99% impossible: A valid, or falsifiable, internal meta-analysis.

Several researchers have relied on, or advocated for, internal meta-analysis, which involves statistically aggregating multiple studies in a paper to assess their overall evidential value. Advocates of internal meta-analysis argue that it provides an efficient approach to increasing statistical power and solving the file-drawer problem. Here we show that the validity of internal meta-analysis rests on the assumption that no studies or analyses...