S0191-8869(00)00217-8

1274

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

From science fairs to spelling bees, children's e€orts are often propelled by the desire to win against others. The value of and potential problems with this emphasis on competition has been a frequent topic of psychological inquiry (Garza & Borchert, 1990; Johnson & Engelhard, 1992; Knight & Chao, 1989; McGlynn, Gibbs, & Roberts, 1982; Vallerand, Gauvin, & Halliwell, 1986; Weinberg & Ragan, 1979). Although children are easily engaged by nearly any activity posed as a competitive game, there is concern that their energy becomes too focused on winning at the cost of other motivating aspects of the activity (Kohn, 1985; Nicholls, 1989). Because losers typically outnumber winners, competition may have negative motivational aftere€ects, as well (Reeve & Deci, 1996; Reeve, Olson, & Cole, 1987). Amabile (1982a) was the ®rst to propose that competition can have a negative impact on creativity. In an initial study to investigate this idea, 21 girls, aged 7±11, made paper collages at one of two ``parties.'' At one party, prizes were awarded for the three best collages, at the other they were ra‚ed o€. Artists later rated collages for creativity. Those collages produced by girls who were at the party where prizes were awarded were found to be signi®cantly less creative than those produced by girls at the party where prizes were ra‚ed o€ (Amabile). This study, now a classic, is often cited as evidence for the damaging e€ects of competition (Deci & Ryan, 1985; Kohn, 1985; Nicholls, 1989). In addition, these ®ndings were the ®rst to provide support for a more general notion: Amabile's intrinsic motivation hypothesis of creativity. This hypothesis, which now has considerable empirical support, states that intrinsic motivation (derived from enjoyment inherent in a task) is conducive to creativity, while extrinsic motivation (fueled by a goal separate from the task) is detrimental (Amabile, 1979, 1983, 1993, 1996; Amabile & Gitomer, 1984; Amabile, Hennessey, & Grossman, 1986; Ruscio, Whitney, & Amabile, 1998). The intrinsic motivation hypothesis is consistent with a large body of research showing that extrinsic motivators can have negative impact on later intrinsic motivation and qualitative aspects of performance (Deci, 1975; Deci, Koestner, & Ryan, 1999; Deci & Ryan, 1985; Lepper, Greene, & Nisbett, 1973; McGraw, 1978). Competition also has been found to undermine intrinsic interest (Deci, Betley, Kahle, Abrams, & Porac, 1981; Vallerand et al., 1986). Recent theoretical accounts, however, propose that the impact of extrinsic motivators on intrinsic motivation and creativity can vary depending on how the extrinsic motivator is interpreted (Amabile, 1993, 1996; Epstein & Harackiewicz, 1992; Kohn, 1996; Lepper, Keavney, & Drake, 1996; Ryan & Deci, 1996). In addition, empirical studies have sometimes shown positive e€ects of competition (McGlynn et al., 1982; Weinberg & Ragan, 1979). Extrinsic incentives may actually boost intrinsic motivation and creativity when they provide positive competence information or when they make the task more interesting or challenging (Csikszentmihalyi, 1975, 1990). It seems that the negative e€ects of extrinsic incentives appear when they are perceived as constraining, controlling, or when they are accompanied by negative competence information (Deci & Ryan, 1985; Deci et al., 1999; Shalley & Oldham, 1997). While most studies have not found gender di€erences in responses to extrinsic incentives, there are exceptions (Koestner, Zuckerman, & Koestner, 1989). Deci and Ryan (1985) suggested that because of traditional sex role socialization practices, men and boys may be more likely to feel challenged by extrinsic pressure, while women and girls may be more likely to feel controlled. There is reason to believe that this is especially true when the extrinsic incentive is competition. Research has consistently shown that boys tend to be more competitive and to feel more comfortable in competitive

