The Proof Is in the Pudding: Designing for Gender Equity in STEM Interest and Persistence

• In Event: Finding the Hook: Engaging Teachers and Students in Computer Science

Abstract

There is a growing consensus that our formal science, technology, engineering, and mathematics (STEM) education practices are failing to engage the interests of today’s youth in meaningful ways (Lemelson-MIT, 2012; Maltese & Tai, 2011; Symonds, Schwartz, & Ferguson, 2011; Augustine, et al., 2010). If we are to connect a broader and more diverse segment of our youth with the intellectual, creative, and cultural practices needed by our current and future STEM workforce (Casner-Lotto & Barrington, 2006), we must provide more engaging forms of initial participation in STEM activities – in other words, better on-ramps into STEM. We created a new program model called FUSE (www.fusetudio.net) designed for just this purpose. Our design-based research questions include: How do we trigger young people’s initial interests in STEM and then sustain that interest over time? Can we design a model that equally engages both boys’ and girls’ interests and persistence in STEM activities?

FUSE is structured around a set of interest-driven challenge sequences organized into levels. Each level of a challenge increases in difficulty, requiring the skills and knowledge developed at prior levels – similar to video game design principles. FUSE participants are free to select any challenge sequence on our website to work on and can decide how far through the sequence of challenge levels they wish to continue. Thus, the FUSE model provides an ideal venue for exploring the factors that impact early youth engagement with – and persistence in – STEM activities. In developing the set of challenges in FUSE, our goals are twofold: 1) Curating a set of challenge sequences that are equally appealing to both genders, and 2) Designing challenge sequences that would have similar persistence rates across gender.

This presentation explores the question of how well our design of FUSE achieved these two goals. We collected logged user activity data from 3876 registered users of our website (2092 boys and 1784 girls in grades 5-12) from July 2012 through June 2015. A chi-square test of independence (𝛂=0.05) was performed to examine the relationship between gender and challenge interest. For the present analysis we define “interest” when a user clicks through to begin the first level of a challenge sequence.

We found that out of 22 challenge sequences with sufficient data to analyze, eight showed users’ interest to be independent of gender (i.e., an insignificant chi-square value). Seven sequences showed a male preference and seven showed a female preference (i.e., a significant chi-square value). Thus, our initial analysis revealed that we achieved our design goal of equity in meeting the expressed interests of both genders, though the way we accomplished this was somewhat unexpected. About one third of the challenge sequences were universally appealing: girls and boys each had an additional one third of the challenge options with demonstrated gender-specific appeal. The next phase of analysis will examine the extent to which our design supported gender parity in persistence through the challenge sequences and these results will be shared during the session.

Authors

• Kemi Jona, Northwestern University

• Reed Stevens, Northwestern University

• Lucy Yang, Northwestern University