#statschat #statsdon #frequentist
Say you run a simple study (Control v Training) in which you planned for an effect of d=.4, power =.80, alpha =.05 (so N=200), but you couldn't collect the total sample; final sample N= 40.

Q: how would you report the outcome to ensure the limitations of the study are clear, for:
a) a significant result
b) a non-significant result

I care about what stats/metrics/plots you would provide and why (e.g. Gardner-Altman plots, FPR, p-curve, bootstrapped analyses)