Investigating the psychometric properties of a South African adaptation of the Boston Naming Test : evidence for diagnostic validity from a memory clinic population

Master Thesis

2013

Permanent link to this Item
Authors
Supervisors
Journal Title
Link to Journal
Journal ISSN
Volume Title
Publisher
Publisher

University of Cape Town

License
Series
Abstract
The Boston Naming Test (BNT) is a popular confrontation naming test that is frequently used in the detection of naming deficits in Alzheimer's disease (AD). However, the test may not be appropriate when used outside of North America due to the influence of varying word frequency and familiarity between different cultures and languages. This study investigated the diagnostic validity of a South African 15-item adaption of the BNT (the BNT-SA-SF) in a Cape Town memory clinic population of patients with dementia and healthy, community-dwelling control participants. Between-groups comparisons, receiver operating characteristic (ROC) analyses, and other diagnostic efficiency statistics were used to assess the test's discriminative capacity between patients with AD (n = 46), patients with other types of dementia (n = 23), and controls (n = 51), matched on key demographic variables. The AD group performed worse than patients with other types of dementia and controls on the BNT-SA-SF, and patients with other types of dementia scored more poorly than controls. The test showed the most significant discriminative capacity between patients with AD and controls, however. A general linear model examining the effects of socio-demographic variables on test performance found that BNT-SA-SF performance was not significantly affected by the socio-demographic characteristics of participants, including age, education, language, or socio-economic status, with the exception that men appear to achieve higher scores than women. Further, an item analysis identified a number of problematic items and suggestions are made concerning how to deal with these in future studies. Preliminary normative data stratified by sex and education are presented. Results support the clinical utility of the BNT-SA-SF as a screening test to aid in the diagnosis of AD from normal aging with older adults in South Africa. This study is a valuable step forward in the ongoing attempt to provide culturally appropriate and valid neuropsychological tests and norms for clinical and research purposes in South Africa. Future studies should examine the functioning of the test in larger samples, representative of the other major population and language groups in South Africa.
Description

Includes abstract.

Reference:

Collections