Research based on a comparison of individual-level data from different sources has uncovered substantial & systematic retrospective bias in interviews. Aside from these direct measurement procedures, researchers have also used indirect methods that generally involve a comparison of aggregated data; however, these indirect studies have been criticized on logical grounds. Here, results obtained using an indirect measurement procedure (see SA 35:4/87R6870) are compared to those obtained with a direct approach, using the same data set. It is concluded that the indirect procedure does not yield reliable & stable results, leading to invalid conclusions about retrospective bias in interviews. 2 Tables, 20 References. Modified HA