Cross-Group Contact Predicts Positive Beliefs About Girls’ and Black Peers’ STEM Abilities and Occupational Prospects

Abstract

High-quality forms of contact, such as cross-group play and friendships, have been identified as particularly effective for promoting positive beliefs toward outgroup peers. Relations between children’s cross-group play experiences and their beliefs about peers’ math and science competencies and high-status occupational prospects have not yet been examined. Understanding these relations is important given that children from minoritized groups continue to face exclusion and bias in these domains. The present study examined the associations between children’s (N = 983, M_age = 9.64, SD_age = .89) reported cross-group play experiences and their math and science competency beliefs and high-status occupation expectations about girls and Black peers. Results revealed that, for majority group participants (i.e., boys and White children), higher levels of cross-group play were associated with significantly higher beliefs and expectations for girls and Black peers. Further, results demonstrated contexts in which higher levels of cross-group play were positively associated with girls’ and Black children’s beliefs and expectations for their own groups. Together, these findings advance theory and research on the benefits of cross-group contact in childhood by highlighting novel outcomes to which cross-group contact is positively related, as well as by showing that children from both minoritized and majority status groups stand to benefit from cross-group contact.

Children’s peer relationships profoundly influence their social development by providing them with opportunities to learn about and understand the social world (Bukowski et al., 2011; Rubin et al., 2006). Several factors shape children’s friendship and playmate choices, but one of the most salient factors is similarity, particularly across group memberships (e.g., gender, ethnicity, race) (Graham & Echols, 2018; Martin & Fabes, 2001). Children also befriend and play with peers from different social groups, and a robust body of research highlights the unique benefits that this type of intergroup contact affords children. Specifically, high-quality, positive intergroup contact is associated with decreased intergroup prejudice and bias in childhood (Skinner & Meltzoff, 2019; Tropp & Prenvost, 2008). Intergroup contact has also been associated with children’s and adolescents’ negative evaluations of race-based exclusion (Crystal et al., 2008; Ruck et al., 2011, 2015), decreased stereotype usage (Killen et al., 2010), and prosocial bystander behaviors in response to group-based social exclusion (Palmer et al., 2017, 2022).

Despite the wealth of evidence demonstrating the positive outcomes associated with intergroup contact, research has not yet examined how children’s gender and racial intergroup contact relates to beliefs about peers’ math and science competencies and high-status occupational prospects. Associations between intergroup contact and these specific beliefs are important to examine because stereotypes about math and science abilities and high-status occupations emerge early in childhood. Children as young as six years of age have been found to associate math with boys more than with girls (Cvencek et al., 2011) and associate brilliance with White, but not Black, men (Jaxon et al., 2019). The consequences of being a recipient of these stereotypes also emerge early, with girls and ethnic and racial minority youth demonstrating lower math and science belonging and engagement (Cheryan et al., 2015; Master et al., 2016; Wang et al., 2022), and decreased likelihood of envisioning themselves as scientists or other high-status occupations (Bigler et al., 2003; Teig & Susskind, 2008).

These stereotypic beliefs may persist into adulthood and contribute to the underrepresentation of women and ethnic and racial minoritized individuals in science, technology, engineering and mathematics (STEM) fields (Cheryan et al., 2017; McGee et al., 2016) along with other high-status occupations (Collins et al. 2017, Sekaquaptewa & Thompson, 2003). Given that the role of stereotypes in limiting minoritized individuals’ inclusion in these fields can be traced back to early development, it is essential to identify effective strategies for reducing these stereotypic beliefs in childhood. Previous developmental research suggests that intergroup contact may be one promising strategy. Thus, the present study aimed to investigate the associations between children’s gender and racial intergroup contact and beliefs about the math and science competencies and high-status occupational prospects of girls and Black peers, two historically underrepresented groups in these domains.

Theoretical Framework

The present study was motivated by two theoretical perspectives: intergroup contact theory and the social reasoning developmental model. Intergroup contact theory, an extension of Allport’s (1954) original “contact hypothesis,” posits that positive interactions between members of different social groups can reduce prejudice, particularly when the conditions of equal status, common goals, cooperation, and authority support are met (Pettigrew & Tropp, 2006). Intergroup contact theory has received widespread empirical support since its original formulation, and meta-analytic results demonstrate that intergroup contact generally corresponds with decreased intergroup prejudice (Pettigrew & Tropp, 2006; Tropp & Prenvost, 2008). Additionally, a review of developmental literature on childhood experiences associated with intergroup bias found that positive, cooperative intergroup contact was one of the experiences most reliably associated with the reduction of outgroup biases (Skinner & Meltzoff, 2019).

A particularly strong form of intergroup contact that contributes to prejudice reduction is cross-group friendships (Davies et al., 2011; Killen et al., 2022). Cross-group friendships provide children with opportunities to gain knowledge about different social groups, which can be used to disconfirm previously held stereotypic beliefs. The utility of cross-group friendships for reducing prejudice is supported by meta-analytic findings that interracial friendships are more strongly related to more positive intergroup attitudes than mere contact (Davies et al., 2011). The extensive research on intergroup contact has been integral to advances in the understanding of prejudice reduction and interventions aimed at reducing negative outgroup attitudes and improving intergroup relations.

