NPA personality theory
[ Post a Response | NPA personality theory ]
Re: making sense of published medical literature; a statistical question
The best studies, from the point of view of statistics, are those that try to replicate a new finding. First, the researchers are less likely to be biased, in that they should be equally pleased to either confirm or refute the new finding. Second, they would be more likely to have a very specific null hypothesis to be tested, and they would be less likely to rework ("massage") the data with alternative hypotheses and alternate statistical tests.