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

1275

situations than girls (Croxton, Chiacchia, & Wagner, 1987; Geary, 1998; Gill, 1986). Given the choice between cooperative and competitive activities, girls choose cooperation more often than boys do (Garza & Borchert, 1990; Johnson & Engelhard, 1992; Knight & Chao, 1989). Although, much of the research on responses to competition has focused on athletic activities, these gender speci®c orientations seem to generalize to other activities as well. For example, themes of competitiveness are more common in the artwork of boys than of girls (Rubenstein, Feldman, Rubin, & Noveck, 1987). This gender di€erence in children's orientation toward competition may have important implications for their responses to it. Pallak, Costomiris, Sroka, and Pittman (1982) demonstrated that when children are accustomed to a particular type of performance feedback, they tend to respond more favorably to it and display heightened levels of intrinsic motivation. Girls and boys may also have di€erent expectancies for winning versus losing. The outcome of a competitive situation has important implications for its motivational impact (Reeve et al., 1987), and there is evidence suggesting that boys may have more positive expectancies (Corbin, 1981; Hall, 1990). A review of the literature by Weisfeld (1986) shows that when competition is made salient, girls and women are more likely to show decreases in performance than boys and men. These di€erential modes of responding are consistent with boys' and girls' sex-typed modes of interaction. Men are thought to be more independent, competitive and aggressive, while women are thought to be helpful, gentle and understanding (Eagly, Mladinic, & Otto, 1991; Williams & Bennett, 1975). If children are adhering to these cultural stereotypes, we would expect boys to respond with enthusiasm to a competitive situation, and girls to respond hesitantly. 1. Study 1 The purpose of the ®rst study was to examine the impact of competition on the intrinsic motivation and creativity of boys and girls. We used a paradigm similar to that used by Amabile (1982a). In the present study, however, both boys and girls participated. To test the intrinsic motivation hypothesis of creativity more directly, measures of intrinsic and extrinsic motivation toward the task were taken following the art activity. Hypotheses for the study were as follows: Hypothesis 1: For girls, competition would undermine intrinsic motivation and creativity, while for boys, competition would enhance intrinsic motivation and creativity. Hypothesis 2: Intrinsic motivation would be positively related to creativity and extrinsic motivation would be negatively related. 1.1. Method 1.1.1. Participants Fifty children (27 boys, 23 girls) ages 6±10 years that were enrolled in a summer day camp in the Boston area voluntarily participated in the study. 1.1.2. Materials Each child was provided with an identical set of collage materials: 88 colored pieces of paper in various shapes and sizes, a 912 piece of white cardboard, and a non-toxic, water-soluble glue stick.

1276

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

A questionnaire was developed to assess children's level of intrinsic and extrinsic motivation toward the art activity. Two items were intended to measure intrinsic motivation, ``How much fun did you have making your design?'' and ``How much would you like to make another paper design just for fun?'' Two were intended to measure extrinsic motivation, ``How much did you think about winning a prize while making your design?'' and ``How much did you want to make a design that was better than the other kids designs?'' Children were asked to respond on a 5-point scale; each point was labeled (from ``very little'' for 1 to ``a whole lot'' for 5). Because many of the younger children could not read, ®ve circles, of increasing size, were used to represent each scale point. 1.1.3. Procedure Six groups of 7±12 children participated in the study. The experiment was introduced to campers as an ``Art Party'' being given by university students interested in children's art. Participating children from each camp group were brought to an outdoor, pavilion equipped with two long picnic tables. Each child chose his or her own seat. Three groups were randomly assigned to the experimental condition (competition) and three groups to the control condition (non-competition). One experimenter then introduced the activity. She, then, displayed three attractive prizes (art materials). In the control condition, she told the children that these prizes would be ra‚ed o€ at the end of the party. In the experimental condition, the experimenter told the children that her assistants would judge the ®nished designs and that the three prizes would be awarded to the three best designs. Children were encouraged to ``do their own thing'' while making the designs, and were given 10 min to work on them. The motivation questionnaire was then distributed. Each question was read aloud, and children were asked to make an ``X'' in the circle that corresponded to the way they felt. The prizes initially displayed were awarded according to the procedure that was described by the experimenter in that session. Smaller prizes were then given to each child to thank them for participating. Two weeks after the experimental sessions, ®ve judges with experience in children's art rated the creativity of the collages according to Amabile's (1982b) consensual assessment technique. The collages were identi®ed only with numbers, and the judges were blind to experimental condition. Judges' ratings of creativity were found to be highly reliable, alpha=0.82. Therefore, the mean of the ®ve ratings for each collage served as average creativity score for that collage. 1.2. Results At two of the six experimental sessions (one in each condition), children self-segregated into a boys' table and a girls' table. At the remaining 4 sessions, both boys and girls were mixed at each table. Because gender di€erences were of interest in the study, sex segregation was included in the analyses as a possible interacting variable. Exploratory analyses of the motivation questionnaire revealed a severe positive skew on each of the 4 items. Overwhelmingly, children chose to mark the largest circle (``a whole lot''). It was unclear whether these results re¯ected ceiling levels of intrinsic and extrinsic motivation or simply children's preference for larger over smaller circles. Regardless, the variability was so limited that these measures could not be used in further analyses. Creativity assessments for the collages were analyzed using a 2 (gender)  2 (condition)  2 (gender segregation) analysis of variance. While there were no signi®cant main e€ects, there was a signi®cant 2way interaction between gender and condition, F(1,42)=4.22, P<0.05. The pattern of means was such