Past intergroup contact theory research suggests that the link between increased contact and reduced intergroup prejudice is stronger among members of majority status groups compared to members of minority status groups (Tropp & Pettigrew, 2005; Tropp & Prenvost, 2008). Interpretations of this differential effect discuss the possibility that minority status group members do not perceive their intergroup interactions to fully meet the conditions required for positive intergroup contact, namely equal status and cooperation (Rutland & Killen, 2015; Tropp & Pettigrew, 2005). A limitation in the field is the lack of extensive research on how minoritized groups experience and interpret intergroup contact. Much of the research has focused on the presence or absence of contact and not how contact is experienced by minoritized individuals. Further, minoritized individuals remain an understudied population in this field.

Studies examining outcome measures beyond intergroup attitudes provide insight into the unique benefits of intergroup contact for minority status group members. For example, cross-ethnic friendships were found to buffer the negative effects of perceived ethnic discrimination on psychological well-being for South Asian British children in the United Kingdom (Bagci et al., 2014). An examination of the longitudinal effects of cross-race friendships in childhood demonstrated that more cross-race friendships were related to increases in peer support and decreases in peer victimization, above and beyond same-race friendships (Kawabata & Crick, 2011). Further, intergroup contact was related to ethnic-racial minority adolescents’ evaluations of the wrongfulness of racial exclusion (Ruck et al., 2011).

These studies, albeit few, suggest that intergroup contact can benefit children from minority status groups. These benefits appear to relate to outcomes of belonging, peer support, and well-being as well as intergroup attitudes and prejudice. This set of outcomes is especially salient when studying beliefs and expectations about STEM and high-status occupations, given that previous research has highlighted belonging, identity, and motivation as vital for the persistence and success of minoritized youth in these areas (Master, 2021; Mulvey et al., 2022, 2023). Investigating how girls’ and Black peers’ cross-group play experiences are related to their beliefs and expectations about their own groups presents a novel method for exploring the potential unique benefits of intergroup contact for children from these groups.

The evaluations that children and adolescents employ during intergroup interactions have been less examined. Thus, the present study was also motivated by the social reasoning developmental (SRD) model, which proposes that children play an active role in evaluating, interpreting, and making decisions about their social world, specifically in intergroup contexts (Killen & Rutland, 2011; Rutland et al., 2010). Developmental research guided by the SRD model has identified numerous positive outcomes related to cross-group friendships in childhood and adolescence. These outcomes include a greater likelihood of evaluating race-based exclusion as wrong (Crystal et al., 2008; Ruck et al., 2015), decreased stereotype usage (Killen et al., 2010), and less negative expectations about outgroup members’ intentions (McGlothin & Killen, 2006). Recent work examining youth’s reactions to group-based social exclusion and bias-based bullying has demonstrated that intergroup contact predicts prosocial bystander responses, particularly among majority-status youth (Palmer et al., 2017, 2022). What has not been examined is whether reported contact in the form of play with outgroup members is directly related to judgments about social competencies that reflect stereotypic expectations, such as math and science abilities along with the likelihood of pursuing high-status occupations.

Math and Science Stereotypes: Impacts on Belonging, Identity, and Motivation

STEM refers to the academic domains of science, technology, engineering, and mathematics. These domains are similarly stereotyped in many ways, but recent research highlights the domain specificity of STEM stereotypes (Cheryan et al., 2017; Leslie et al., 2015; McGuire et al., 2022). The STEM stereotypes relevant to the current study are math and science ability stereotypes, which have different associations with gender and race.

Gender Stereotypes.

Although young children tend to show a bias favoring members of their own gender group when rating math and science abilities (Kurtz-Costes et al., 2008; McGuire et al., 2020; Passolunghi et al., 2014), prior research suggests this ingroup bias may shift to the endorsement of gender stereotypes around the time children enter formal schooling. For example, children as young as six years old have been found to associate math with boys more than girls (Cvencek et al., 2011). This age is also a time when broader stereotypes about intelligence begin to reflect traditional gender stereotypes, with girls being less likely to exert ingroup bias in the selection of team members when a game is presented as being for “really, really smart” children (Bian et al., 2018). Stereotypes associating males with intelligence more than females have a direct connection to math and science stereotypes because these fields are viewed as requiring innate talent and brilliance to succeed (Leslie et al., 2015; Meyer et al., 2015).

Yet, research on the endorsement of gender stereotypes about math and science in middle childhood remains mixed, with some studies finding the persistence of ingroup bias throughout the elementary school years (Kurtz-Costes et al., 2014). Evidence of gender differences in the persistence of ingroup bias has also been found, with male, but not female, participants in middle childhood expressing similar levels of ingroup bias to those present in early childhood (McGuire et al., 2020). This finding suggests that girls may begin to endorse egalitarian math and science ability beliefs earlier in development than boys.

