Why does anyone pay attention to "scientific studies" about code? If you ready beyond the headline of 90% of them, it's methodologically garbage. Code quality as a single score from 1-10?! Like epidemiological studies of nutrition: "coffee causes cancer!"; next week it doesn't.