A third and still more difficult measurement scenario, again drawing from our own work, involves three studies in which participants were asked whether they believe or expect alcohol to relax them (see bottom half of Table 1). The response formats for this item varied greatly between studies and it is not immediately obvious how to construct a harmonized item. Without additional information to bridge the different item structures (see our later section on bridging studies), there is little justification for pooling these responses to enable IDA.
