Lightweight Feedback Loops in Undergraduate Software Projects: A Controlled Study Across Four Cohorts

Sabbir Hosen Mamun^†, Helena R. Novak, Joris T. Vandenberghe

IEEE Transactions on Education, Vol. 65, No. 4, pp. 88–102 · IEEEQ1, IF 3.6, Scopus

Journal202234 citations

Abstract

Frequent, low-overhead feedback during undergraduate software projects improves both product quality and student confidence. This paper examines three lightweight feedback practices — weekly check-ins, automated diff reviews, and peer walkthroughs — applied across four cohorts totalling 312 students. We find statistically significant improvements in on-time delivery (p < 0.01) and self-reported understanding of version control workflows (Cohen's d = 0.61). Peer walkthroughs produced the largest effect on code quality metrics, while automated diff reviews had the strongest correlation with final-examination performance.

Cite this work

@article{Mamun2022Lightweight,
  title = {Lightweight Feedback Loops in Undergraduate Software Projects: A Controlled Study Across Four Cohorts},
  author = {Sabbir Hosen Mamun and Helena R. Novak and Joris T. Vandenberghe},
  year = {2022},
  journal = {IEEE Transactions on Education},
  volume = {65},
  number = {4},
  pages = {88–102},
  publisher = {IEEE},
  doi = {10.1109/TE.2022.3168402},
  abstract = {Frequent, low-overhead feedback during undergraduate software projects improves both product quality and student confidence. This paper examines three lightweight feedback practices — weekly check-ins, automated diff reviews, and peer walkthroughs — applied across four cohorts totalling 312 students. We find statistically significant improvements in on-time delivery (p < 0.01) and self-reported understanding of version control workflows (Cohen's d = 0.61). Peer walkthroughs produced the largest effect on code quality metrics, while automated diff reviews had the strongest correlation with final-examination performance.},
}