College students were presented with a series of four problems, each presented in four different formats: verbally, in abstract symbols, in negative abstract symbols, and in geometric figures. Problem-solving performance was evaluated in terms of the sequence of questions that a subject asks in order to solve a problem. They were also tested by the Holtzman Inkblot Technique, the California Psychological Inventory, and the Rokeach Dogmatism Scale. The data for the 161 students were analyzed into nine factors by the principal axis method and graphically rotated to an oblique structure. Subjects were classified in three groups (high, medium, and low) on each factor; and the groups were compared for their success in problem solving. Three personality factors appeared to be related to problem solving: a factor related to body image; an ability to define good forms; and a well balanced emotional life exhibiting control of hostility and of anxiety. The language used to present the problems was also discussed. (Author/CTM)