Differential Power of Distinct Kinds of STEM Stereotypes: Interest Versus Ability Stereotypes

• In Event: 1-080 - Do People Like Me Belong in STEM? Social Cognitive Influences on Children’s STEM Motivation

Integrative Statement

Societal stereotypes about gender and STEM may contribute to girls’ underrepresentation in STEM (Cheryan et al., 2017). Previous research has primarily examined stereotypes about ability (“which gender has more ability in STEM?”) rather than interest (“which gender likes STEM more?”). Yet the sense that one’s group has less interest in STEM may be a more powerful deterrent from feeling a sense of belonging and interest in STEM than the sense that one’s group has less STEM ability (Wigfield & Eccles, 2002). Here we focused on the subtle but critical difference between ability versus interest stereotypes. We examined stereotypes about computer science because it has one of the lowest representations of women among STEM fields, with only 18% of US bachelor’s degrees earned by women (NSF, 2017), and because gender gaps in technology interest emerge early (Master et al., 2017). We predicted that endorsement of interest stereotypes about computer science would be more strongly negatively related to girls’ motivation in computer science than endorsement of ability stereotypes.

Participants were 879 3rd-7th grade students in two Rhode Island school districts, which presented a unique opportunity due to a statewide initiative to put computer science courses in all K-12 schools. Participants took an online survey, with all items on a Likert scale from 1-6. We used computer science terms familiar to participants (e.g., “coding”). The survey measured interest stereotypes (calculated as the difference score between ratings for “how much most boys like coding” and “how much most girls like coding”), ability stereotypes (difference score between “how good most boys are at coding” and “how good most girls are at coding”), and three measures of motivation: sense of belonging (3 items, α=.77; e.g., “I feel like I belong in my coding class”), self-efficacy (2 items, α=.81; e.g., “I am good at coding”), and interest (2 items, α=.83; e.g., “I choose to do coding activities”).

As predicted, interest stereotypes were more strongly related to girls’ interest in computer science than ability stereotypes. The more that girls believed that boys were more interested than girls in computer science, the lower their motivation in computer science, see Table 1. While ability stereotypes were negatively correlated with girls’ self-efficacy, they were not correlated with belonging or interest. Supporting the theory that stereotypes about girls’ lower interest predict lower motivation only for girls, the correlations between stereotypes and motivation were negative for girls and positive for boys. Further, in linear regressions including both interest and ability stereotypes, only interest stereotypes predicted girls’ interest, b=-0.28, SE=0.06, p<.001 (ability stereotypes: b=0.05, SE=0.06, p=.42), and boys’ interest, b=0.19, SE=0.06, p<.001 (ability stereotypes, b=0.07, SE=0.06, p=.20).

Girls who believe boys are more interested than girls in computer science may decide that computer science is not for them, and pursue other fields instead. Interventions that counteract the negative effects of interest stereotypes may be a promising target for future programs aimed at recruiting more girls into STEM majors and careers. Theoretical implications for the development of stereotypes, self-concepts, and motivation will be discussed.

Authors

• Allison Master, University of Washington
  Presenting Author

• Sapna Cheryan, University of Washington
  Non-Presenting Author

• Andrew Meltzoff, University of Washington
  Non-Presenting Author