[virtual_void]

N is not the only thing that matters here. It is important but if you’re disregarding results solely on that basis then it’s unfounded.

I noted from reading the paper that they based their sample size on a previous paper that shows the effect size for contractual muscle strength to be 0.6 and turned that in to a sample size estimate (in that paper). The same technique was used here and they nearly doubled the sample size from the previous study. It seems from a first reading that they have the statistical power to make the conclusions they do.

The study that shows 0.6 for effect size was a meta analysis that i do not have access to via my institution, but I’m curious as to why you think this paper is no better than anecdote. Perhaps some other part of their methods?

Their results show at the very least that it might be worth a bigger study (more N) but there’s nothing outrageously wrong with their methodology, is there?

[rd11235]

My comment isn’t about the paper. It’s only about the surprising response it’s getting from HN. This small level of evidence can be found in some paper for almost any fitness claim one can imagine. (And that is not at all a knock against the paper - as you said, it can e.g. warrant larger studies.)