Investigators are urged to plan factor analytic studies prior to collecting the data, to formulate a hypothesized factor structure, to develop several relatively pure measures of each factor expected, and to select an appropriate sample of at least 200 cases. Continuous variables should be used rather than dichotomous variables. (Author)

Accounting for verbal reports requires explication of the mechanisms by which the reports are generated and influenced by experimental factors. We discuss different cognitive processes underlying verbalization and present a model of how subjects, when asked to think aloud, verbalize information from their short-term memory. (Author/GDC)