Hi Szabolcs,

Correct. Multiple testing problem will appear whenever there are multiple p-values in a family of tests. Linear mixed effects models aren't immune to them.

That said, there seems to be some misunderstanding between authors and reviewers, and a suggestion would be for the parties to make it all clear how the multiple testing has arisen in this case and how it was or could be mitigated.

All the best.

Anderson

On Mon, 12 Nov 2018 at 17:58, Szabolcs David <[log in to unmask]> wrote:

Dear Anderson and co.,

I would like to ask for some advice: In a paper, we pointed out that someone did not correct for multiple comparison (or didn't communicate it clearly..), but the reviewer commented on:
"The authors point out that ** et al did not control for multiple comparisons. Unclear what authors mean by this? Multiple comparisons by multiple patients? In fact they did if the authors read the statistical analysis section of that paper: they did this with a linear mixed effects model which is ideal for this purpose, they controlled for correlated observations within subjects by testing a subject-specific random slope. This paper was not looking at multiple regions. Perhaps changing wording here."

Linear mixed effects models are not immune to such problem, aren't they?

Kind regards,
Szabolcs

---

To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1

---

To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1