Seven experiments were conducted concerning decision making and information processing under conditions of uncertainty. Several different experimental tasks were used; all presented the subject with multiple independent sources of information regarding the likelihood that some event would occur. Study 1 subjects were Air Force pilots; all other subjects were undergraduate college students. The independent variables included the number of inputs, the inputs' reliability, the effects of discrepant inputs, the format of the problem, and the time available to respond. All tasks and subsequent analyses were non-Bayesian, yielding both normative and idiographic information. The subjects' basic instruction was to indicate the event most likely to occur or to estimate the likelihood of a given event. Results indicated no significant differences between F-16 pilots and student pilots in their use of averaging as the predominant strategy chosen. Simpler strategies were adopted when the time allocated for the task was reduced. The consistent use of a strategy was disrupted when the subject's first experience to the task was under time restrictions. When postdecision feedback was unreliable, the consistency of subsequent decision-making patterns was disrupted. The effects of information reliability were equivocal. Unreliable information was sometimes incorporated into the decision-making processes and sometimes ignored. (Author/YLB)