Researching Pre-College Factors that Lead to Persistence in Computer Science

Harvard University, Cambridge MA

Investigators

Abstract

This research project is studying the wide range of efforts underway nationwide to offer computer science (CS) and computational thinking (CT) to K-12 students, by measuring their impact on students' career interest in computer-related fields and on their attitudes toward computing. While there is widespread support for offering CS and CT experiences to pre-college students, there are few national statistics on the prevalence of in-school/out-of-school offerings, as well as personal hobbies and explorations. However great the enthusiasm about pre-college CS and CT initiatives, little definitive evidence exists about their long-term effects on students. Whereas many programs have been evaluated individually, the field has yet to examine the relative impact of pre-college students' CS and CT experiences on their interest in CS and STEM careers, their attitudes toward CS, and their CS identity. This study measures the impact of decisions about the use of computers and CT activities, made by CS and STEM teachers, as well as teachers of other subjects, along with those made by the creators of online resources, out-of-school time educators, and other involved professionals. This study has the capability to reveal the most promising educational practices and interventions (many developed with NSF support), including in-school computing and CT instruction, after-school programs, competitions, and clubs in a nationally representative sample of first-year college students. Because the under-representation of females and certain racial and ethnic minorities in the STEM and CS workforce has been a long-standing concern, this project will have a strong focus on girls and underrepresented minorities. It will determine both the degree to which these groups have access to the various CS and CT-related opportunities and the extent to which these opportunities translate into positive long-term outcomes for them. Using epidemiological methods that control for prior interest and background variables, this retrospective cohort study will collect a nationally representative, stratified random sample of 8,000 college students (enrolled in a mandatory freshman course, so that students with all levels of interest and experience in CS and CT are included). It will model the degree to which any prior CS- and CT-related experiences predict students' CS attitudes, identity, and career interests, using linear and logistic regression. Epidemiological techniques offer a cost effective and well-understood methodology to simultaneously test multiple hypotheses, while controlling for a host of demographic and background factors that differ for individual subjects. This project is funded by the CS for All: Research and RPPs program. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.