Are nonsensical models useful?
Statistics
Despite that regression models sometimes include parameters that are theoretically invalid, they are still used to explain phenomena. Why? This will be explained by fitting different models on data with theoretical limitations.