Simple calculations seem to show that larger studies should have higher statistical power, but empirical meta-analyses of published work in criminology have found zero or weak correlations between sample size and estimated ...