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

1277

that boys were more creative in the competition condition than in the non-competition condition, and girls were less creative in the competition condition than in the non-competition condition (Fig. 1). Planned t-tests, however, did not show signi®cant di€erences for boys across conditions [competition: M=4.09, S.D.=1.45; non-competition: M=3.52, S.D.=1.14; t(25)= 1.12, ns], or for girls [competition: M=3.75, S.D.=1.23; non-competition: M=4.36, S.D.=1.72; t(21)=0.99, ns]. A signi®cant 3-way interaction, F(1,42)=7.16, P<0.05, revealed that the 2-way interaction pattern was striking when children were segregated by sex, and did not appear when children were integrated by gender (Fig. 2). T-tests showed that girls were more creative in the non-competition condition than in the competition condition when segregated by gender [competition: M=4.10, S.D.=1.21; non-competition: M=6.07, S.D.=.31; t(7)=2.70, P<0.05], but not when they were at gender integrated tables [competition: M=3.40, S.D.=1.25; non-competition: M=3.73, S.D.=1.58; t(12)=0.42, ns]. Likewise, boys were more creative under competitive than under noncompetitive conditions when they were gender segregated, [competition: M=4.70, S.D.=1.69; non-competition: M=2.60, S.D.=.54; t(8)= 2.36, P<0.05], but not when they were gender integrated, [competition: M=3.63, S.D.=1.13; non-competition: M=3.93, S.D.=1.11; t(15)=0.57, ns]. 1.3. Discussion The results of Study 1 provide support for our ®rst hypothesis. Girls were less creative when competing, while boys were more creative. These di€erential modes of responding are consistent with boys' and girls' sex-typed modes of interaction (Bem, 1974; Geary, 1998; Jacklin & Maccoby, 1978; Maccoby, 1990). Based on the creativity of their collages, it appears that gender linked tendencies were in¯uencing children's feelings about the experimental situation. These results have important implications for the intrinsic motivation hypothesis of creativity. Although competition, in this case, represented a goal that was clearly separate from the activity

Fig. 1. Average artistic creativity scores for boys and girls under competitive and non-competitive conditions in Study 1.

1278

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

Fig. 2. Average artistic creativity scores at gender-segregated and gender-integrated experimental sessions for boys and girls under competitive and non-competitive conditions in Study 1.

itself, this extrinsic motivator did not undermine the creativity of all children. For boys, competing may have added to the interest value and challenge of the activity, making it more intrinsically enjoyable. This higher level of intrinsic motivation could account for the higher level of creativity displayed by boys in the competition condition. Unfortunately, the problems with the motivation measure make it impossible to assess the impact of competition on boys' and girls' intrinsic motivation, or the relations between motivation and creativity.

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

1279

Despite this disappointment, the results of Study 1 were quite exciting because they raised a new and intriguing issue to explore. When girls and boys segregated themselves by gender, the interaction was quite pronounced; when girls and boys were gender-integrated, they did not show marked di€erences in their responses to competition. This ®nding is consistent with research showing that people will display more masculine behavior when interacting with males, and more feminine behavior when interacting with females (Carli, 1989; Hartup, 1983; Moskowitz, Suh, & Desaulniers, 1994). If the masculine response to competition is to consider it a challenge, enjoy it and arrive at a higher quality product, then the boys showed this response more markedly when they were interacting with only other boys. If the feminine response is to ®nd the competition aversive and distracting, thus leading to a lower quality of response, then girls showed this response most markedly when interacting only with other girls. Groups of boys are more likely to orient their interactions toward competition and dominance, an orientation that girls seem to ®nd aversive (Maccoby, 1988). That boys are more competitive when with other boys may have something to do with the fact that they are more concerned with the reactions of their male peers. Fagot (1985) found that boys are more likely to modify their behavior in response to the reaction of a boy than in response to a girl. The competition manipulation apparently facilitated a situation that was natural and engaging for these boys, as compared with the non-competition manipulation. In contrast with the confrontational style of boys, girls tend to support and encourage one another in their social interactions (Maccoby, 1990; Maccoby & Jacklin, 1987). Competition is not part of the social interaction style adopted by groups of females (Maccoby, 1990). And, women are more communal and less likely to respond competitively when interacting with other women, perhaps because they expect men to be more competitive (King, Miles, & Knisha, 1991; Moskowitz et al., 1994). For example, in the context of a debate, women expressed more disagreement when paired with a male, and more agreement and positive social behavior when paired with a female (Carli, 1989). Thus, when girls segregated by gender in Study 1, the noncompetition manipulation may have set-up a social situation that was comfortable for these girls and, thus, facilitated their involvement in the task. The competition manipulation, on the other hand, may have disrupted the ordinary interaction style of all-girl groups and, in doing so, interfered with their performance. Although it is interesting to speculate about why boys and girls responded so di€erently when gender-segregated, it is dicult to reach any ®rm conclusions. Because children self-segregated, the results may be accounted for by the gender make-up of the groups, the gender role orientation of the children who chose to segregate or an interaction of the two. Because our motivation measure could not be used, it was not possible to explore the mediating e€ects of intrinsic and extrinsic motivation. And, the small sample size limited the statistical reliability of the results. For these reasons, a second study was conducted. 2. Study 2 Study 2 was designed to further explore the ways in which gender in¯uences children's responses to competition. The same ``Art Party'' paradigm was used; several methodological improvements were made and additional measures were taken to clarify and extend the results of

