Age of acquisition, imagery, recall, and the limitations of multiple-regression analysis

Abstract

The first part of the paper describes the limitations of the stepwise multiple-regression technique used by Gilhooly and Gilhooly (1979) to assess the roles of age of acquisition and other variables in memory. The high intercorrelations between the variables make the technique inappropriate for the identification of those variables that influence recall and recognition. An experiment in which age of acquisition and imagery were manipulated is then reported. Better free recall occurred for late-acquired words, and it is suggested that this resulted from their greater distinctiveness and saliency. © 1981 Psychonomic Society, Inc.