Protocol analysis as a tool for behavior analysis.

The study of thinking is made difficult by the fact that many of the relevant stimuli and responses are not apparent. Although the use of verbal reports has a long history in psychology, it is only recently that Ericsson and Simon's (1993) book on verbal reports explicated the conditions under which such reports may be reliable and valid. We review some studies in behavior analysis and cognitive psychology that have used talk-aloud reporting. We review particular methods for collecting reliable and valid verbal reports using the "talk-aloud" method as well as discuss alternatives to the talk-aloud procedure that are effective under different task conditions, such as the use of reports after completion of very rapid task performances. We specifically caution against the practice of asking subjects to reflect on the causes of their own behavior and the less frequently discussed problems associated with providing inappropriate social stimulation to participants during experimental sessions.