Date of Award

Spring 1-1-2013

Document Type


Degree Name

Doctor of Philosophy (PhD)


Speech, Language & Hearing Sciences

First Advisor

Gail Ramsberger

Second Advisor

David B. Arciniegas

Third Advisor

Albert Kim

Fourth Advisor

Neeraja Sadagopan

Fifth Advisor

Brenda Schick


Patient Characteristics and Treatment Components that Mediate Improvements in Connected Speech in Persons with Chronic Post-Stroke Aphasia: A Delphi study involving a Communication Disorders Expert Panel Dissertation directed by Associate Professor Gail Ramsberger Purpose: To achieve agreement amongst a group of aphasia experts regarding the most important patient characteristics and treatment components that mediate improvements in connected speech in persons with chronic post-stroke aphasia. Method: A three-round Delphi study obtained the views of experts regarding patient characteristics and treatment components (e.g., focus, materials, task, duration/intensity, other) that positively, neutrally, or negatively influence connected speech. A structured eight-stage process included two data acquisition methods. First, a systematic literature review extracted patient characteristics and treatment components in studies for which connected speech was an outcome variable of interest. Next, over three-rounds, experts identified and rated patient characteristics and treatment components according to their relative influence on connected speech. Means, standard deviations, percentiles, and Spearman rank correlation coefficients were used to investigate the importance of items, agreement amongst experts, and stability of responses across rounds. Results: Fifty-two experts identified in the literature review were contacted to participate in the study, seven of whom completed all three Delphi rounds. Categorization of the narrative responses to first round questions resulted in 175 specific, five-point Likert-scale questions After expert ratings of Round two and three questions, 53 final items achieved greatest certainty to positively or negatively influence connected speech. Of these 53, 38 achieved high expert consensus, 11 moderate consensus, and 4 low consensus. Twenty-three of the ratings had strong stability across rounds, 38% had moderate, and 26% had weak. Discussion: Comparison of the literature and expert responses in the Delphi study revealed contrasting thoughts on variables that influence connected speech. Studies identified in the literature review emphasized treatment of microlinguistic elements of language. In contrast, experts' responding to the Delphi questionnaire indicated that treatment should be directed toward macrolinguistic elements of language. Through this Delphi study, a panel of international aphasia rehabilitation experts identified and agreed upon key elements of aphasia therapy aimed at improving connected speech. When the tasks, materials, and types of feedback are considered, and treatment is constructed, these key elements can be integrated and manipulated and thus be used as mechanisms of, not just variables in, change.