issue-certain designs??that enable a full specification of your distribution of the result by making use of random outcomes to capture in-issue correlations. The selection concerning which method of opt for partly will depend on the purpose in the investigation and the desired interpretation on the estimated consequences (inhabitants-average compar