1280

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

the ®rst study. Gender-segregation was manipulated by assigning children seats at the experimental sessions. To investigate the possibility that the characteristics of the children who gender-segregated in¯uenced their responses to competition, gender role orientation was measured using the Children's Sex Role Inventory (CSRI; Boldizar, 1991). Children's di€erential responses to competition when in single-sex groups seems to be accounted for by an increased tendency toward gender stereotyped behavior. Children who identify more strongly with gender-stereotyped roles also show this tendency (King et al., 1991). Although investigations of gender-segregation have not found the tendency to segregate to be linked with sex-typed activity preferences or interaction styles (Maccoby, 1990; Powlishta, Serbin, & Moller, 1993), both variables would seem to have a similar impact on behavior. Therefore, based on the ®ndings of the ®rst study, we suspected that gender role orientation would in¯uence children's responses to competition, such that more masculine children would be more intrinsically motivated and creative when competing and feminine children would be less intrinsically motivated and creative when competing. Research has shown that although most people experience some anxiety when competing, this anxiety seems to be less pronounced for highly masculine individuals, especially when those individuals are men (Alagna, 1982; Swain & Jones, 1991; Wittig, 1984). To summarize, the purposes of Study 2 were to attempt to replicate the support for hypothesis 1 found in Study 1. Study 2 was also intended to test hypothesis 2 from Study 1, the intrinsic motivation hypothesis of creativity: intrinsic motivation was expected to be positively related to creativity and extrinsic motivation was expected to be negatively related. In addition, Study 2 was designed to test 2 additional hypotheses: Hypothesis 3: When children are gender segregated, girls will show lower levels of intrinsic motivation and creativity when faced with competition, while boys will show higher levels of intrinsic motivation and creativity. Hypothesis 4: More masculine children will show higher levels of intrinsic motivation and creativity under competitive conditions, while feminine children will show lower levels of intrinsic motivation and creativity under competitive conditions. 2.1. Method 2.1.1. Participants Subjects were 143 children (71 boys and 72 girls) ages 6±11 years (1st through 5th graders) enrolled at a public elementary school in the Boston area. Children participated with their classmates for both portions of the study. 2.1.2. Materials The short (30-item) version of the CSRI was used to assess masculinity and femininity (Boldizar, 1991). Both scales show good reliability (average alpha=0.81), and stable test-retest reliability (average coecient=0.60). Validity has been demonstrated with signi®cant di€erences on the scales between boys and girls and has been shown to correlate with children's toy and play preferences (Boldizar). To facilitate administration for children unable to read, each point on the response scale was accompanied by a small circle. For the ``not at all true of me'' option, the circle was not at all ®lled in, for the ``entirely true of me'' option the circle was entirely ®lled in,

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

1281

and for the two middle options the circle was partially ®lled. Art materials were identical to those used in the ®rst study. The motivation questionnaire was revised to avoid the response bias encountered in Study 1. The visual aide used was the same as that used for the CSRI. A small circle represented each scale point and the degree to which the circle was ®lled in corresponded to the amount of the response. For example, the circle indicating ``very little'' was only slightly ®lled (10%), while the ``a whole lot'' circle was entirely ®lled. 2.1.3. Procedure All children responded to the CSRI with their entire class one week prior to the experimental sessions. For the experimental sessions, each class was divided into two groups of 10±12 children. Each group was then randomly assigned to one of four experimental conditions: (1) competition/genderintegrated, (2) competition/gender-segregated, (3) non-competition/gender-integrated; and (4) non-competition/gender-segregated. The experimental sessions were introduced to children as ``Art Parties.'' Each session was held in a room with two long tables. At gender-integrated parties both boys and girls were seated at each table, and at gender-segregated parties only boys were seated at one table and only girls were seated at the other. The sessions were otherwise identical to those conducted for Study 1. As in Study 1, the ®ve judges showed good reliability in their assessments of creativity (alpha=0.75), and their scores were averaged to form a creativity score for each collage. 2.2. Results Results from the motivation questionnaire showed substantially more variability than they did in Study 1. The two intrinsic items were positively correlated: r (143)=0.51, P<0.001, as were the two extrinsic items: r (143)=0.34, P<0.001. Therefore, these item pairs were averaged to form measures of intrinsic and extrinsic motivation toward the collage activity. Results from the CSRI con®rmed that gender role is moderately related to gender. A t-test showed that boys reported a signi®cant, but not substantially higher degree of masculinity than girls did [boys' M=3.02, S.D.=0.50; girls' M=2.81, S.D.=0.52; t(131)= 2.40, P<0.05]. A second t-test showed that girls reported a much higher degree of femininity than boys reported [boys' M=2.97, S.D.=0.52; girls' M=3.42, S.D.=0.47; t(131)=0.533, P<0.001]. To check the e€ectiveness of the competition manipulation, t-tests were conducted comparing children's responses to the two extrinsic items in the competition condition and in the non-competition condition. In the competition condition, children were signi®cantly more likely to indicate that they were trying to make a design that was better than the other kids' designs [t(141)=2.98, P<0.005]. Children in the two conditions were not signi®cantly di€erent in the degree to which they thought about winning a prize. This indicates that the manipulation was e€ective in manipulating the degree to which children were focused on competing (doing better than others), and that this competition orientation was not confounded with the degree to which children were reward focused (wanting to win a prize). To test hypothesis 2, that intrinsic motivation would have a positive impact on creativity and extrinsic motivation would have a negative impact, hierarchical regression analyses were conducted.

