Help stats gurus! Am reviewing a paper: N=24 w 7 2x2x2x2 RM anovas reporting 4-way interactions (& no correction for mult comparisons). ISO authoritative source to cite in support of statement that N = 24 is way underpowered for 3 and 4-way within-subject interactions. thx!
Is it possible there is a potential issue with RM designs with 4 factors being falsely over-powered due to so many repeated measures?
-
-
Impossible to say without context - overpowering is relative to the effect you want to detect and its risk/cost. Generally less of an issue in designed experiments (which aren't usually addressing applied impact)
-
It could still be massively underpowered - for example if they sample stimuli but treat stimuli as a fixed factor then then power could be extremely low or if there are very few trials per person. I think the main point is that n = 24 isn't a clear indicator on its own
- 1 more reply
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.