Using acceptance tests to predict merge conflict risk

RQ4: Is TestI a better predictor of conflict risk than TextI?



Tasks with non disjoint TextI interfaces more likely change a common file




TextI is a more precise predictor of conflict risk than TestI