Measurement of stuttering in adults: comparison of stuttering-rate and severity-scaling methods

This study investigated the comparative reliability of 2 stuttering measurement tools when used by experienced judges: percentage of syllables stuttered (%SS) and a 9-point severity scale (SEV). The study also investigated the degree to which scores on 1 tool predict scores on the other and the distributions of stuttering when measured by these tools. Twelve experienced judges watched 3-min videotapes of 90 stuttering and 10 nonstuttering participants. Half the judges rated %SS, and half made severity ratings. Results showed very high intrajudge and interjudge agreement for both measures. There was a strong linear correlation between %SS scores and SEV scores. Based on this finding, it seems that the 2 measures can be regarded as largely interchangeable. The exception to this, however, was in cases where there was either a small number of significant fixed postures (blocks and prolongations) or a large number of innocuous repeated movements (repetitions) in the speech sample. In such cases, it appears that %SS and SEV scores combined would be needed to provide a valid measure of stuttering. SEV scores were more normally distributed than %SS scores, which were clearly skewed. The advantages and limitations of the SEV scale are discussed.