Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
Date of Graduation
5-6-2021
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Department of Communication Sciences and Disorders
Advisor(s)
Geralyn R. Timler
Stacey L. Pavelko
Carol C. Dudding
Dannette A. Bronaugh
Abstract
Purpose: Conversation, narrative, and expository language sampling contexts are recommended for school-aged children (Pezold et al., 2020), and multiple ways to analyze these samples have been promoted in the clinical literature. This dissertation addressed two gaps in the literature related to analyses and sampling contexts. The purpose of study one was to examine differences in two commonly-used language sample analysis methods, Sampling Utterances Grammatical Analysis Revised (SUGAR) and Systematic Analysis of Language Transcripts (SALT). The purpose of study two was to examine the presence of age-related changes in conversation, narrative, and expository contexts for older school-aged children.
Method: Conversational, narrative, and expository language samples were elicited from 85 typically-developing children ages 8-11. In study one, the conversation language samples were transcribed and analyzed using both the SUGAR and SALT conventions for all children ages 8-10 (n=68).Four paired language performance metrics were calculated including (a) Total Number of Words (TNW) and Number of Total Words (NTW), (b) Mean Length of Utterance SUGAR (MLUs) and Mean Length of Utterance in Morphemes (MLUM), (c) Words Per Sentence (WPS) and Mean Length of Utterance in Words (MLUW), and (d) Clauses Per Sentence (CPS) and Subordination Index (SI). Z-scores were calculated by comparing the computed values to their respective normative databases. Differences in z-scores were examined for each paired metric. Next, to determine if SALT and SUGAR conventions would classify children differently based on language status, these z-score values were compared to thresholds suggestive of typical development versus suggestive of language impairment. In study two, the conversation, narrative, and expository language samples were transcribed and analyzed for mean length of utterance in words (MLUW) and subordination index (SI), a measure of complex language, using SALT conventions.
Results: Results for study one revealed statistically significant differences in mean z-score values for all paired language sample analysis metrics. The SUGAR z-scores were significantly lower for the TNW/NTW, MLUs/MLUM, and WPS/MLUM comparisons, but SUGAR z-scores were significantly higher for the CPS/SI comparison. In addition, while no children fell below the thresholds suggestive of language impairment in SUGAR, four children fell below the threshold using the SALT conventions. Study two revealed no age-related changes in MLUW or SI. Sampling contexts did yield significant differences. Narrative and expository language samples elicited longer utterances and more complex language than conversation samples.
Conclusions: Comparisons between different normative databases for language sample analyses may yield differing clinical conclusions. Although age-related changes were not detected in any context, the results indicate that narrative and expository contexts may be more appropriate language sampling contexts to examine complex sentence use in school-aged children ages 8-12 years.
Recommended Citation
Lenhart, Michelle, "Language sampling with older school-aged children: Examination of analyses and sampling contexts" (2021). Dissertations, 2020-current. 50.
https://commons.lib.jmu.edu/diss202029/50
