Intact Implicit Learning in Schizophrenia: Evidence from Three Implicit Learning Tasks

Abstract

The aim of this thesis was to determine whether people with schizophrenia show signs of implicit learning on a variety of implicit learning tasks. Experiment 1 aimed to replicate Green, Kern, Williams, McGurk, and Kee's (1997) study by comparing people with schizophrenia to controls on a serial reaction time (SRT) task. The hypotheses that both groups would show signs of implicit learning but that the patients would have slower reaction times (RT) overall were supported. However, one-third of the patients was impaired on the SRT task (non-learners). The only discernible difference between the learners and non-learners was that the non-learners had had schizophrenia for considerably less time than the learners. Experiment 2 explored this finding further by comparing patients who had had schizophrenia for more than ten years (long duration) with patients who had had schizophrenia for less than ten years (short duration) on the SRT task. Taken together, both groups showed signs of implicit learning on the SRT task, although a subgroup of people from the long duration group were impaired on the SRT task. Additional analyses that pooled subjects from Experiments 1 and 2 suggested that the learners-non-learners dichotomy was likely to be just an artifact of small sample sizes. The aim of Experiment 3 was to determine whether SRT non-learners were impaired on other implicit learning tasks. To investigate this, SRT learners and SRT non-learners from Experiments 1 and 2 were compared on the SRT task, a Tower of Hanoi (TOH) task and a mirror-drawing task. Furthermore, all the SRT non-learners from the previous experiments completed the SRT task for a second time. As expected, people with schizophrenia demonstrated implicit learning abilities on the SRT task but overall RT was slower and rate of learning was somewhat less than those of controls. Despite taking longer than controls to complete the TOH task, overall rates of learning between patients and controls were comparable. As a group, patients with schizophrenia also demonstrated implicit learning on the mirror drawing task, although controls were faster and made fewer errors than patients did. However, because controls were almost at ceiling level on this task, patients actually had higher learning rates, both in terms of speed and accuracy. In conclusion, although some people with schizophrenia may take longer to respond and make more errors overall compared to controls, implicit learning on a variety of tasks appears to be largely unaffected by schizophrenia.