1282

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

For all regression analyses, variables measured on arbitrary scales were converted to z-scores to aide in interpretation of the results and to avoid multicolinearity. On the ®rst step of the analysis, the background variables of age (in months) and gender were entered. Age was related to creativity (beta=0.27, t=3.17, P<0.01), with older children producing more creative collages than younger children. Gender did not predict creativity. The overall equation was signi®cant [Multiple R=0.28, F(2,126)=5.15, P<0.01]. On the next step intrinsic and extrinsic motivation were entered. Intrinsic motivation was a signi®cant predictor (beta=0.18, t=2.12, P<0.05); extrinsic motivation was not (beta=0.09, t= 0.81, ns). The ®nal equation accounted for a signi®cant proportion of the variance [Multiple R=0.33, F(4,124)=3.85, P<0.05]. Thus, partial support for the intrinsic motivation hypothesis of creativity was observed: intrinsic motivation was found to be conducive to creativity, but extrinsic motivation was found to be unrelated to creativity. To test hypothesis 3, that gender segregation would in¯uence boys' and girls' responses to competition, a series of 2 (competition)  2 (gender segregation)  2 (gender) analyses of variance were carried out. Dependent variables were intrinsic motivation, extrinsic motivation and creativity. No signi®cant e€ects were observed in these analyses. These null results were especially surprising because they failed to replicate the gender by competition interaction on creativity observed in Study 1 [F(1, 133)=0.09, ns]. Speci®cally, neither the negative impact of competition on girls' creativity observed both by Amabile (1983) and in Study 1 nor the positive impact of competition on boys' creativity observed in Study 1, were observed in Study 2. To test hypothesis 4, that gender role orientation would interact with experimental condition to in¯uence children's motivation and creativity, additional hierarchical regression analyses were conducted. These equations predicted intrinsic and extrinsic motivation1 from: (1) background variables (gender and age); (2) gender role orientation (masculinity and femininity); (3) experimental conditions (competition and gender segregation); and (4) the interactions between gender role and experimental condition (masculinitycompetition, masculinitygender segregation, femininitycompetition and femininitygender segregation). Each category of variables was entered on a separate step. For the intrinsic motivation equation, gender was a marginally signi®cant predictor when entered, with girls expressing more intrinsic motivation than boys (beta= 0.16, t= 1.86, P=0.06). Two interactions were signi®cant. The interaction between masculinity and competition showed the predicted pattern: highly masculine children were more intrinsically motivated when competing, while children low in masculinity were less intrinsically motivated when competing (beta= 0.66, t= 2.06, P<0.05). This interaction is illustrated in Fig. 3. The interaction between masculinity and gender segregation showed that children high in masculinity were more intrinsically motivated when seated with their own gender than when at a table with both genders (see Fig. 4). The ®nal equation was signi®cant [Multiple R=0.39, F(11,117)=1.90, P<0.05]. For the extrinsic motivation equation, both gender and age were signi®cant when entered (for gender: beta=0.19, t=2.28, P<0.05; for age: beta= 0.33, t= 4.00, P<0.001). Boys were more extrinsically motivated than girls, and older children were less extrinsically motivated than younger children. Masculinity was also signi®cant when entered (beta=0.33, t=3.63, P<0.001). The interaction between masculinity and gender-segregation was also signi®cant and is illustrated in Fig. 5 (beta= 0.63, t= 2.51, P<0.05). Children who were low in masculinity expressed less 1 The same analysis was also carried out with creativity as the dependent variable. No signi®cant direct or interactive e€ects of gender role were observed.

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

1283

Fig. 3. Intrinsic motivation for children low, average and high in masculinity under competitive and non-competitive conditions.

extrinsic motivation when gender-segregated, while children high in masculinity expressed more extrinsic motivation when gender-segregated. The ®nal equation was signi®cant (Multiple R=0.53, F(11,117)=4.22, P<0.001). 2.3. Discussion The support for hypothesis 1, that girls and boys would respond di€erentially to competition, found in Study 1 was not replicated in Study 2. Neither intrinsic motivation nor creativity showed a signi®cant gender by competition interaction. Based on the results of Study 2, it seems as though the signi®cant interaction observed in Study 1, was the result of the degree of masculinity of the boys and girls who chose to segregate themselves by gender. While this interpretation ®ts well with the ®ndings reported here, additional research is needed to con®rm that the gender-role composition of children's playgroups will in¯uence boys' and girls' reactions to competition. The results of Study 2 further show that gender segregation alone does not lead to di€erent patterns of responding for boys and girls. Gender segregation, as a manipulated variable did not in¯uence how boys and girls responded to competition. Perhaps, in order for gender segregation to have the e€ect that was observed in Study 1, the groups of children need to be more cohesive groups of friends who have chosen to sit together. Maccoby (1990) explains the distinct patterns of responding observed in all-boy and all-girl play groups as resulting from a higher degree of social interaction. Because the experimenters assigned seats and the study took place in a more formal school setting, children may not have had as much of a group feeling in the second study.

1284

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

Fig. 4. Intrinsic motivation for children low, average and high in masculinity under gender-segregated and gender integrated conditions.

