Ex post facto hypothesizing

Ex post facto" literally means "after the fact". So ex post facto hypothesizing occurs when you notice a correlation between two variables, andthen make your hypothesis. (rather than the other way around)

For example, I may start with a hypothesis like "sugary foods cause child obesity". I do my research and find that TV watching actually shows a much higher correlation with child obesity. So then I change my original hypothesis and disseminate my findings as if that had been my hypothesis all along - to make it look as if I had a correct hypothesis.