Author ORCID Identifier

Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Robin Morris

Second Advisor

Lee Branum-Martin

Third Advisor

Jan Frijters

Fourth Advisor

Rose Sevcik


This study explored three theoretical models for understanding working memory (WM) in struggling readers. Research has shown relations between working memory and reading, especially on foundational phonologically based tasks like non-word repetition, however there is limited research exploring the predicted relations between non-language based WM tasks and basic reading abilities. Here three models of WM are used to explore the optimal theoretical framework for understanding both foundational reading abilities and response to intervention among a group of elementary school diagnosed with developmental dyslexia (DD). The three models were driven by theories from: 1. Baddeley and Hitch model; 2. Engle et al., model; and 3. Brown and Hulme model. 108 children (mean age: 9.01) in grades 3-4 were assessed as meeting criteria for a reading disability and subsequently participated in a 70-hour intensive reading intervention. Children were administered nine working memory, intelligence, and language tasks prior to the intervention and then tested at four time points on four single word reading tasks at baseline, 23, 45 and 70 hours over the course of the intervention. Three theoretical models were fit to the resulting data using structural equation modeling. Model 1 (Baddeley and Hitch, 1974) revealed the best fit for the data (RMSEA = 0.04; Model Chi Squared = 43.65, p=.18 (df=33); CFI =0.96), with Language, Intelligence and Phonological Working Memory factors showing strong relations to initial reading scores, but no factor showed relations to single word reading intervention change scores. These results suggest that phonological WM, visuo-spatial WM, and central executive WM reflect separate but related constructs. These data also confirm the fundamental connection between single word reading and phonological working memory and supports previous data showing minimal support for cognitive assessment as a means of predicting outcomes on reading intervention.


File Upload Confirmation