What can be done when in meta-analysis the variance of effects size is zero?

When effect size (prevalence) is 1 or 0 variance p(1-p)/n is zero and in the pooled effect size the weight of such studies is 0.

09 May 2014 1,180 3 View