Combining $p$-values in Meta Analysis
I came across the term meta-analysis in the previous post, and I had another question about nominal size while reading the paper of the previous post, which reminds me Keith’s notes. By coincidence, I also find the topic about meta-analysis in the same notes. Hence, this post is mainly based on Keith’s notes, and reproduce the power curves by myself.
Story about P value
“The p value was never meant to be used the way it’s used today.” –Goodman