Figure 3

Power values for each method under 10 different conditions: for situation i, there are i out of K = 100 studies each has the same effect size 2/i and the remaining 100-i studies have zero effect. (a) Sample size n = 20, and standard deviation σ = 1 for each study; (b) Same as (a), but the sample size n is a random sample from Poi(20); (c) Same as (a), but the standard deviation σ is a random sample from gamma(10, 0.1); (d) Same as (b), but the standard deviation σ is a random sample from gamma(10, 0.1).