An investigation of the impact of interexaminer reliability on the outcome of phonological assessment

Includes bibliographical references (pages [38]-40)The purpose of this study was to investigate the impact of interexaminer reliability of whole word transcription on phonological assessment. Specifically, the present study was designed to determine how examiner variability in phonetic transcription influenced the assessment results of fifteen phonological processes. The Goldman-Fristoe Test of Articulation, Sounds-In-Words Subtest was utilized as a part of a complete speech and language evaluation for three preschool age children. The results of these evaluations indicated that all of the children demonstrated at least four different phonological processes and therefore, all were chosen for inclusion in this study. The tape recorded administrations of the Goldman-Fristoe Test of Articulation, Sounds-In-Words Subtest were played for ten speech-language pathologists to transcribe. The transcriptions were then transferred to Khan-Lewis Phonological Analysis forms and analyzed by the researcher. The results indicated that differences in phonetic transcriptions did influence the assessment results of the fifteen phonological processes, with the processes of velar fronting, liquid simplification, cluster simplification, deaffrication, palatal fronting and deletion of final consonants having the highest standard deviations for most of the children. However, it was noted that when a 40% criterion level was established for therapy placement, the differences in phonetic transcription had little effect on placement decisions.M.A. (Master of Arts