Stimulus recognizability judgments as a function of the utility of physical dimensions in recognition-discrimination problems

The effect of different Retraining conditions on the scaling of random forms for recognizability was examined. Pretraining conditions were defined by exposure to recognition-discrimination problems structured to emphasize or de-emphasize metron variability. Analyses of the results of the scaling task indicated that exposure to problems designed to emphasize metron variability within the “compactness” and “jaggedness” physical dimensions resulted in more extensive use of those dimensions, while no similar effect resulted from training with problems emphasizing an “x-axis areal asymmetry” physical dimension.