False memory and schizophrenia: evidence for gist memory impairment.
BACKGROUND: In this study we examined patients' false memory, that is memory for a non-presented event, to search for a further source of converging evidence for the impairment of semantic memory in individuals with schizophrenia. In two experiments we compared the pattern of false memory created by the Deese-Roediger-McDermott (DRM) paradigm between individuals with schizophrenia and those of a normal control group. METHOD: Experiment 1 tested participants on both recall and recognition of lists of semantically related words. Experiment 2 adopted the meaning recognition test, in addition to the standard recognition test, to assess the participants' gist memory. RESULTS: Individuals with schizophrenia performed worse than normal controls on both recall and recognition of studied words. The schizophrenia patients had higher rates of false recall and false recognition for semantically unrelated words than did the normal controls, suggesting an abnormal pattern of semantic activation in the former group. More importantly, no differences were found between the two groups with regard to false recall and false recognition of semantically related words. When the participants were tested for meaning recognition, however, the schizophrenia patients gave fewer 'old' responses to non-studied semantically related words than did the control group, indicating an impaired gist memory in schizophrenia patients. CONCLUSIONS: When asked to consciously retrieve word lists, individuals with schizophrenia showed impairment not only in item-specific memory but also in gist memory. The pattern of results is consistent with the storage deficit view of semantic memory in schizophrenia.