Pair Programming as a Pedagogical Approach for Promoting Success and Equity in Computer Science Coursework

University Of Iowa, Iowa City IA

Investigators

Abstract

Most economic forecasts foresee a shortage of qualified computer scientists and engineers by the year 2020, leading to growing interest in computer science courses at many universities. Yet the usual first course for computer science majors typically consists of learning to program, and necessarily involves at least several challenging programming assignments, which can lead to poor student engagement and eventual disenchantment. Growing enrollments only exacerbate the challenges inherent in these courses, which often have low retention rates and are particularly unpopular among female students. One promising technique to improve engagement and retention is to allow students to work on programming assignments in pairs, i.e., pair programming. Pair programming is not a new concept itself, but quantitative measures of the resulting student learning are few. The significance of this award is a cluster-randomized control trial on the effectiveness of pair programming in the introductory computer science curriculum. This project will consider differential effects for female and male students across different introductory computing courses using 2,000 undergraduates. Outcomes of interest include level of programming skill, attitudes toward computing, course attrition rates, course grades, selection of and performance in subsequent computational coursework, and changes in undergraduate major. The project investigation will study the effects of pair programming on student short-term and long-term programming skills, computing attitudes, academic achievement, and persistence in computer science, particularly paying attention to female students.