issue-precise designs??that make it possible for a complete specification in the distribution of the outcome by using random outcomes to seize in just-issue correlations. The selection regarding which method of pick out partly will depend on the aim from the analysis and the desired interpretation of your estimated effects (population-typical vs .