topic-distinct versions??that let a complete specification on the distribution of the outcome through the use of random results to capture within-topic correlations. The selection regarding which approach to decide on partly is determined by the aim on the study and the specified interpretation on the believed consequences (population-regular vs .