1

I am running a series of non-linear regressions (exponential and logarithmic) and came across this blog post from Minitab's statistician. I delved further and found this Cross Validated question, which was helpful. However, I am wondering if anyone knows of some literature supporting not reporting p-values along with the other results of the regression? P-values are ubiquitous in my field and not reporting them requires a reference, despite the logical explanation for it.

Thanks!

Community
  • 1
mlav
  • 11
  • 2
  • This doesn't seem to be a Minitab specific question - perhaps that tag should be removed? Or is it there because you're querying the default setting in Minitab? It seems to me that this Q touches on a more general, underlying statistical issue – Silverfish Sep 18 '16 at 22:09
  • +1: that blog post is very confusing, especially because it advocates computing confidence intervals--and those can almost always be used for hypothesis testing! – whuber Sep 20 '16 at 15:46

0 Answers0