The Importance of Role Models Who Show Enthusiasm About STEM

• In Event: Beyond Seeing Success: Exploring Optimal Features of Role Models and Mentors to Broaden STEM Participation

Abstract

Objectives
Broadening participation in computer science (CS) is important. Girls represent only 29% of AP CS test-takers, with Black and Hispanic/Latinx students similarly underrepresented (Code.org et al., 2022). Psychological factors such as negative stereotypes can deter students from introductory STEM courses (Cheryan et al., 2015). In particular, stereotypes that girls are less interested than boys in CS have negative causal effects on girls’ interest (Master et al., 2021). One way to counter interest stereotypes may be through role models who send messages about their enthusiasm and interest in CS. The purpose of this user-centered design study (Yeager et al., 2016) was to examine important features of role models to create more effective interventions to counteract interest stereotypes. We asked participants how different features of role models (including enthusiasm/interest) would affect their personal interest in a CS class. We had two research questions: (1) Which features would girls rate most highly (especially compared to boys)? (2) For which features would race/ethnicity similarity matter most?

Theoretical Framework
Social cognitive theory posits that students’ cognitions, behavior, and emotions underlie their motivation and academic choices (Dweck & Master, 2009). According to social cognitive theory, role models are most effective when they are similar to students (in terms of gender and race/ethnicity) and their success appears attainable (Lee et al., 2023). However, this theory has not examined how features such as interest might impact effectiveness (Figure 1).

Methods/Materials
Participants were 475 students in Grades 8-9 (49% girls, 45% boys, 6% missing; 57% Black, 30% Hispanic/Latinx, 3% Asian, 1% White, 9% missing/other, Mage = 14.05, SD = .53) because early adolescence is an ideal stage to influence academic interests (Maltese & Tai, 2010). Study design and analyses were preregistered. Data were collected during class time through partnership with the Character Lab Research Network using an online Qualtrics survey in 2022-2023. Participants watched a video of three Hispanic/Latinx middle school role models talking about their CS class (Table 1). Participants responded to questions about how various collective characteristics and ages of the students in the video affected their interest (Table 2).

Results
We conducted mixed model ANOVAs to examine interest based on survey item set and participant characteristics (gender and race/ethnicity) with post-hoc pairwise comparisons (Tables 2-3). Girls reported their interest was most affected by enthusiasm/enjoyment, experiences, and similarity of the role model, ps<.005. Girls were more likely than boys to report their interest was affected by discussions of belonging, p<.001. Girls also reported their interest would be more affected by a role model their own age than a professional, p=.018. In terms of race/ethnicity, Hispanic/Latinx participants were more likely than Black participants to say race/ethnicity, experiences, and “type of person” increased their interest, ps≤.032.

Significance
Overall, enthusiasm and enjoyment were among the most highly rated features across measures. These findings indicate that features of role models that relate to enthusiasm, enjoyment, and interest may be critical for countering interest stereotypes to broaden the participation of girls and minoritized students in STEM fields like CS.

Authors

• Kahyun Lee, University of Houston

• Jennifer L. Thompson, University of Houston

• Allison Master, University of Houston