Memory for everyday texts and the WMS : the usefulness of the Logical Memory subtest in clinical settings

Abstract

The aim of this study was to explore the usefulness of the Logical Memory subtest from the Wechsler Memory Scale (WMS) as an effective means of assessing everyday text-processing ability in clinical settings. The text recall performance of a group of well-educated normal memory and memory impaired subjects was assessed on the paragraph-length texts (Anna Thompson & the American liner New York)from the WMS Form 1and on a set of three everyday texts (representative of different text genres and modes of presentation). The everyday texts were a short story which was read to subjects, a newspaper article subjects read and a documentary film. Text recall was gauged across three time periods - immediately, 30 minutes and one week after presentation. Logical Memory scores were computed for each subject as per the test instructions. All texts including the WMS texts were analysed into gist (i.e. the key information which contains the main meaning of the text) and details (supporting information). Scores for gist and details were computed across texts and subjects. The results of the study found that the American liner passage (WMS) did not adequately distinguish between well-educated normal and memory impaired subjects. Furthermore, the WMS passages were found to be poorly written examples of news reports and judged by subjects to be difficult to recall. Some items from each text were found to be substantially and consistently more memorable over time than others. The Logical Memory subtest, however, does not acknowledge the different memorability of items and, thereby loses power as an assessment device. Brain trauma clients who are able to recall key information from a text are less disabled than those who cannot. Recall of the everyday texts was also found to be different in nature to that of the WMS texts across a number of dimensions. For the everyday texts but not for the WMS texts, gist was consistently more memorable than details across subjects and time. This indicates that with everyday texts subjects use strategies which appropriately select the gist or the main meaningful structure from a text given free-recall instructions. The WMS texts, on the other hand, were found to encourage details recall and rehearsal strategies because of the details focus in the texts themselves and the verbatim recall instructions. For the everyday texts, the recall of the memory impaired subjects (although diminished and prone to decay) was characteristic of normal memory functioning i.e. memory impaired like normal memory subjects recalled the most memorable items and more gist than details. In contrast, the WMS texts presented lists of details which were largely irrelevant to the meaning of the text. Normal memory subjects could cope with this difficult and unnatural task but it overtaxed memory impaired individuals. Logical Memory scores were found to be the most predictive of the recall of the short story for memory impaired subjects but did not successfully predict the recall of the documentary film or the newspaper article. Case study data found that Logical Memory scores did not convincingly predict the recall of everyday texts (but particularly the gist recall) for most memory impaired individuals over time. The study discussed the problems inherent in the Logical Memory subtest, specifically problems with: the texts not being representative of everyday texts; the lack of acknowledgement of the different memorability of items and of the important distinction between gist and details items; the verbatim test instructions which discourage high-level text­ processing strategies; the delayed recall period of only 30 minutes not being adequate to assess text recall for educational purposes; and the questionable predictive validity of the subtest for everyday text recall. The study highlighted the need for an assessment device for clinical purposes to be developed which uses everyday texts and everyday contexts as recall tasks. Show more