P.Mean: Borderline p-values (created 2012-09-19).

News: Sign up for "The Monthly Mean," the newsletter that dares to call itself average, www.pmean.com/news.

Dear Professor Mean, I originally reported a p-value of 0.04 for a Chi-Square test, but I was told to use the Fisher's Exact Test instead. The p-value for Fisher's Exact Test is 0.06. Do I have to drop the discussion of statistical significance?

You don't have to stop talking about a result when the p-value crosses the border to 0.06. There's a term for this: borderline statistical significance. Some people hate that word, so you may have to change it, depending on who the peer reviewer is. The only thing that bothers me about the term is that people always seem to remember the adjective "borderline" when the p-value is 0.06, but they forget that it applies equally well when the p-value is 0.04.

In any case, you should consider de-emphasizing the p-values and consider confidence intervals instead for the odds ratio, relative risk, or Number Needed to Treat. These are far more important to discuss than whether the p-value is significant or borderline significant.

Creative Commons License This page was written by Steve Simon and is licensed under the Creative Commons Attribution 3.0 United States License. Need more information? I have a page with general help resources. You can also browse for pages similar to this one at Hypothesis Testing.