Very often both experimentalists and theorists engage in testing and comparing models and theories by confrontation with data on atomic nuclei. This talk will use very simple examples to show some of the dangers, pitfalls, and incorrect conclusions that can arise by the use of standard methods, such a 2 analyses. The discussion will address the (perhaps counterintuitive) dangers of highly precise data, the sensitivity of specific observables to different physics, the tendency to multiply-test the same physics while bypassing other physics, the critical idea of theoretical uncertainties, and the role of parameters in assessing different models. The guidelines that result may give insights into the kinds of data that are needed and how best to assess different models.
