The degree to which design- versus model-based approaches are used in practice varies by discipline and the nature of the research question under study, and this is further complicated by recent hybrids of the two approaches (e.g., Lenhard, 2006). Nonetheless, model-based sampling procedures implicitly underlie the majority of data sets within clinical psychology. Thus from a strictly practical perspective, most applications of IDA in clinical psychology will likely not pool data from samples in which probability weights are even available.1 For this reason, we focus our attention on the use of model-based approaches in IDA.
