Peer instruction contributes to self-efficacy in CS1

Abstract

Recent work in computing suggests that Peer Instruction (PI) is a valuable interactive learning pedagogy: it lowers fail rates, increases retention, and is enjoyed by students and instructors alike. While these findings are promising, they are somewhat incidental if our goal is to understand whether PI is "better" than lecture in terms of student outcomes. Only one recent study in computing has made such a comparison, finding that PI students outperform traditionallytaught students on a CS0 final exam. That work was conducted in a CS0, where the same instructor taught both courses, and where the only outcome measure was final exam grade. Here, I offer a study that complements their work in two ways. First, I argue for and measure self-efficacy as a valued outcome, in addition to that of final exam grade. Second, I offer an inter-instructor CS1 study, whose biases differ from those of intra-instructor studies. I find evidence that PI significantly increases self-efficacy and suggestively increases exam scores compared to a traditional lecture-based CS1 class. I note validity concerns of such an in-situ study and offer a synthesis of this work with the extant PI literature.

Predicting student success using fine grain clicker data

“…However, even in classes employing these instructional practices, there appears to be a vast divide among the students who succeed and those who fail. For example, researchers have shown independent of pedagogical style that students with prior experience in programming fare much better than those without prior experience [17].…”

A multi-institutional study of peer instruction in introductory computing

“…N/A denotes data not available. CS, a number of studies have reported on the success of PI, including improved student satisfaction [5,8,10], student learning [7,15], final exam grades [11,13], failure rates [6], and retention of majors [9].…”

“…Students value PI [5,8,10] and learn more in PI classes compared to traditional lecture classes [11,13]. PI is also associated with low failure rates [6] and increased retention of majors [9].…”

Comparing outcomes in inverted and traditional CS1

“…Considerable research shows that students with no prior programming experience ("beginners") perform more poorly than students with experience [11,4]. In the inverted offering, students participated in supported classroom practise three times a week and prepared for these activities by watching videos.…”