Effects of procedure on memory scanning of mildly mentally retarded adults

The performance of mildly mentally retarded adults on recognition memory tasks was compared with that of nonretarded control subjects. Following Sternberg's (1969) procedure, material to be remembered was presented in a fixed-set procedure, during which subjects were tested repeatedly on the same well-learned set of material, and a varied-set procedure, during which they were only tested on a memory set once before having to learn a new set. Mean reaction times (RTs) in all groups increased linearly as the number of items in the fixed memorized set increased; but "no" RTs of retarded adults tested under the varied-set procedure did not show this pattern. There was a gradation of slopes for the linear regression functions of RT on memory set size in both procedures, from less steep for nonretarded adults to increasingly steep values for nonretarded children and retarded adults. These results suggest that retarded adults use different processing strategies in the two procedures and that rate of processing increases with increasing mental age.