LogFAQs > #971630724

Studies based on self-report were excluded from the meta-analysis:

Of the remaining 172 papers, 97 were further excluded as they either did not report penis measurements (n=63), reported measurements after major pelvic surgeries (n=12), or reported self-measurements (n=22).

Makes sense, self-report would be wildly unreliable on this subject.

That said I can imagine there's still a selection bias. The conclusion might as well be "small dick-havers have become more shy and less likely to have their dicklet measured by a scientist compared to 30yrs ago".