Gender ability stereotypes impact the inclusion of girls in math and science activities, as well as girls’ self-perceptions. In a study in which children and adolescents chose a peer to help them with a science task, participants who endorsed traditional gender stereotypes were more likely to ask a male peer for help than a female peer (McGuire et al., 2022). Much of the research on the negative impact of gender stereotypes on girls’ perceived math and science belonging and identity focuses on adolescents (Master et al., 2016; Rogers et al., 2021; Starr & Simpkins, 2021). However, the few studies focusing on elementary school-aged children demonstrate early evidence of the same negative effects. A study of Italian elementary school children found that exposure to information about the underrepresentation of women in mathematics produced performance effects consistent with the stereotype threat model (Muzzatti & Agnoli, 2007). Additionally, implicit math-gender stereotypes have been found to predict math self-concepts in German girls, but not boys, by nine years of age (Steffens et al., 2010). Math and science expectations for Black peers are also often negatively biased.

Racial Stereotypes.

Stereotypic beliefs about the math and science abilities of members of different ethnic and racial groups also develop in childhood. Although little research has explicitly examined elementary school-aged children’s ethnic and racial math and science stereotypes, related studies on the development of broader race-intelligence stereotypes provide insight into the origins of such beliefs. For example, an examination of the acquisition of race-brilliance stereotypes suggests that children associate White, but not Black, men with brilliance around six years of age (Jaxon et al., 2019). Previous findings regarding children’s broader academic stereotypes demonstrate that White children endorse stereotypes that favor White children over Black children by middle childhood, but Black children’s beliefs reflect an ingroup bias in middle childhood and then shift to the endorsement of traditional stereotypes as they transition into late childhood (Rowley et al., 2007). Understanding how young children think about their peers’ math and science abilities based on ethnicity and race is important given the prominent underrepresentation of individuals from certain ethnic and racial minority groups, such as Black individuals, across all STEM domains (Pew Research Center, 2021).

Research on the academic self-concepts of Black middle school students showed that the endorsement of traditional race stereotypes was linked to lower self-perceptions of academic competence (Okeke et al., 2009). However, this relation was only significant for youth with high racial centrality, which measures the extent to which race is a central component of one’s self-definition (Okeke et al., 2009). Addressing the gap in the research on ethnic and racial stereotypes during middle childhood is crucial because evidence of Black adolescents’ academic self-perceptions and evidence of adolescents’ ethnic and racial math and science stereotypes suggests these stereotypic beliefs emerge and may negatively impact children before they enter middle school. In addition to math and science expectations, stereotypes exist about high-status occupations.

High-Status Occupation Stereotypes

Occupational status broadly refers to the esteem and power associated with a certain career choice (Xu & Leffler, 1992). Individuals holding high-status occupations, including lawyers, doctors, and pilots, tend to have greater education levels and higher incomes compared to those not holding high-status occupations (U.S. Bureau of Labor Statistics, 2016). High-status occupations also provide additional resources and opportunities to individuals, including greater access to benefits and increased economic stability (Center for American Progress, 2022). As the SRD model contends, children are aware of social status hierarchies, and by adolescence often acquire a critical lens toward the lack of upward social mobility (Gönül et al., 2023). Children often view high-status occupations as achieved for some groups more than others due to early stereotypic expectations about who belongs in occupations of differing status (Heck et al., 2022).

Gender stereotypes shape children’s views of high-status occupations from an early age. For example, Teig and Susskind (2008) found that traditionally feminine occupations were underrepresented in 6- to 12-year-old children’s perceptions of high-status jobs. Children also demonstrate expectations of high-status occupations based on ethnicity and race. A study of Black children’s perceptions of occupational status found that participants associated lower-status occupations with being performed by ethnic and racial minority individuals, and associated novel occupations depicted by White people with being high-status (Bigler et al., 2003). These stereotypical occupation expectations have consequences for how children imagine futures for themselves. Research suggests that girls are less likely to aspire to high-status occupations than are boys (Conlon et al., 2023; Teig & Susskind, 2008), and this likelihood decreases into adolescence (Sadler et al, 2012; Saw et al., 2018). Similar patterns in the occupational aspirations of Black children (Bigler et al., 2003) and adolescents (Saw et al., 2018) have been observed as well. Thus, in addition to measuring math and science competency beliefs, we also assessed expectations about others’ high-status occupation prospects to provide another indicator of how cross-group play may be related to fewer biases and prejudicial attitudes.

The Present Study

The present study aimed to better understand the relationship between children’s reported cross-group play and their math and science competency beliefs as well as high-status occupation expectations about girls and Black peers. Based on intergroup contact theory, the social reasoning developmental model, and prior research demonstrating the effects of intergroup friendships on intergroup attitudes (Davies et al., 2011; Killen et al., 2022), we predicted the following:

Gender.

