Date of Award


Document Type

Campus Access Thesis


Communication Sciences and Disorders

First Advisor

Allen A. Montgomery


The examination of stuttered words dates back over 70 years. While recent studies have examined the speech of children who stutter (CWS) (Anderson & Byrd, 2008) or small parts of speech of adults who stutter (AWS) (Au-Yeung, Howell, & Pilgrim, 1998, little is known about the words produced disfluently by AWS. In this study, 803 disfluent words from 65 AWS were analyzed with the Irvine Phonotactic Online Dictionary - IPhOD (Vaden, Hickok, & Halpin 2009) to generate a population of words matched by phoneme structure with the disfluent words. The IPhOD list of 54,000 words (reduced to 19,000+ that matched by consonant-vowel syllable structure) and the 803 words attained from the AWS were compared by density, phonotactic probability and log10 frequency. In addition, the analyses were broken down by the type of word produced disfluently (content, function, or proper name.). Observations were also made regarding disfluency type (block, prolongation, part-word repetition, whole-word repetition and starter).

The results indicated that the word's neighborhood density for stuttered words was significantly greater than that of the general population of words. However, the statistical difference was not judged to be clinically or theoretically meaningful. The bi-phonotactic probability of stuttered words was not significantly significant when compared to that of the general population of words. The log10 frequency for stuttered words was significantly higher than the general population of words across all examined word structure types. Furthermore, among the 803 disfluent words obtained, content words were more commonly produced disfluently than function words, agreeing with the Au-Yeung et. al. (1998) findings for speech of AWS. Blocks and prolongations were the most common types of disfluencies. However, the most common type of disfluency for proper names was blocks. It was concluded that the speech of AWS differs from the speech of CWS when compared by neighborhood density, frequency, phonotactic probability, and word type (function vs. content words).