Toward a model of dissociating implicit and explicit memory components.

点击翻译按钮获取中文摘要

The aim of this research was to dissociate implicit and explicit components in a memory task and to explicate the nature of these two components. Three different methods were used : remember/know judgment, pathological dissociation, and developmental dissociation. In the first experiment, college students either read target nouns or generated them in response to their definitions. A recognition test was then administered : The subjects were required to make remember/know judgment and confidence rating on the items they had called old. The results showed that remember/know responses significantly correlated with confidence ratings. In the second experiment, alcoholic Korsakoff patients and control subjects studied a list of Japanese nouns written in different scripts. When the writing script was changed between study and test phases, priming in word-fragment completion was attenuated but was still reliable as compared with baseline performance. This was confirmed in both Korsakoff patients and control subjects. In contrast, the script change had little effect on recognition memory, which was severely impaired in Korsakoff patients. In the third experiment, elementary school children and college students were given implicit and explicit memory tests following either a read or a generate task. As a result, children's priming under the generate condition was lower than that under the read condition, whereas there was no difference in adults' priming between these two conditions. In the fourth experiment, mentally retarded persons exhibited a profile similar to that of children. Taken together, these findings suggest that there are two different components in implicit memory, one that shows no developmental or pathological difference and heavily relies on perceptual processes and another that shows an age-related or intelligence-related increase and heavily relies on lexical or conceptual processes.