We hypothesized (H1) that for boys, but not for girls, higher levels of reported cross-gender play would be associated with more positive math and science competency beliefs about girls. We also hypothesized that fifth graders would report higher math and science competency beliefs (H2) for girls than would third graders. We also hypothesized (H3) that for boys, but not for girls, higher levels of reported cross-gender play would be associated with more positive high-status occupation expectations for girls and that fifth graders would have higher high-status occupation expectations (H4) for girls than would third graders. These hypotheses were driven by research suggesting that, with age, children less strongly endorse gender stereotypes about math and science (Kurtz-Costes et al., 2014; McGuire et al., 2020).

Ethnicity and Race.

We hypothesized (H5) that for non-Black participants, but not for Black participants, higher levels of reported cross-race play would be associated with more positive math and science competency beliefs about Black peers. We also hypothesized that fifth graders would report higher math and science competency beliefs (H6) about Black peers than would third graders. Lastly, we hypothesized (H7) that for non-Black participants, but not for Black participants, higher levels of reported cross-race play would be associated with more positive high-status occupation expectations for Black peers and that fifth graders would report higher high-status occupation expectations for Black peers than would third graders (H8). The SRD model contends that, with age, children are increasingly aware of systemic discrimination and are therefore likely to report less biased attitudes and more likely to challenge unfair stereotypic expectations (Rutland & Killen, 2015). We expected that fifth graders would be more aware than third graders that stereotypic expectations were an outcome of systemic bias towards Black peers, and thus older participants would refrain from responding in stereotypical ways due to their greater understanding of the sociopolitical context about the origins of racial bias.

Method

Participants

Participants (N = 983, 502 females) were children from ages 8 to 11 (M_age = 9.64 years, SD_age= .89 years) from the Mid-Atlantic region of the United States. Participants were 58.5% White/European American, 17.5% Multiethnic, 8.3% Asian American, 5.6% Black/African American, 4.2% Latinx, 0.6% other ethnic and racial groups, and 5.3% not reported. Due to the school and parent report forms, gender was coded as binary (girl/boy). Participants were recruited from six elementary schools with a range of low-middle to middle-high income and a mean of 8.1% of students receiving free and reduced-price meals. The schools from which participants were recruited shared similar school-level diversity such that they were racially and ethnically diverse with a White numeric majority.

Procedure

These data were collected as part of a program evaluation for a multi-site, randomized control trial to reduce prejudice and promote inclusive friendships in 2018 and 2019 (see Killen et al., 2022 for full details and outcomes of the intervention study). A survey was administered using Qualtrics. Children independently completed the survey in supervised classrooms with a researcher present to answer questions when needed. The analyses reported here do not represent a program evaluation, but instead use pretest responses to represent general survey data. We evaluated children’s pretest reported levels of gender and race cross-group play, beliefs about the math and science competencies of girls and Black peers, and expectations about high-status occupations for these peers. The Institutional Review Board at the University of Maryland, College Park approved this study. Children received a small prize (e.g., an eraser) for returning parent consent forms (independent of consent or no consent from parents). Parental consent rate was 86.3%.

Measures

Reported cross-group play.

Participants’ reported cross-group play was measured by items modified from Bierman and McCauley (1987) that asked how often they played with peers from different groups (Figure 1). Children responded on a 5-point Likert-type scale (1 = never, 5 = all of the time). All participants completed items about play with different gender and ethnic and racial groups. The measure used for each child was the one that referenced the relevant outgroup. For example, the cross-gender play measure analyzed for boys read, “Here are some girls. How often do you play with girls?” For cross-race play the item analyzed for Black participants was the one that assessed play with White peers. For all non-Black participants, the item analyzed for cross-race play was the one that assessed play with Black peers. Race of the characters was not explicitly identified such that participants read: “Here are some kids who look like this. How often do you play with kids like this?” All reported cross-group play measures were split to high/low dichotomous measures and treated as categorical variables, with 1=high cross-group play and 0=low cross-group play. Specifically, responses indicating never and a little of the time were categorized as low cross-group play and responses indicating sometimes, most of the time, and all of the time were categorized as high cross-group play.

Figure 1.

Reported cross-group play items

Math and science competency beliefs.

Participants’ math and science competency beliefs were measured by items modified from Liben and Bigler (2002) that asked participants to indicate how many members of a depicted group of peers they thought were really good at math and science (Figure 2). Participants responded on a 5-point Likert-type scale (1 = none, 5 = all).

Figure 2.

Math and science competency beliefs items

High-status occupation expectations.

Participants’ high-status occupation expectations were measured by items modified from Liben & Bigler (2002) that asked participants to indicate how many members of a depicted group of peers they thought would grow up to be lawyers, doctors, or pilots (Figure 3). Participants responded on a 5-point Likert-type scale (1 = none, 5 = all).

Figure 3.

