Establishing the Reliability of a Verbal Self-Monitoring Paradigm

Background: Auditory hallucinations in schizophrenia are associated with defective monitoring of inner speech. This process can be examined using a computerised task of verbal self-monitoring. Cognitive therapy for voices aims to improve the patients' self-monitoring ability, and changes in self-monitoring could be assessed using this task. Before using the paradigm in this way, we sought to determine its test re-test reliability. Sampling and Methods: 20 normal volunteers were tested on the paradigm on two occasions, 3 months apart. The task involved reading single adjectives under the following randomized conditions: reading aloud; reading aloud with acoustic distortion of their own voice; reading aloud with alien feedback (someone else's voice), and reading aloud with distorted alien feedback. Immediately after articulating each word, the participants were required to identify the source of the speech they heard ('self'/'other'/'unsure') via a button press. Response choice and reaction time were recorded. Results: There was no change in error rate over the two testing occasions; however, participants had faster reaction times when tested for the second time. This reduction in reaction time was not specific to type of response (correct, error, unsure). Conclusions: Participants showed no evidence of practice on the paradigm in terms of making fewer errors or responding faster for correct responses. The overall reduction in reaction times on the second occasion may reflect increased familiarity with the task. The verbal self-monitoring paradigm appears to be a reliable measure, which could be used to evaluate changes in self-monitoring ability in hallucinators following psychological therapy. Copyright © 2003 S. Karger AG, Basel.