1
$\begingroup$

As an early-career applied researcher and consumer of statistics, I'm aware of the Neyman-Pearson and Fisher lemmas and how they have become intertwined and misused, particularly around P-values. While I tend to use each lemma in separate occasions, I have several colleagues that uses the hybrid mixture (unknowingly), without acknowledging the limitations and caveats.

My field is mainly bioinformatics and basic experimental biomedical research, which all suffer from the small-sample issues.

While I do like likelihood and Bayesian alternatives, most of my colleagues resist them, under the argument that they add another layer of complexity and lead to similar results when using uniform or flat priors.

In this regard, is there any solid guideline for NHST (assuming this is how people call the NP-Fisher hybrid)? I'm aware of the discussions here and some papers, mainly from psychology, but wanted to make sure I'm not missing any relevant standards.

Above that, many journal reviewers and editors are unaware of such discussions and make the peer-reviewing process harder for researchers trying to comply with statistical rigor... to be honest, this all have been demotivating.

Thanks

Improve this question
$\endgroup$
1
  • 1
    $\begingroup$ Most researchers do make a very poor use of statistics, and indeed most data analysis you find outside of top material is complete rubbish. $\endgroup$
    – David
    Commented May 19, 2021 at 13:01

1 Answer 1

Reset to default
0
$\begingroup$

is there any solid guideline for NHST

Have you come across the special issue of the American Statistician with the ASA Statement on p-Values?

Improve this answer
$\endgroup$
1
  • $\begingroup$ Thanks! Yes, I did. In fact I usually recommend it to those confused about P-values and how/if they report it. I should have clarified that I'm preocupied about the pre-data stage of NHST, i.e., whether to stick to sample size planning, but not adopting the decision aspect of NP and using P-values continuously without cutoffs etc, more like an informal strenght of evidence. That's where I get confused, because I've seen different mixtures in how NHST is used in my field. $\endgroup$
    – BioLeal
    Commented May 19, 2021 at 12:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.