High-status occupations expectations items

Data Analytic Plan

We conducted 2 x 2 x 3 ANOVAs to test each of our hypotheses. For all analyses regarding beliefs about girls, we included participant gender as a variable but not ethnicity and race; for all analyses regarding beliefs about Black peers, we included participant ethnicity and race as a variable but not gender. Gender was coded as girl or boy. Ethnicity and race were coded as Black, White, or non-Black ethnic and racial minority.

To test our hypotheses that for boys, but not for girls, higher levels of reported cross-gender play would be associated with more positive math and science competency beliefs (H1) and high-status occupation expectations (H3) about girls, as well as for the effects of grade (H2 and H4) we conducted 2 (Cross-gender play: high, low) x 2 (Gender: girl, boy) x 3 (Grade level: 3, 4, 5) ANOVAs. Bonferroni-adjusted pairwise comparisons were conducted to probe for significant interactions.

To test our hypotheses that for non-Black participants, but not for Black participants, higher levels of reported cross-race play would be associated with more positive math and science competency beliefs (H5) and high-status occupation expectations (H7) about Black peers, as well as the effects of grade (H6 and H8) we conducted 2 (Cross-race play: high, low) x 3 (Ethnicity/Race: Black, White, non-Black ethnic and racial minority) x 3 (Grade level: 3^rd, 4^th, 5^th) ANOVAs. Bonferroni-adjusted pairwise comparisons were conducted to probe for significant interactions.

Results

Math and Science Competency Beliefs about Girls

There were significant main effects of cross-gender play, F(1, 953) = 16.51, p < .001, ${\eta }_p^2=.017$, and gender, F(1, 953) = 15.50, p < .001, ${\eta }_p^2=.016$, on participants’ math and science competency beliefs about girls. Supporting our hypothesis (H1), we found a significant interaction between gender and cross-gender play, F(1, 953) = 9.40, p = .002, ${\eta }_p^2=.006$, such that for girls, high levels of reported cross-gender play were not associated with significantly higher math and science competency beliefs about girls. For boys, however, those who reported high levels of reported cross-gender play reported significantly higher math and science competency beliefs about girls (M = 3.58, SD = .06) compared to those who reported low levels of cross-gender play (M = 3.21, SD = .05). See Figure 4.

Figure 4. Mean math and science competency beliefs about girls by gender and reported cross-gender play level.

Note. N = 965. ***p < .001.

Supporting our hypothesis (H2) we also found a main effect for grade, F(2, 953) = 5.52, p = .004, ${\eta }_p^2=.011$, on participants’ math and science competency beliefs about girls. Follow-up pairwise comparisons showed significantly higher math and science competency beliefs about girls (p < .001) among fifth graders (M = 3.66, SD = .76) than among third graders (M = 3.41, SD = .76). There was also a significant difference (p = .001) between fourth (M = 3.46, SD = .76) and fifth grades, but no significant difference between third and fourth grades in math and science competency beliefs about girls.

High-Status Occupation Expectations for Girls

Partially supporting our hypothesis (H3), we found significant main effects of reported cross-gender play, F(1, 953) = 12.54, p < .001, ${\eta }_p^2=.013$, and gender, F(1, 953) = 8.68, p =.003, ${\eta }_p^2=.009$, on participants’ high-status occupation expectations about girls. Overall, participants with high levels of reported cross-gender play reported higher high-status occupation expectations about girls (M = 3.11, SD = .88) than did participants with low levels of reported cross-gender play (M = 2.90, SD = .93). Girls reported higher levels of high-status occupations expectations about girls (M = 3.08, SD = .88) than did boys (M = 2.92, SD = .95). We did not find a significant gender by cross-gender play interaction, F(1, 953) = 2.76, p =.097, ${\eta }_p^2=.003$.

Grade (H4) was associated with significant differences in high-status occupation expectations about girls, F(2, 953) = 3.57, p = .029, ${\eta }_p^2=.007$. Follow-up pairwise t-tests showed significantly higher high-status occupation expectations about girls (p = .003) among fifth graders (M = 3.12, SD = .83) than among third graders (M = 2.90, SD = 1.01). See Figure 5. There were no significant differences between third and fourth grades or between fourth and fifth grades in high-status occupation expectations about girls.

Figure 5. Mean high-status occupation expectations about girls by grade and reported cross-gender play level.

Note. N = 965. **p < .01; ***p < .001.

Math and Science Competency Beliefs about Black Peers

Partially supporting our hypothesis (H5), we found a main effect of reported cross-race play, F(1, 950) = 19.77, p < .001, ${\eta }_p^2=.020$, but no main effect of race, F(2, 950) = 1.89, p = .151 , ${\eta }_p^2=.004$, on participants’ math and science competency beliefs about Black peers.

However, pairwise comparisons revealed that among White participants, those with high levels of reported cross-race play reported higher math and science competency beliefs about Black peers (M = 3.72, SD = .07) than did those with low levels of reported cross-race play (M = 3.46, SD = .04). We found the same pattern among non-Black ethnic and racial minority participants, as those with high levels of reported cross-race play reported higher math and science competency beliefs about Black peers (M = 3.90, SD = .08) than did those with low levels of reported cross-race play (M = 3.34, SD = .06). See Figure 6.

