Effects of Measurement Method and Transcript Availability on Inexperienced Raters’ Stuttering Frequency Scores

Publication date: Available online 24 April 2018
Source:Journal of Communication Disorders
Author(s): Nalanda Chakraborty, Kenneth J. Logan
PurposeTo examine the effects of measurement method and transcript availability on the accuracy, reliability, and efficiency of inexperienced raters' stuttering frequency measurements.Method44 adults, all inexperienced at evaluating stuttered speech, underwent 20 minutes of preliminary training in stuttering measurement and then analyzed a series of sentences, with and without access to transcripts of sentence stimuli, using either a syllable-based analysis (SBA) or an utterance-based analysis (UBA). Participants' analyses were compared between groups and to a composite analysis from two experienced evaluators.ResultsStuttering frequency scores from the SBA and UBA groups differed significantly from the experienced evaluators' scores; however, UBA scores were significantly closer to the experienced evaluators' scores and were completed significantly faster than the SBA scores. Transcript availability facilitated scoring accuracy and efficiency in both groups. The internal reliability of stuttering frequency scores was acceptable for the SBA and UBA groups; however, the SBA group demonstrated only modest point-by-point agreement with ratings from the experienced evaluators.ConclusionsGiven its accuracy and efficiency advantages over syllable-based analysis, utterance-based fluency analysis appears to be an appropriate context for introducing stuttering frequency measurement to raters who have limited experience in stuttering measurement. To address accuracy gaps between experienced and inexperienced raters, however, use of either analysis must be supplemented with training activities that expose inexperienced raters to the decision-making processes used by experienced raters when identifying stuttered syllables.



https://ift.tt/2JpK1q8