Promoting a Diverse Computing Workforce: Using National Survey Data to Understand Persistence Across Undergraduate Student Groups
The need for increased participation of individuals from underrepresented groups in computer science is well documented. This project makes use of a large, diverse data set to better understand the issues of identify and self-efficacy that contribute to persistence in underrepresented groups in computer science. The project capitalizes on an existing NSF-funded infrastructure that annually surveys a national sample of students enrolled in a diverse array of computing programs. Specifically, this project leverages the Computing Research Association's (CRA) Center for Evaluating the Research Pipeline (CERP), which collects large scale, cross sectional survey data concerning the experiences of students pursuing computing career tracks from a variety of institution types and from a wide array of demographic groups. By evaluating self-efficacy and identity across a large and broad data set from a diverse population of students and institutions, a much better picture of factors influencing student persistence is being generated.
The proposed research is interested in predicting persistence among a variety of student groups in computing. The hypotheses that will guide this study are: (1) Self-efficacy, occupational values, and social identity threat predict engagement and persistence in computing, (2) Group differences in perceived threat will predict group differences in self-efficacy and, subsequently, group disparities in engagement and persistence, and (3) Social support will attenuate the negative effects of self-efficacy and social identity threat on engagement and prediction. Two surveys will be administered over the course of the project to assess the proposed theoretical model of student persistence. The first survey will be administered to a national sample of undergraduate and graduate computing students that will measure engagement and intentions to persist in computing and their hypothesized predictors. The second survey will measure engagement and intentions to persist again (i.e., changes over time), as well as actual persistence. A series of analyses will examine each of the hypotheses implicated by the theorized model. A structural equation modeling will be used to assess overall model fit and examine relations among variables.
This research informs the development of best practices and new interventions that positively impact retention of a wide variety of underrepresented populations in computer science nationwide. This project combines a strong theoretical background and unique social science perspective with an unmatched data collection infrastructure provided by CERP. Project evaluation is conducted by an advisory board throughout all phases of the project. The objectives that will be measured are the scope of the dissemination of the data produced and the number of interventions developed by educators and administrators who use the data. Consistent contact with an advisory board results in the collection of high caliber data from a diverse population of students and the dissemination of results in a way that is a accessible and useful to a general audience with interest in STEM education.