Software Project Management

APPENDIXAThe Rationale for “Mixed-Methods” Measures

Throughout this book, you have encountered instances where conclusions were drawn from a very limited data set. Sometimes, those conclusions also drew from qualitative information, such as other factors that “might” be affecting the data or knowing a bit about the team’s history. The guidelines for interpreting measures presented in the book were certainly not statistically valid inferences. So it is quite reasonable to ask whether these (or any such guidelines) have any validity as analytic tools.

The short answer is: It depends on what you mean by validity. Can our common-sense judgments and interpretations be “valid,” or is validity only possible in logically or statistically sound claims? ...