Abstract
BackgroundThe current study addresses the need for improvement in the assessment of executive function (EF). An unpublished measure of working memory (WM) ability, the Cleveland Verbal Working Memory Test with Interference (CVWMT-I), incorporates an element of distraction, which enables more accurate prediction of how WM ability is used in daily life. This study presents psychometric properties and normative data for the CVWMT-I.
Method
Ninety-six volunteers from the general population, between the ages of 21-79 were recruited for participation. Participants were selected based on education level (degree, college or secondary school), age and biological sex, to ensure even representation across each category, within the sample. Exclusion criteria included current depression or anxiety, psychiatric or neurological diagnosis, or history of brain injury. Participants completed a neuropsychological test battery, which included the CVWMT-I. Fortyeight participants completed version A of the CVWMT-I, whilst 46 completed version B, in order to measure parallel-form reliability. Thirty-five participants were audio recorded with consent, in order to measure inter-rater reliability, and 35 were re-tested on alternate forms four weeks later, to measure test-retest reliability. Convergent validity was measured via comparison of participant performance on background neuropsychological measures and the CVWMT-I. Pre-existing data on the CVWMT-I from 55 participants with traumatic brain injury (TBI) were compared to a demographically matched sample of 40 participants from the normative sample, to measure construct validity.
Results
Performance on the CVWMT-I was found to be mediated by IQ and age. A regressionbased norming method was used, stratifying for IQ and age, using the Reg_Build_MR_Raw.exe programme. Normative data is presented for use with this package, along with guidance on how to use it. