Construct and ecological validity of short-term memory measures in retarded persons.

The performance of retarded subjects on five short-term memory tasks was analyzed in order to determine whether these tasks possessed construct and ecological validity. I predicted that certain components of each task required an active strategic process for optimal performance, whereas other components would only require a passive process. A principal axes factor analysis was performed on the intercorrelations of these components, and three factors were isolated that accounted for performance on the five tasks. Two of these factors were viewed as representing active and passive processes, and the third factor was determined by a recognition-memory task and, apparently, did not require the same processes as the recall tasks. These findings indicated that some of the short-term memory tasks possessed construct validity. An additional analysis was the investigation of the ecological validity of the various short-term memory tasks. Several multiple-correlation coefficients between memory scores and measures of adaptive behavior were significant and support the interpretation that the measures of short-term memory are ecologically valid.