The trouble with babies and the value of bathwater: Complexities in the use of verbal reports as data.

Recent interest among behavior analysts in protocol analysis techniques prompts a consideration of some general measurement issues and some special issues relevant to protocol analysis. The development of behavior- analytic method and theory specific to verbal report research is a good thing, and Ericsson and Simon's (1984) book, Protocol Analysis, provides a useful model of integrating psychological theory and the craft of research. But protocol analysis techniques do not provide a magic window to the "world within the skin," and individual researchers should adopt these techniques only after confronting thorny issues such as how to determine the operating characteristics of verbal reports about private events, how to identify public performances to which protocol analysis can be applied productively, and how to maintain theoretical integrity in the empirical search for private events. We also caution against letting enthusiasm (and controversy) regarding protocol analysis distract behavior analysts from the benefits of using verbal report methods to study interesting events that are public in principle but difficult to measure in practice.