The Impact of Code Review Coverage and Code Review Participation on Software Quality
https://rebels.cs.uwaterloo.ca/papers/msr2014_mcintosh.pdf
Method
"we extract code review data from the Gerrit review databases of the studied systems, and link the review data to the integrated patches recorded in the corresponding VCSs"
Population
"Qt 5.0.0 (5,560,317 LOC, 1,339 components, 254 with post-release defects, 10,163 commits, 435 authors/358 reviewers), Qt 5.1.0, VTK 5.10.0 and ITK 4.3.0"
What it does not show
The paper’s own counter-signal is the important part: coverage “only provides significant explanatory power to two of the four studied releases, suggesting that review coverage alone does not guarantee a low incidence rate of post-release defects”, and “several defect-prone components have high coverage rates”. Three open-source projects with unusually traceable review data.
Shane McIntosh, Yasutaka Kamei, Bram Adams, Ahmed E. Hassan
“Low code review coverage and participation are estimated to produce components with up to two and five additional post-release defects respectively.” Changes approved with zero comments were significantly associated with defects across all three releases studied, as were self-approved and hastily reviewed changes.
Tier II: Observational study on real defect data from three large open-source systems, with regression models controlling for size, complexity and churn. Correlational, no manipulation.