Working memory after acquired brain injury: listening span recall.

University of Minnesota M.A. thesis. June 2011. Major: Speech-language-hearing sciences. Advisor: Dr. Mary R. T. Kennedy. 1 computer file (PDF); vi, 68 pages, appendices A-C.Twenty-three mildly impaired adults with acquired brain injury (ABI) and eighteen carefully matched healthy controls performed three commonly used working memory tasks (WM): the digit span, n-back task, and listening span task (Tompkins, Bloise, Timko, & Baumgaertner,1994). In a preliminary study, Baumgarten (2009) administered these to a small group of mildly impaired adults with ABI and controls and found that participants with ABI made more errors on the listening span task, but did not perform worse on the n-back or digit span tasks compared to controls. The present study followed the same methods and procedures used in Baumgarten (2009) with the addition of error analysis by type for recall errors made on the listening span task. Recall errors were coded as either intrusions or omissions. Intrusion errors were broken down into within-task intrusions, categorical intrusions, non-categorical intrusions and phonemic intrusions. Analysis of variance (ANOVA) was used when there were both between- and within-group comparisons, and simple group comparisons were done in the absence of within-group variables. Adults with ABI made more total errors and more omission errors than controls on the listening span task, however groups iii Abstract Twenty-three mildly impaired adults with acquired brain injury (ABI) and eighteen carefully matched healthy controls performed three commonly used working memory tasks (WM): the digit span, n-back task, and listening span task (Tompkins, Bloise, Timko, & Baumgaertner,1994). In a preliminary study, Baumgarten (2009) administered these to a small group of mildly impaired adults with ABI and controls and found that participants with ABI made more errors on the listening span task, but did not perform worse on the n-back or digit span tasks compared to controls. The present study followed the same methods and procedures used in Baumgarten (2009) with the addition of error analysis by type for recall errors made on the listening span task. Recall errors were coded as either intrusions or omissions. Intrusion errors were broken down into within-task intrusions, categorical intrusions, non-categorical intrusions and phonemic intrusions. Analysis of variance (ANOVA) was used when there were both between- and within-group comparisons, and simple group comparisons were done in the absence of within-group variables. Adults with ABI made more total errors and more omission errors than controls on the listening span task, however groups did not differ in total intrusion errors or in specific intrusion error types. All participants made more omission than intrusion errors on the listening span task. Performance on the digit span task and the n-back task were similar between groups. The listening span task appears to capture WM that has a linguistic base, including pre-injury vocabulary and post-injury word fluency. The clinical significance of this is discussed