Inhibition of inappropriate responses is preserved in the Think-No-Think and impaired in the random number generation tasks in schizophrenia.

We examined the ability of 23 schizophrenia patients and 24 healthy controls to exert intentional inhibition of prepotent responses in the Think-No-Think (TNT) paired-associate learning paradigm (Anderson & Green, 2001). TNT manipulates the frequency (1, 8, 16 times) of intentional attempts to suppress (inhibit) some target words and to respond to most cue words. Following a TNT practice-phase, recall of suppressed words was tested in two ways, using the same cue words initially learned, and the category name plus letter-stem of the target words. Inhibition of prepotent responses was also examined in a random number generation (RNG) task. In TNT, speed results showed longer reaction times after 16 suppress attempts in patients, not in controls, reflecting increased difficulty with retrieving the memory traces of the overridden items. In accuracy, no between-groups differences were evidenced, and overall patterns replicated those of Anderson & Green. In RNG, patients produced more stereotyped responses and ascending and descending counting than controls, pointing to on-line failures to inhibit prepotent responses. These findings suggest that schizophrenia patients' difficulties to inhibit prepotent responses appear specific, not widespread, the intentional inhibition addressed in TNT being preserved, and on-line inhibition in RNG being impaired.