Abstract

Four mechanisms have been proposed to explain a developmental increase in short-term memory (STM) capacity: Rehearsal, encoding, organization, and persistence of memory trace. Two experiments were designed to test these mechanisms. Experiment I dealt with three factors, subjects' age, intelligence, and presentation time of stimuli, in a digit STM task. A significant interaction effect was observed between age and presentation time, but not between intelligence and presentation time. These findings suggest that intelligence and age of the subject affect the STM capacity via different mechanisms. In particular, age influences the STM capacity through speeding up encoding operation. Primacy effect was obtained, regardless of age, intelligence, and presentation time, suggesting that rehearsal is not a major determinant of the increase in STM capacity. Experment II further included the manipulation of categorical relatedness of words in a stimulus list. It was found that age affects encoding speed, whereas intelligence influences organization process. Rehearsal is again rejected on the ground that primacy effect was obtained in all conditions. Experiment II also revealed signicant interactions between presentation time and clustering size and between intelligence and clustering size at retrieval. These findings suggest that persistence of memory trace depends to a large extent on organization of items at input. Hence, three out of four mechanisms are proved to be major determinants of the STM capacity. These findings call for shift to a multi-process view of the development of the STM capacity.