The interactions between masculinity and experimental condition suggest that masculinity plays an important role in determining responses to competition. Highly masculine children were more intrinsically motivated when competing; they reported more intrinsic and extrinsic motivation when in gender segregated groups. Children low in masculinity were less intrinsically motivated when competing; they reported less intrinsic and extrinsic motivation when in gender segregated groups. Possibly, the groups who gender segregated in Study 1 had boys who were high in masculinity and girls who were low in masculinity. By manipulating gender-role composition of the groups, a future study can examine this possibility. According to the results of the second study, children's level of masculinity is central in determining their response to competition. The interactions observed provide partial support for our fourth hypothesis. Counter to predictions, femininity did not have an e€ect on children's responses. The greater in¯uence of masculinity, as compared with femininity on behavior has been observed in research assessing the validity of the CSRI (Boldizar, 1991). CSRI masculinity scores were found to be positively associated with children's global sense of self-worth, scholastic competence, social acceptance, athletic competence and physical attractiveness. Only perceptions of behavioral competence were associated with femininity scores. Study 2 provided partial support for the intrinsic motivation hypothesis of creativity; intrinsic motivation was positively related to creativity, although extrinsic motivation was not negatively related to creativity. This result is consistent with other recent work showing that although intrinsic motivation and creativity are closely linked in a wide variety of circumstances, the impact of extrinsic incentives on intrinsic motivation and creativity depend on how the incentive

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

1285

Fig. 5. Extrinsic motivation for children low, average and high in masculinity under gender-segregated and gender integrated conditions.

is interpreted (Amabile, 1993, 1996). Competition, it appears, can be interpreted as adding to the excitement and challenge of an activity, thus heightening intrinsic motivation. Or, it can cause the pressure and tension that reduce intrinsic motivation toward an activity. Because the contextual factors manipulated in this study did not have main e€ects on intrinsic motivation, the intrinsic motivation hypothesis could not be tested experimentally, but did receive correlational support. In sum, these results extended those of Study 1 in several important respects. Because neither central ®nding of Study 1 was directly replicated in Study 2, we know that the in¯uence of gender on children's response to competition is dependent on the degree of masculinity the children display. Masculinity also appears to in¯uence how children respond to the gender composition of a group, with more masculine children feeling more highly motivated when with same gender peers. Piecing together the mechanisms by which masculinity produces these e€ects is an important problem for future research. 3. General discussion This research began with two seemingly simple questions: Will the undermining e€ect of competition on girls' artistic creativity that was observed in Amabile (1982a) occur with boys? And, is the impact of competition on motivation mediated by its e€ect on intrinsic and extrinsic motivation as the intrinsic motivation hypothesis predicts? The results of the two studies conducted,

1286

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

although not as straightforward as the questions posed, provide some insight into these issues. As the research progressed, additional questions arose: Does gender segregation in¯uence children's responses to competition? Does gender role in¯uence children's responses to competition? The results from the second study help to clarify these issues. As we initially suspected, competition does not seem to undermine boys' artistic creativity or their intrinsic motivation toward an art activity. Results from the ®rst study suggested that boys were actually more creative when competing. However, because this result was observed only at gender-segregated experimental sessions, and because it did not replicate in the second study, it is clearly not generalizable. In order for competition to enhance intrinsic motivation and creativity, a speci®c social situation must be present. These ®ndings suggest that this social situation somehow supports children's sense of masculinity or agentic/task oriented responding. More surprising, perhaps, was the ®nding that competition does not always undermine girls' intrinsic motivation and creativity. In Study 1, girls' creativity was undermined only when they were in gender segregated groups. Girls were not di€erentially a€ected by competition in Study 2. The set of circumstances present in Amabile's (1983) study may have combined to cause competition to lead to lowered levels of creativity in girls. If the setting were somewhat di€erent, these same results may not have emerged. The Amabile study was conducted nearly two decades before the present research. It is possible that, because of changes in our culture, the girls who participated were less masculine than girls of the same age today. Also, because only girls participated in the study, the sessions were, unavoidably, gender segregated. Interacting only with other girls may have led to an even more gender stereotyped pattern of responding. Thus, the results of the present research are reassuring in that they suggest that, in most circumstances, girls are not negatively a€ected by competition. This study adds to the substantial body of evidence behind a revised understanding of the intrinsic motivation hypothesis of creativity. Intrinsic motivation has been consistently found to be positively related to creativity. But, in certain situations, extrinsic motivation has not been observed to be negatively related to creativity. When extrinsic incentives are enabling present challenge or provide positive competence information they can contribute to intrinsic motivation and creativity (Amabile, 1991, 1996; Deci & Ryan, 1985; Deci et al., 1999). Amabile (1983) has also speculated that an individual's personality characteristics may in¯uence how he or she responds to an extrinsic motivator. The present ®ndings suggest that masculinity is one personality characteristic that in¯uences responses to competition, and that competition may facilitate intrinsic motivation and creativity for those that are highly masculine. Acknowledgements We wish to thank the day camp directors, counsellors and campers from the day camp where Study 1 took place, the principal, teachers and students of the Boston-area elementary school where this study was conducted and our undergraduate research assistants from Brandeis University for their enthusiastic participation in the research. We are grateful to our colleagues who have provided useful feedback on earlier versions of this manuscript; in alphabetical order, they are: Teresa M. Amabile, Heather Coon, John F. Dovidio, Caroline Keating, Gail Martino, and Elise Phillips. This study was supported by a grant from the Sachar Foundation of Brandeis University to Mary Ann Collins and Regina Conti.

