Validity and Reliability of Judgments of Authentic and Simulated Stuttering

¹ University of South Carolina, Columbia
² University of Southern California, Los Angeles

This exploratory study was undertaken to determine if the concept has merit that the information by which a stutterer identifies stuttering at time of occurrence is qualitatively different from that by which listeners identify stuttering. If it is not the same, then perceptually equivalent stuttered and nonstuttered speech disruptions should be experienced by the stutterer as qualitatively different, even though recordings of these disruptions would sound alike to listeners. To test this hypothesis a criterion was developed to validate a stutterer's ability to accurately identify her stuttering at time of occurrence. She simulated her own stuttering and then judged acoustical recordings of authentic and simulated samples at five intervals following occurrence. Listener judgments were also obtained. Listeners were able to distinguish simulated and authentic samples with 57% accuracy. The stutterer's judgments were never inaccurate at time of occurrence of a speech disruption, but her accuracy decreased rapidly following occurrence, plateauing at 54%. These results supported the concepts that the production of stuttered and nonstuttered speech disruptions are experienced as being qualitatively different, that the difference is involuntary blockage, that only stutterers can validly recognize this difference, and only when it occurs, and that stuttering is a speaker/production rather than a listener/perceptual disorder.

Key Words: stuttering • production definition • validity • reliability • listener agreement

Submitted on September 30, 1988
Accepted on June 21, 1989