Analyzing Pivoting Among Weekly Many Small Programs in a CS1 Course

In Fall 2018, our university fully switched from using a weekly one large program (OLP) approach to using a many small programs (MSP) approach in our CS1 course, utilizing a program auto-grader with immediate points feedback and partial credit possible. The switch led to positive results such as an increase in student grade performance and a reduction of student stress. We also saw students making good use of MSPs in their learning, such as spending sufficient time programming each week, and starting earlier on programming assignments. A unique benefit of MSPs is the ability for students to pivot, meaning to switch among programs if they get stuck. This paper investigates such pivoting, and seeks to answer common questions related to pivoting. We analyze how many students pivot and the number of pivots done each week. Given a full-credit threshold (50 of 70 points on 7 programs worth 10 points each with partial credit possible), we examine how students complete the subset of required points. We compare pivot data between a class with a full-credit threshold and a class without. We examine whether students who pivot eventually return to the program from which they pivoted, or if they leave the program unsolved. Finally, we analyze student workflow to observe various pivot patterns. By analyzing student pivot behavior, we hope the community can better understand the pros and cons of pivoting, to help decide whether to adopt an MSP approach and possibly a full-credit threshold.