Item analyses of memory differences.

OBJECTIVE Although performance on memory and other cognitive tests is usually assessed with a score aggregated across multiple items, potentially valuable information is also available at the level of individual items. METHOD The current study illustrates how analyses of variance with item as one of the factors, and memorability analyses in which item accuracy in one group is plotted as a function of item accuracy in another group, can provide a more detailed characterization of the nature of group differences in memory. Data are reported for two memory tasks, word recall and story memory, across age, ability, repetition, delay, and longitudinal contrasts. RESULTS The item-level analyses revealed evidence for largely uniform differences across items in the age, ability, and longitudinal contrasts, but differential patterns across items in the repetition contrast, and unsystematic item relations in the delay contrast. CONCLUSION Analyses at the level of individual items have the potential to indicate the manner by which group differences in the aggregate test score are achieved.