Inferring Cognitive Processes Associated with Alcohol-Related Behavior: Reply to Kerby (1995)

With the acceptance of cognitive science, the biological substrate linking stimulus to response has come to be usefully viewed as hardware that can support learned software. A recent article focused on the development of software associated with alcohol consumption in the form of theoretical network models of alcohol expectancies. The comment by D. S. Kerby inappropriately removed from the context of theoretical modeling out discussion of the cluster analyses and suggested that these analyses were incorrectly interpreted to offer conclusion about real-time behavior and cognitive events. In fact, our interpretation is entirely consistent with model-building approaches, which are further explained. Other strategies for clarifying issues raised by this discussion also are reviewed