Reliability and validity of an auditory working memory measure: Data from elderly and right-hemisphere damaged adults

Abstract

The use of non-standardized measures in research and clinical assessments creates difficulties with interpretation and generalization of results obtained. One example of a widely used non-standardized tool is the reading/listening span paradigm for assessment of working memory (WM). WM is an important construct because of its purported relationship to language comprehension and capacity theories of cognition. This paper investigates several facets of reliability and validity for an auditory working memory measure designed for older adults and individuals with right hemisphere brain damage (RHD). Results from 28 non-brain-damaged subjects (NBD) and 11 RHD subjects indicate that the measure is internally consistent and reliable over time. Construct validity evidence, which compares favourably with evidence from existing literature, suggests that for NBD sujects this tool differentiates WM from simple short term memory. RHD subjects do not demonstrate the same pattern of validity results as the NBD group. Further evaluation with RHD patients is warranted, because clinically this tool may be useful as a measure of severity or a prognostic indicator of language comprehension abilities for this population.