Gender Gaps in Technology and Engineering Literacy, Self-Efficacy, and Course-Taking Over Time

• In Event: Contextualizing Student Academic Success: Using NAEP Data to Understand Student Achievement

Abstract

Girls tend to get better grades than boys across academic subjects, including science (e.g., Voyer & Voyer, 2014). However, girls also tend to underestimate their abilities in science compared to boys, creating a confidence gap (Ehrlinger & Dunning, 2003) that contributes to the underrepresentation of women in STEM fields (National Science Board, 2016). This tendency to underestimate may be reflected in gender differences in self-efficacy, or perceptions of one’s ability to succeed at or accomplish certain tasks. Self-efficacy is a positive predictor of performance outcomes across academic domains and levels (Pajares & Valiante, 2001; Usher & Pajares, 2008), but few studies have examined gender differences in self-efficacy in STEM-related domains, and how self-efficacy relates to gender performance gaps in these domains. While research on the underrepresentation of women in STEM fields notes the importance of engaging girls in STEM-related experiences (e.g., through coursework) to increase their self-identification with STEM professions and narrow the confidence gap (Microsoft, 2018), the relationship between involvement in STEM-related coursework and self-efficacy has received little empirical attention.

The current study addresses how self-efficacy and coursetaking relate to student achievement among girls and boys in the technology and engineering literacy (TEL) domain. These relationships are examined in large, nationally-representative student samples from two assessment administrations, allowing for exploration of changes to these variables over time.

Data were collected as part of the 2014 and 2018 National Assessment of Educational Progress (NAEP) Technology and Engineering Literacy (TEL) grade 8 assessments (n = 21,500 and n = 15,400, approximately). Survey questionnaires asked students about their TEL-related self-efficacy and coursetaking in engineering. Multiple regression was used to examine the utility of TEL self-efficacy and coursetaking in predicting student achievement. Gender was used as a control variable, self-efficacy and coursetaking were entered as independent variables, and TEL achievement scores were used as the criterion variable. Gender (β = 0.07, p < .01), high self-efficacy (β = 0.70, p < .01), and involvement in engineering coursework (β = 0.07, p < .01) significantly predicted higher TEL student achievement in 2018, replicating results of regression analyses for the same variables administered in 2014. Preliminary analyses indicate that more students who took engineering coursework reported high levels of TEL self-efficacy and scored higher than those who did not (p < .01).

The confidence gap appears to have narrowed: more boys than girls reported having high TEL self-efficacy in 2014 (p < .01), but no gender differences were observed across all levels of self-efficacy in 2018 (ps > .05). Examination of average achievement scores associated with different levels of self-efficacy revealed a change in girls’ interpretation of their TEL-related abilities. In 2014, girls who reported low and moderate TEL self-efficacy scored higher on average than boys (p < .01). However, in 2018, girls who reported moderate or high self-efficacy scored higher on average than boys. Girls’ involvement in engineering courses increased significantly between 2014 (15%) and 2018 (21%), with average scores of girls taking engineering courses also increasing significantly between these years (ps < .05).

Authors

• Jan Alegre, ETS

• Ryan P. Whorton, Educational Testing Service