The effect of instructional bias on listener ratings of specific disfluency types in children

This study evaluated the influence of instructional bias on severity ratings of children's disfluencies. Normal speaking children were trained to simulate specific types of disfluency and severity judgments were made by judges who had been exposed to instructional bias. The 75 judges who participated in the study were divided into three groups of 25 each and were given a bias “toward stuttering,” bias “toward fluency,” and a “neutral” bias toward neither perceptual set. Both univariate and multivariate analyses were employed for data analysis.Although predictable differences were found between the specific types of speech disfluency, no significant differences were found for the experimental bias conditions. These findings, based on molecular analyses, differ from previous results based on molar speech samples. Molecular approaches to understanding the nature of stuttering were found to be powerful modes of assessment, and it is suggested that the type of disfluency emitted by children may be more important than the instructional bias or perceptual set of the listener.