Date of Award

Spring 2024

Project Type

Thesis

Program or Major

Communication Sciences and Disorders

Degree Name

Master of Science

First Advisor

Jill C Thorson

Second Advisor

Amy Plante

Third Advisor

Donald Robin

Abstract

Clinical assessment of prosodic skills is essential when investigating linguistic and affective abilities. Prosodic elements include speech melody (intonation), phrasing distinctions, phrase level accentuation, lexical stress, affect, tempo, and rhythm. Prosody can be described as a hierarchy which allows for various prosodic skills, such as lexical stress, to be investigated independently. To identify a child who is displaying receptive or expressive prosodic deficits, which potentially impact language skills, an assessment tool for clinical settings must be developed. Speech-language pathologists (SLPs) often rely on perceptual skills to identify speech disorders, as there is no agreed upon prosodic assessment for evaluation. The Profiling Elements of Prosody in Speech-Communication (PEPS-C) is the most comprehensive assessment, but is not widely used due to length, administration issues, lack of normative data, and no acoustic counterpart to the perceptual analysis. To improve the assessment of linguistic skills, the Naturalistic Lexical Stress (NLS) subtests were developed to elicit lexical stress in a more natural and spontaneous interaction. The NLS expressive subtest controls for effects such as uncertainty contours (sounds like a question) and phrase final lengthening (last word in an utterance has a longer duration). The NLS receptive subtest assesses orthographic English word pairs which can be represented and recognized by children and adults. These word pairs differ in meaning based on stress. Thirteen typically developing children (TD), one diagnosed with childhood apraxia of speech (CAS), and seven children identified with a phonological impairment (PI) aged 5-11 were recruited to complete the PEPS-C and NLS tasks. Results from the receptive tasks were automatically scored by the PEPS-C program and the examiner. Expressive tasks were assessed perceptually and acoustically. The normalized pairwise variability index (nPVI) was measured using fundamental frequency (PVI_f0), duration (PVI_dur), and intensity (PVI_int). TD and PI groups demonstrated higher accuracy on the NLS subtests than the PEPS-C. Acoustic measures of nPVI_dur and nPVI_int were found to be significant indicators of lexical stress. This supports the use of more naturalistic tasks in assessment.

Available for download on Tuesday, May 01, 2029

Share

COinS