The Emergence of Ability Stereotypes and Gender Differences Across STEM Fields

• In Event: "Yes We Can": Improving Equity in Education Through STEM Motivation and Engagement

Abstract

Objectives
The gender gap in motivation in science, technology, engineering, and mathematics (STEM) remains a large educational problem. One source of this gap is cultural stereotypes associating boys with STEM (Cheryan et al., 2017). Previous research has found links between gender stereotypes, ability self-concepts, sense of belonging, and motivation to pursue STEM (Authors, 2016, 2017; Eccles & Wang, 2016; Passolunghi et al., 2014; Plante et al., 2013).
Our main objective was to examine the emergence of these links across multiple STEM fields (math, science, and computer science) in 1st-12th grades. We examined when gender gaps emerged in self-concepts and sense of belonging, and when students began to endorse gender stereotypes about ability. We also examined when endorsement of stereotypes began to correlate with girls’ lower self-concepts and belonging.

Theoretical Framework
Our model is rooted in expectancy-value theory (Eccles, 2011; see also Authors, 2020), which shows how individuals’ perceptions of stereotypes and gender influence their ability self-concepts, which in turn influence their academic choices and engagement.

Methods
Participants were 1,459 students (728 girls, 731 boys; 37% White, 23% Latinx, 15% Multiple, 10% Asian American, 8% Black; and 7% Other) across 1st-12th grades in a suburban school district. Students completed an online survey. All measures were repeated for three fields: math, science, and computer coding. Stereotype score was calculated by taking the difference between ratings about boys (“How good do you think most boys are at [X]?”) minus ratings about girls (“How good do you think most girls are at [X]?” on Likert scales from 1-6), so positive scores indicated traditional stereotypes that boys were better, 0 represented neutral, and negative scores represented counter-stereotypes that girls were better. Self-concepts were measured with two items per field (e.g., “I am good at [X] activities”), αs > .84. Belonging was measured with three items per field (e.g., “I feel like I belong in my [X] class”), αs > .67.

Results
Grade levels were paired to improve statistical power and ease interpretation. For full results, see Figure 1A and Table 1 for ability self-concepts, Figure 1B and Table 2 for sense of belonging, Table 3 for stereotype endorsement, and Table 4 for correlations between stereotypes, self-concepts, and sense of belonging. Taken together, students reported strongest stereotypes favoring boys (at all grade levels) and earliest gender gaps (5th-8th grade) for computer science, compared to math and science. These findings show that gender gaps were evident and stereotypes were linked to self-concepts and belonging in computer science by middle school.

Scholarly Significance
As part of the 2021 AERA theme, education researchers are called to accept responsibility for improving equity in education. This includes removing psychological barriers that create gender gaps and prevent girls from pursuing STEM. These findings highlight important variation across STEM fields, with the greatest need for equity in computer science rather than math or science. The results point to middle school as a critical point for interventions to counteract negative effects of stereotypes and reduce gender gaps in computer science.

Authors

• Allison Master, University of Houston

• Andrew N Meltzoff, University of Washington

• Sapna Cheryan, University of Washington - Seattle