Figure 6. Mean math and science competency beliefs about Black peers by grade and reported cross-race play level.

Note. N = 968. ***p < .001.

In testing for grade effects (H6), we found a significant three-way interaction of reported cross-race play by race by grade, F(4, 950) = 2.41, p = .048, ${\eta }_p^2=.010$. Follow up t-tests revealed that among Black third and fourth graders, there was no significant difference between those with high and low levels of reported cross-race play. However, there was a significant effect of cross race play among Black fifth graders, t(16) = −3.29, p = .005. Participants who reported high levels of cross-race play reported higher math and science competency beliefs about Black peers (M = 4.44, SD = .73) than did those with low levels of reported cross-race play (M = 3.33, SD = .71). There was also no significant difference (p = .39) in math and science competency beliefs about Black peers between those with high and low levels of reported cross-race play among White third graders. However, there was a significant effect of reported cross-race play for White fourth, t(205) = −3.10, p = .002, and fifth, t(171) = −1.99, p = .047, graders. Among non-Black ethnic and racial minority participants, there was a significant effect at all grade levels of reported cross-race play on math and science competency beliefs about Black peers.

High-Status Occupation Expectations for Black Peers

We found main effects of reported cross-race play, F(1, 950) = 16.71, p < .001, ${\eta }_p^2=.017$, and race, F(2, 950) = 7.21, p = .001, ${\eta }_p^2=.015$, on participants’ high-status occupation expectations for Black peers. There was not a significant interaction between cross-race play and race, such that within each ethnic and racial group those with high levels of reported cross-race play reported higher high-status occupation expectations for Black peers. However, among Black participants, this difference was nonsignificant. Among White participants, those with high levels of reported cross-race play reported significantly higher high-status occupation expectations for Black peers (M = 3.29, SD = .07) than did those with low levels of reported cross-race play (M = 2.93, SD = .04). Similarly, among non-Black ethnic and racial minority participants, those with high levels of reported cross-race play reported significantly higher high-status occupation expectations for Black peers (M = 3.56, SD = .08) than did those with low levels of reported cross-race play (M = 2.98, SD = .06).

Bonferroni-adjusted pairwise comparisons between racial groups indicated that Black participants overall reported significantly higher high-status occupation expectations for Black peers than did White participants (p < .001) and non-Black ethnic and racial minority participants (p = .021). See Figure 7. We did not find differences by grade level (H8).

Figure 7. Mean high-status occupation expectations about Black peers by ethnic and racial group and reported cross-race play level.

Note. N = 968. *p < .05; ***p < .001.

Discussion

The present study revealed novel findings regarding the associations between children’s cross-group play experiences and their beliefs about girls and Black peers. Supporting our SRD framework, the findings demonstrate that intergroup contact is directly related to children’s beliefs about math and science competencies. For boys, higher levels of reported cross-gender play were associated with significantly higher math and science competency beliefs about girls. For girls, however, there was no effect of reported cross-gender play on their math and science competency beliefs about girls. This finding aligns with prior intergroup contact theory research by demonstrating that positive contact with outgroup members can promote more positive attitudes towards that group (Tropp & Prenvost, 2008). Additionally, this finding contributes to the growing body of research demonstrating the effectiveness of gender intergroup contact on improving children’s attitudes and peer relations (Halim et al., 2021; Hanish et al., 2021; Martin et al., 2017). Specifically, our findings reveal the benefits of intergroup contact in the novel context of gender STEM stereotypes, as cross-group play was found to be associated with more positive beliefs about girls’ math and science competencies.

The present study also demonstrated the novel application of intergroup contact theory to children’s high-status occupation expectations for Black peers. Consistent with developmental intergroup contact theory literature (Aboud et al., 2003; Feddes et al., 2009) and SRD (Rutland & Killen, 2015), we found that non-Black participants with high levels of reported cross-race play reported significantly higher high-status occupation expectations for Black peers than did those with low levels of reported cross-race play. For Black participants, however, levels of reported cross-race play were not significantly associated with their high-status occupation expectations for Black peers. Children are aware of occupational status differences between ethnic and racial groups from an early age (Bigler et al., 2003). This finding suggests that intergroup contact may be an effective strategy to promote more positive expectations about Black peers’ occupational prospects, which may reduce the negative impacts these stereotypic expectations have on Black children’s self-beliefs. In addition to highlighting the positive value of high cross-race contact, our findings also highlight the negative outcomes associated with low cross-race contact. Non-Black participants with low levels of contact with Black peers were less likely to expect their Black peers to have high-status occupations compared to Black participants without contact. This finding demonstrates that a lack of cross-race contact can perpetuate stereotypic beliefs about racially minoritized groups. The negative outcomes associated with a lack of cross-race contact provide additional support for anti-racist interventions in schools (Killen et al., 2022).

