Date of Award

1-2011

Degree Name

Doctor of Philosophy

Department

Interdisciplinary Health Sciences

First Advisor

Dr. Nickola W. Nelson

Abstract

The purpose of this three-paper dissertation was to explore the relationship between performance on verbal memory and language assessment tasks in school-age children. Study one used data from whole-class administration of three language assessment tasks to 2nd, 4th, 6th, and 9th graders (N= 187). Studies two and three used data from individualized administration of language assessment and verbal memory tasks to comparable samples of children between the ages of 6 and 18 years with typical language (TL) and language impairment (LI) (n = 37 per group).

The first study addressed practical and theoretical questions regarding memory and language demands of language tasks assessing nonword spelling, listening comprehension, and procedural direction following. Correlational analysis of the three tasks showed that they were only moderately correlated, suggesting that they were measuring different constructs. It was hypothesized that verbal memory might play a significant role in mediating each of the language tasks, but in different ways.

The second study was designed to explore the hypothesis developed in study one regarding the role of verbal memory in language assessment by incorporating measures of short-term and working memory with low language loads (digit span forward and reversed). Vocabulary awareness scores were combined to create a composite language variable with low demands on memory. Regression models showed different patterns of variables (age, vocabulary awareness, and memory) predicting nonword spelling, listening comprehension, and direction following performances of children with and without LI. Memory contributed unique variance to nonword spelling for both groups and to direction following for the students with TL but not LI.

The third study investigated patterns of association between memory skills and language skills. Four theory-driven hypotheses were presented. Results of classifications based on cut-off scores for a memory composite and language composite created profiles (high memory/low language, high memory/high language, low memory/low language, and low memory/high language). Analysis of these profiles using chi-square indicated statistically significant results disproving the null hypothesis of no association between memory and language. Further analysis concluded that the hypothesis that best fit the data was that memory is necessary, but not sufficient for language.

Access Setting

Dissertation-Open Access

Share

COinS