A third consideration in planning IDA within primary data collection is sampling. To the extent that individual studies are completely confounded with sampling designs (e.g., all males in one study and all females in another), researchers will be unable to use IDA to test the unique influence of confounded factors in hypothesis testing apart from the influence of study membership. If a priori hypotheses indicate that particular group differences are of interest, the inclusion of important sub-groups within each of the contributing studies can provide leverage for distinguishing the effects of theoretically meaningful factors associated with sampling design (e.g., gender, age, ethnicity) and study membership.