Regardless of gender, participants with high levels of reported cross-gender play reported significantly higher high-status occupation expectations for girls than did those with low levels of reported cross-gender play. Although we expected to observe this effect of cross-gender play on male participants’ expectations for girls, the finding that this pattern held regardless of participant gender was unexpected. We interpret this finding with respect to the optimal conditions of intergroup contact present in children’s cross-group friendships and play experiences. In cross-gender play, children typically engage with gender outgroup peers in activities involving cooperation, shared goals, and authority support, achieving the conditions of effective intergroup contact for improving intergroup attitudes (Tropp & Prenvost, 2008). Play that occurs under these conditions may allow boys to learn that they share interests with girls, which in turn may expand their expectations for the type of occupations girls can attain in the future. At the same time, cross-gender play may provide girls with opportunities to learn about and explore stereotypically male interests and gain self-esteem and belonging in those areas, which may impact how girls envision futures for themselves. This novel finding suggests that gender intergroup contact can benefit both boys’ and girls’ expectations for girls. This type of effect has been underemphasized by intergroup contact research due to the predominant focus on outgroup attitudes (Tropp et al., 2022; Vezzali & Stathi, 2016).

For the association between children’s cross-race play and beliefs about Black peers’ math and science competencies we found that non-Black participants with high levels of reported cross-race play reported significantly higher math and science competency beliefs about Black peers than those with low levels of reported cross-race play. Interestingly, effects of cross-race play emerged for Black fifth grade participants as well. We again draw upon the optimal conditions of intergroup contact in conjunction with the SRD framework to interpret this finding. When Black children have opportunities to engage in cross-race play with White peers through activities involving cooperation, shared goals, and authority support, they may gain experiences of feeling supported and accepted by their peers, which may bolster their abilities to reject stereotypical expectations about their group. This interpretation is in line with previous developmental research suggesting that the benefits of intergroup contact for children from ethnic and racial minority groups may have more to do with belonging and peer support as opposed to reduced intergroup prejudice (Bagci et al., 2014; Kawabata & Crick, 2011).

This finding was based on a relatively small number of Black participants in the sample, and thus its generalizability should be interpreted appropriately. Nonetheless, this significant finding suggests that cross-group play with White majority status peers may afford Black children unique benefits related to their own self-competencies. Further, we reason that the effect of cross-race play emerged only among Black fifth graders in part due to children’s increasing awareness of race STEM stereotypes between middle and late childhood (Copping et al., 2013; Rowley et al., 2007). Younger children may not be fully aware of stereotypes about race and math and science abilities, and thus cross-race play does not afford them opportunities to reject stereotypes because they are unaware of them to begin with. Yet, by fifth grade, children may possess greater knowledge of these stereotypic beliefs, and thus cross-race play with White peers provides Black children with experiences they can use to actively challenge stereotypes about their group.

Together, these findings offer several theoretical advancements. First, this study adds to the growing literature documenting gender intergroup contact as a mechanism for improving children’s peer attitudes and relations (Halim et al., 2021; Hanish et al., 2021; Martin et al., 2017). Compared to race and ethnicity, gender has received relatively little attention in intergroup contact research, which has left a gap in the understanding of the effectiveness of intergroup contact in reducing gender biases. Given that persistent gender biases contribute to the underrepresentation of women in STEM and other high-status occupations (Cheryan et al., 2017; Collins et al., 2017), understanding how intergroup contact can promote more positive attitudes towards women and girls is essential in reducing current disparities. Our findings highlight childhood as an opportune time to leverage gender intergroup contact to mitigate the formation of negative gender intergroup attitudes before they become deeply entrenched in adulthood.

Second, this study advances intergroup contact theory by highlighting novel contexts in which cross-group contact may promote positive beliefs about one’s own group. Intergroup contact has mainly been cited as an effective prejudice reduction strategy (Pettigrew & Tropp, 2006; Tropp & Prenvost, 2008), yet these developmental findings offer an expanded view of the benefits of intergroup contact and align with recent suggestions to examine contact outcomes beyond just prejudice reduction (Tropp et al., 2022). Positive contact with members from other groups may allow children, particularly those from minoritized groups, to gain experiences of peer support that strengthen belonging, self-esteem, and capacities for rejecting stereotypes about their own group. Researchers and scholars invested in reducing prejudice and promoting positive intergroup relations in childhood have argued that diversity alone is not enough, but rather true inclusion and belonging must be fostered among youth for change to occur (Brown, 2019; Killen & Rutland, 2022; Rivas-Drake et al., 2019). The findings presented here on the benefits of intergroup contact for children from minoritized groups underscore the importance of such inclusion efforts. Our findings serve as further rationale for the development and implementation of intervention programs integrating principles of intergroup contact because they suggest contact can benefit members of all groups, albeit in different ways.

