Date of Award


Document Type

Campus Access Dissertation


Communication Sciences and Disorders

First Advisor

Julius Fridriksson


Purpose: Treatment for aphasia, a communication disorder resulting from damage to speech and language areas of the brain, has often been guided by clinical intuition and pragmatism. This manuscript aimed to identify predictors of effective aphasia treatment that could better indicate a patient's prognosis.

Method: This retrospective study analyzed behavioral and neurophysiological data collected from 26 individuals with chronic aphasia who had undergone intensive semantic and phonological treatment for anomia, a naming impairment common among all aphasic patients. Using trained and untrained items as measures of correct naming improvement, the following analyses were performed: 1) a partial correlation examined whether participants who began the semantic treatment first improved to a greater degree than participants who began the phonological treatment first; 2) a simple linear regression tested whether the average number of paradigm completions (APC) predicted an individual's change in correct naming; 3) paired t-tests ascertained whether fewer cues were needed for the phonological treatment than for the semantic treatment. An exploratory analysis used stepwise multiple regression to determine the best predictors of improved naming.

Results: Statistical analyses revealed: 1) treatment order, semantic first versus phonological first, did not significantly predict participant improvement; 2) the amount of practice, APC, significantly predicted improvement on treated items; 3) a phonological cueing hierarchy required significantly fewer cues than a semantic hierarchy, though no significant differences were found for correct naming improvement for either treatment. Results of a multiple regression exploratory analysis revealed that correct naming on the Boston Naming Test (BNT) and APC predicted correct naming improvement for treated items, whereas phonological errors on the Philadelphia Naming Test (PNT) pre-test predicted correct naming improvement for untreated items.

Discussion: The present findings support the use of behavioral measures to predict aphasia treatment outcome, though combinations of behavioral and neurophysiological variables may provide still more reliable outcome predictors.


© 2010, Julius Fridriksson