Two studies, both part of a research program based on a theory of clinical problem solving behavior, used a computer to simulate a clinician interacting with a child having reading difficulties. The impact of routine cue collection on a clinician's performance is the focus of the first study; the process of hypothesis generation as it is affected by different certainty thresholds for considering a hypothesis is the subject of the second. The results of these studies suggest that the use of routine cue collection can increase diagnostic accuracy by as much as 15% to 20%, and that early hypotheses generation may be a result of hypothesis directed inquiry. (AEA)