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

1287

References Alagna, S. W. (1982). Sex role identity, peer evaluation of competition, and the responses made by males and females in a competitive situation. Journal of Personality and Social Psychology, 43, 546±554. Amabile, T. M. (1979). E€ects of external evaluation on artistic creativity. Journal of Personality and Social Psychology, 37, 221±233. Amabile, T. M. (1982aa). Children's artistic creativity: detrimental e€ects of competition in a ®eld setting. Personality and Social Psychology Bulletin, 8, 573±578. Amabile, T. M. (1982bb). Social psychology of creativity: a consensual assessment technique. Journal of Personality and Social Psychology, 43, 997±1013. Amabile, T. M. (1983). The social psychology of creativity. New York: Springer-Verlag. Amabile, T. M. (1991). Perspectives on research classics: motivations to study motivation. Contemporary Social Psychology, 15, 57±59. Amabile, T. M. (1993). Motivational synergy: toward new conceptualizations of intrinsic and extrinsic motivation in the workplace. Human Resource Management Review, 3, 185±201. Amabile, T. M. (1996). Creativity in context: update to the social psychology of creativity. Boulder, CO: Westview Press. Amabile, T. M., & Gitomer, J. (1984). Children's artistic creativity: e€ects of choice in task materials. Personality and Social Psychology Bulletin, 10, 209±215. Amabile, T. M., Hennessey, B. A., & Grossman, B. S. (1986). Social in¯uences on creativity: the e€ects of contractedfor reward. Journal of Personality and Social Psychology, 50, 14±23. Bem, S. L. (1974). The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 42, 155±162. Boldizar, J. P. (1991). Assessing sex typing and androgyny in children: the Children's Sex Role Inventory. Developmental Psychology, 27, 505±515. Carli, L. L. (1989). Gender di€erences in interaction style and in¯uence. Journal of Personality and Social Psychology, 56, 565±576. Corbin, C. B. (1981). Sex of subject, sex of opponent, and opponent ability as factors a€ecting self-con®dence in a competitive situation. Journal of Sport Psychology, 4, 265±270. Croxton, J. S., Chiacchia, D., & Wagner, C. (1987). Gender di€erences in attitudes toward sports and reactions to competitive situations. Journal of Sports Behavior, 10, 167±177. Csikszentmihalyi, M. (1975). Beyond boredom and anxiety. San Francisco, CA: Jossey-Bass. Csikszentmihalyi, M. (1990). Flow: the psychology of optimal experience. New York: Harper Perennial. Deci, E. L. (1975). Intrinsic motivation. New York: Plenum Press. Deci, E. L., Betley, G., Kahle, J., Abrams, L., & Porac, J. (1981). When trying to win: competition and intrinsic motivation. Personality and Social Psychology Bulletin, 7, 79±83. Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. New York: Plenum. Deci, E. L., Koestner, R., & Ryan, R. M. (1999). A meta-analytic review of experiments examining the e€ects of extrinsic rewards on intrinsic motivation. Psychological Bulletin, 125(6), 627±668. Eagly, A. H., Mladinic, A., & Otto, S. (1991). Are women evaluated more favorably than men? An analysis of attitudes, beliefs, and emotions. Psychology of Women Quarterly, 15, 203±216. Epstein, J. A., & Harackiewicz, J. M. (1992). Winning is not enough: the e€ects of competition and achievement orientation on intrinsic interest. Personality and Social Psychology Bulletin, 18, 128±138. Fagot, B. I. (1985). Beyond the reinforcement principle: another step toward understanding sex roles. Developmental Psychology, 21, 1097±1104. Garza, R. T., & Borchert, J. E. (1990). Maintaining social identity in a mixed-gender setting: minority/majority status and cooperative/competitive feedback. Sex Roles, 22, 679±691. Geary, D. C. (1998). Male, female: the evolution of human sex di€erences. Washington, DC: American Psychological Association. Gill, D. L. (1986). Competitiveness among females and males in physical activity classes. Sex Roles, 15, 233±247. Hall, E. G. (1990). The e€ect of performer gender, performer skill level, and opponent gender on self-con®dence in a competitive situation. Sex Roles, 23, 33±41.

