Seventy-two college students were divided into three groups: Button Push-Speech (BP-S), Speech-Button Push (S-BP), and Control. BP-S subjects pushed one of two buttons on signal for 8 min. During the last 4 min, depression of the criterion button caused a buzzer to sound. After the button-push task, subjects spoke spontaneously for 30 min. During the last 20 min, the buzzer was presented contingent upon each disfluency. S-BP subjects were run under the same procedures, but the order of button-push and speech tasks was reversed. Control subjects followed the same procedures as S-BP subjects, but no buzzer signal was presented at any time.
Both S-BP and BP-S subjects emitted significantly fewer disfluencies during the last 20 min (Conditioning) than during the first 10 min (Baserate) of the speaking task. The frequency of disfluencies for Control subjects did not change significantly from Baserate to Conditioning. In none of the three groups did the frequency of pushes on the criterion button change significantly from minute to minute throughout the 8-min button-push session.
