Punishment of stuttering: contingency and stimulus parameters

The influence of two major parameters of time out from speaking were investigated with 36 adolescent and adult stutterers. Subjects were assigned to four matched groups in an attempt to examine whether time out has to be contingent to be effective and whether simply signaling stuttering (no time out) produces response reduction. The major finding was that reliable reductions in stuttering frequency were produced only when tone presentations were contingent and signaled time out from speaking. Overall results are discussed in terms of their implications for the various major models used to account for the effects of response-contingent procedures on stuttering. Assigning a label of distraction, attention, or punishment is less important than identifying the parameters responsible for observed effects.