Abstract:
|
Too often we, as statisticians and data scientists, focus on the part of the problem we are most comfortable with -- the analysis. But, as Alan Kimball once said, there is a third type of error in hypothesis testing -- finding the right answer to the wrong problem. Usually this can be avoiding by spending more time listening to the collaborator and coming to an agreement on what the problem really is. Through a series of case studies where I committed one type of error or another, I will attempt to highlight some of the most common ways to go wrong when collaborating with others and in doing so, hope to show some ways of avoiding these errors as well.
|