I felt really good about the data analysis work I performed this morning, until I discovered that the data looked good, but was wrong, wrong, wrong. A large part of my job involves testing and then rejecting data. I wonder if it is like that for everybody in my line of work.