Hacker News new | ask | show | jobs
by whitneyrzoller 2553 days ago
"I then tallied the main result, the sample size and citation patterns for each of the 10 studies, as reported here"

(Honest) stats question: Given that n=10 is itself a small sample size, how does the computation of statistical power in meta studies differ from that of the underlying studies?

Is there an accepted methodology? I couldn't remember doing this ages ago when I took stats.