# does p-value distribution indicate the suitability of the statistical test

This post suggests that p-value follows uniform distribution in case of point null hypothesis and continuous data.

In my project, I have millions of p-values from actual data. I don't want to get into the details of my data and statistic test. I just want to know the following:

In general, for point null hypothesis and continuous data, if the p-values do not follow uniform distribution, does it imply that 1) the null hypothesis is not correct, and/or 2)that applied statistic test is not appropriate? could there be other reasons?

Thanks,

• I don't understand your question. Under these conditions the p-value under the null hypothesis is uniformly distributed by definition, as explained in the comments on & answers to the question you linked to. – Scortchi - Reinstate Monica Apr 14 '16 at 15:22
• No, it could mean that the null hypothesis is wrong... but to compare the p-value distribution with the uniform distribution, you need to perform the same study dozends of times, which is normally pathetic. – Michael M Apr 15 '16 at 16:13
• As @MichaelM says, the distribution under an alternative hypothesis won't necessarily be uniform - indeed shouldn't be if it's an alternative you want your test to have power against. I'm also puzzled as to how you're getting the distribution of actual p-values. Perhaps add an illustration of what you mean to your question. – Scortchi - Reinstate Monica Apr 15 '16 at 16:24
• I want to raise pretty much the same points as @Scortchi. How do you know "if the p-value of a statistic test does not follow uniform distribution"? You should make it clear whether you are talking about running a simulation to generate simulated p-values, or if you intend to use p-values from actual studies. – Silverfish Apr 15 '16 at 18:06
• Are you obtaining different p values for the same hypothesis or just the same type of hypothesis (but e.g. from different variables)? – Michael M Apr 18 '16 at 14:28