Lastly, the present study drew upon both intergroup contact theory and the social reasoning developmental model to examine how children’s cross-group experiences relate to their beliefs and expectations about their peers. Even though intergroup contact has always been embedded within the SRD model (Rutland et al., 2010), this integrated approach highlights the utility of examining reasoning and contact together. Despite the wealth of evidence demonstrating the effectiveness of intergroup contact in reducing both children’s and adults’ negative outgroup attitudes (Pettigrew & Tropp, 2006; Tropp & Prenvost, 2008), individuals’ reasoning about their contact experiences remains underexamined. This integrated theoretical approach offers a method for understanding how individuals make sense of and interpret their cross-group interactions, which has the potential to identify specific aspects of intergroup contact that contribute to its effectiveness.

The correlational design of the present study provides novel findings that speak to the relationship between cross-group contact and beliefs and expectations about girls’ and Black peers’ math and science abilities and occupational prospects but does not provide insight into the direction of this relationship. We discuss our findings with respect to the robust body of research on the benefits of intergroup contact on prejudice reduction, but it could also be the case that children with more positive beliefs and expectations about girls and Black peers are more likely to engage in cross-group play. The relationship between children’s cross-group interactions and their beliefs and expectations about peers is a reciprocal, interactional process in which contact and attitudes constantly inform one another. Although investigating the causal relationship between contact and these specific beliefs through longitudinal studies is a vital next step for research, the present study ultimately highlights the positive relationship between cross-group play experiences and beliefs and expectations about girls and Black peers in childhood.

Limitations and Future Directions

The present study provided novel insights into the associations between children’s cross-group play and their beliefs and expectations about girls and Black peers, but there are some limitations. One limitation is that the non-Black ethnic and racial minority group included participants identifying as any ethnicity or race other than White or Black. This decision was made in part due to small sample sizes of other ethnic and racial groups, as well as the fact that the target group in this study was Black peers. However, given that some participants in this group (i.e., Asian American children) are stereotyped more similarly to White children in STEM (Cvencek et al., 2015), and others (i.e., Latinx children) are stereotyped more similarly to Black children in STEM (Okeke et al., 2009), it is conceptually imperfect to group them together. Future studies should aim to parse out these potential differences by collecting data with equal numbers of participants from several different ethnic and racial groups. Relatedly, the research questions and analyses explored in the present study do not consider intersectionality or holding multiple identities. Understanding how children incorporate information about multiple social identities into their beliefs and expectations about peers is a current focus for developmental science research on social cognition (Lei & Rhodes, 2021). This focus is particularly important for future research on STEM beliefs given that children have been observed to hold intersectional stereotypes about brilliance in early childhood (Jaxon et al., 2019).

Another limitation is that the measures used to assess children’s cross-group play asked children about their play experiences broadly, and thus the specific contexts in which children reported engaging in cross-group play are unknown. Future research could utilize more specific measures to assess children’s cross-group play experiences in particular contexts, such as those that pertain to math and science or high-status occupations. These measures may provide a better understanding of the type of play experiences that contribute to more positive beliefs about outgroup peers, which could help guide the construction and implementation of early intervention programs. Relatedly, the measures assessing high-status occupation expectations ask children to indicate how many members of a depicted group they think will grow up to attain certain careers, but the specific reasoning underlying children’s expectations remains unclear. It’s possible that children may factor their knowledge of the structural barriers that make it difficult for certain groups to enter high-status occupations into their responses, and thus their expectations are not just pure reflections of stereotypic expectations. Future studies should follow up these assessments with open-ended reasoning questions to achieve a clearer understanding of how children are thinking about these issues.

Finally, math and science competency beliefs represent just one STEM-related belief that may be associated with children’s contact with outgroup peers. Recent research has examined the domain specificity of children’s STEM stereotypes (McGuire et al., 2022), and future work could examine the utility of intergroup contact in promoting children’s beliefs about girls’ and Black peers’ abilities in other STEM domains, such as technology or engineering. Additionally, future research could examine associations between intergroup contact and STEM interest beliefs, as these beliefs have distinct developmental patterns and outcomes as compared to ability beliefs (Master et al., 2021; Master & Meltzoff, 2020). Examining relations between children’s self-competency beliefs in math and science, their peer contexts, and their beliefs about others’ math and science competencies also presents interesting areas of future research.

Conclusion

Children’s experiences of cross-group contact uniquely shape their social development by positively influencing their intergroup attitudes and relations. The present study revealed novel associations between children’s cross-group contact experiences and their beliefs and expectations about girls and Black peers. Higher levels of cross-group play were associated with significantly higher math and science competency beliefs and high-status occupation expectations for girls and Black peers among majority group participants. Results also demonstrated positive relations between girls’ and Black children’s cross-group play experiences and their beliefs about their own groups in certain contexts. Overall, these findings highlight novel ways in which cross-group contact can benefit children from all groups and provide support for the integration of principles of intergroup contact into early intervention programs aimed at reducing prejudice and improving intergroup relations in childhood.

Data Availability Statement

This study was not pre-registered; data and code will be available upon request from the corresponding author.