The association value of random shapes.

It is well known that verbal materials vary in meaningfulness or association value, and that these variations are related to learning and retention. Patterns and shapes may vary similarly; however, little systematic control over such variation has been exercised in studies of perceptual learning and retention. Indeed, little effort to standardize such materials has been expended (cf., e.g., the discussions by Hilgard [1951, p. 547] and Graham [1951, pp. 911-915]). A number of experiments have appeared in which random shapes have been used as stimuli in tasks involving perceptual learning and retention. Random shapes have also been employed in studies of mediated transfer and "predifferentiation." In most of these studies, control has not been exercised over possible effects of association value of the shapes upon performance of 5s. The present experiment was undertaken to provide a pool of random shapes with known association value for use in studies of the effects of certain stimulus variables and pretraining upon recognitive performance. It was considered desirable to provide for variation of association value and stimulus