1288

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

Hartup, W. W. (1983). Peer relations. In P. H. Mussen, & E. M. Hetherington, Handbook of child psychology. Vol. 3: Socialization, personality, and development (pp. 103±196). New York: Wiley. Jacklin, C. N., & Maccoby, E. E. (1978). Social behavior at 33 months in same-sex and mixed-sex dyads. Child Development, 49, 557±569. Johnson, C., & Engelhard, G. (1992). Gender, academic achievement, and preferences for cooperative, competitive, and individualistic learning. Journal of Psychology, 126, 385±392. King, W. C., Miles, E. W., & Kniska, J. (1991). Boys will be boys (and girls will be girls): the attribution of gender role stereotypes in a gaming situation. Sex Roles, 25, 607±623. Knight, G. P., & Chao, C. (1989). Gender di€erences in the cooperative, competitive, and individualistic social values of children. Motivation and Emotion, 13, 125±141. Koestner, R., Zuckerman, M., & Koestner, J. (1989). Attributional focus of praise and children's intrinsic motivation: the moderating role of gender. Personality and Social Psychology Bulletin, 15, 61±72. Kohn, A. (1985). No contest: the case against competition. Boston, MA: Houghton Mi‚in. Kohn, A. (1996). By all available means: Cameron and Pierce's defense of extrinsic motivators. Review of Educational Research, 66, 1±4. Lepper, M. R., Greene, D., & Nisbett, R. E. (1973). Undermining children's intrinsic interest with extrinsic reward: a test of the ``overjusti®cation'' hypothesis. Journal of Personality and Social Psychology, 28, 129±137. Lepper, M. R., Keavney, M., & Drake, M. (1996). Intrinsic motivation and extrinsic rewards: a commentary on Cameron and Pierce's meta-analysis. Review of Educational Research, 66, 5±32. Maccoby, E. E. (1988). Gender as a social category. Developmental Psychology, 26, 755±765. Maccoby, E. E. (1990). Gender and relationships: a developmental account. American Psychologist, 45, 513±520. Maccoby, E. E., & Jacklin, C. N. (1987). Gender segregation in childhood. In H. W. Reese, Advances in child development and behavior, Vol. 20 (pp. 239±288). New York: Academic Press. McGlynn, R. P., Gibbs, M. E., & Roberts, S. J. (1982). E€ects of cooperative versus competitive set and coaction on creative responding. Journal of Social Psychology, 118, 281±282. McGraw, K. (1978). The detrimental e€ects of reward on performance: a literature review and a prediction model. In M. Lepper, & D. Greene (Eds.), The hidden costs of reward: new perspectives on the psychology of human motivation (pp. 33±60). Hillsdale, NJ: Lawrence Erlbaum. Moskowitz, D. S., Suh, E. J., & Desaulniers, J. (1994). Situational in¯uences on gender di€erences in agency and communion. Journal of Personality and Social Psychology, 66, 561±573. Nicholls, J. (1989). The democratic ethos and competitive education. Cambridge, MA: Harvard University Press. Pallak, S. R., Costomiris, S., Sroka, S., & Pittman, T. S. (1982). School experience, reward characteristics, and intrinsic motivation. Child-Development, 53, 1382±1391. Powlishta, K. K., Serbin, L. A., & Moller, L. C. (1993). the stability of individual de€erence in gender typing: implications fro understanding gender segregation. Sex Roles, 29, 723±737. Reeve, J., & Deci, E. L. (1996). Elements of the competitive situation that a€ect intrinsic motivation. Personality and Social Psychology Bulletin, 22, 24±33. Reeve, J., Olson, B. C., & Cole, S. G. (1987). Intrinsic motivation in competition: the intervening role of four individual di€erences following objective competence information. Journal of Research on Personality, 21, 148±170. Rubenstein, J., Feldman, S. S., Rubin, C., & Noveck, I. (1987). A cross-cultural comparison of children's drawings of same and mixed-sex peer interactions. Journal of Cross-Cultural Psychology, 21, 148±170. Ruscio, J., Whitney, D. M., & Amabile, T. M. (1998). Looking inside the ®shbowl of creativity: Verbal and behavioral predictors of creative performance. Creativity Research Journal, 11, 243±263. Ryan, R. M., & Deci, E. L. (1996). When paradigms clash: comments on Cameron and Pierce's claim that rewards do not undermine intrinsic motivation. Review of Educational Research, 66, 33±38. Shalley, C. E., & Oldham, G. R. (1997). Competition and creative performance: e€ects of competitor presence and visibility. Creativity Research Journal, 19, 337±345. Swain, A., & Jones, G. (1991). Gender role endorsement and competitive anxiety. International Journal of Sport Psychology, 22, 50±65. Vallerand, R. J., Gauvin, L. I., & Halliwell, W. R. (1986). Negative e€ects of competition on children's intrinsic motivation. Journal of Social Psychology, 126, 649±657.

R. Conti et al. / Personality and Individual Di€erences 30 (2001) 1273±1289

1289

Weinberg, R. S., & Ragan, J. (1979). E€ects of competition, success/failure, and sex on intrinsic motivation. Research Quarterly, 50, 503±510. Weisfeld, C. C. (1986). Females' behavior in mixed-sex competition: a review of the literature. Developmental Review, 6, 278±299. Williams, J. E., & Bennett, S. M. (1975). The de®nition of sex stereotypes via the adjective check list. Sex Roles, 1, 327± 337. Wittig, A. F. (1984). Sport competition anxiety and sex role. Sex Roles, 10, 469±473.