Developer experience · Retrospective quantitative case study using repository/code-review histories and statistical models relating review-practice measures to post-release defects.
An empirical study of the impact of modern code review practices on software quality
Empirical Software Engineering · 2016
Summary
Review coverage, participation, and reviewer expertise each share a significant link with software quality; poorly reviewed code is associated with more post-release defects.
For your marketing
Possible founder application
Prioritize review coverage, meaningful reviewer participation, and domain-relevant reviewer assignment in developer-tool metrics and workflow design; treat these as risk signals rather than simplistic productivity scores.
Reported result
What the study found
Review coverage, participation, and reviewer expertise each share a significant link with software quality; poorly reviewed code is associated with more post-release defects.
Research question
What the authors investigated
How are code-review coverage, reviewer participation, and reviewer expertise related to post-release software quality in modern lightweight review?
Design
Retrospective quantitative case study using repository/code-review histories and statistical models relating review-practice measures to post-release defects.
Sample and setting
Three large open-source projects using modern review tools: Qt, VTK, and ITK; outcomes are post-release defects at component/change level. The accessible abstract does not state a single participant count because the unit of analysis is repository history rather than surveyed people.
Scope and transfer
Limitations to keep in view
Observational archival data support association rather than a clean causal estimate; the study is a case study of three open-source projects, so organizational and proprietary settings may differ; post-release defects are a proxy for quality and review measures may be confounded by change complexity and project process.
Research sources
Bibliographic fields and abstract: https://link.springer.com/article/10.1007/s10664-015-9381-9; accessible paper: https://rebels.cs.uwaterloo.ca/papers/emse2016_mcintosh.pdf Firecrawl research job 01a14371-854f-71aa-ba3d-455ec397717d returned this citation and metadata.