Don't Touch My Code! Examining the Effects of Ownership on Software Quality
https://www.microsoft.com/en-us/research/wp-content/uploads/2016/02/bird2011dtm.pdf
Method
"We compare the amount of variance in failures explained by a model that includes the ownership metrics to a model that does not include them."
Population
"Windows Vista and 7 were developed by 2,000+ software developers and is composed of thousands of individual executable files (.exe), shared libraries (.dll), and drivers (.sys)"
What it does not show
The authors state their own limit: results “are much more likely to hold in other industrial settings where similar policies are in place”, and the effect where ownership is deliberately not stressed, as in much open source, is untested. Correlational, not a manipulation of ownership.
Christian Bird, Nachiappan Nagappan, Brendan Murphy, Harald Gall, Premkumar Devanbu
Adding ownership metrics — particularly the count of low-ownership ‘minor’ contributors — substantially increased explained variance in failures, from 26% to 46% pre-release in one system and from 24% to 70% in the other. Minor contributor count correlated with defects more strongly than any other metric the organisation collected.
Tier II: Observational study on real defect and failure data from two very large systems, using regression to control for size, complexity and churn. Correlational despite the regression framing.