Predicting variability from perceived situational similarity

An interactional approach to the issue of transsituational consistency which relates persons and situations via the mediating variable of a person's world view was suggested. A model for the mediating variable was proposed which emphasized the idiographic distortions of a shared nomothetic structure of the world. This model was tested by using the INDSCAL algorithm to recreate 100 individual's judgments of similarity between 28 pairs of social situations. These INDSCAL-derived similarities (which conform to the model of the proposed mediating cognitive structure) significantly predicted variability across the situations for 58% of the subjects. The most predictable 50 subjects had an average correlation of 38 between the InDSCAL-recreated similarities and their self-reported transsituational variability. The least predictable 50 subjects had an average correlation of only .09. The degree of fit of the INDSCAL model to the initial data was significantly related (r = .39) to how well a particular subject's variability scores could be predicted from the INDSCAL model. The advantages of the proposed idiographic/nomothetic model over a purely idiographic model were discussed.