Verbal Punishment of Disfluencies During Spontaneous Speech

The subjects were 59 college adults who spoke spontaneously from a set of stimulus words for 36 minutes. The first 12 minutes was a Baseline segment during which disfluencies were counted. During the Treatment segment Contingent subjects were presented the word "wrong" through an earphone, contingent on each disfluency, Random subjects heard "wrong" on a random schedule, and no stimuli were introduced for Control subjects. During the Recovery segment the stimulus " wrong " was removed. Subjects were divided into high and low disfluency levels according to their performance during Baseline. A significant (p < 0.05) decrease in disfluencies was obtained from Baseline to Treatment for high level Contingent subjects. Low level Contingent subjects did not significantly alter their disfluencies, and Control and Random subjects maintained stable disfluency rates throughout the experiment. Rate of speech was not related to variations in fluency and also remained stable.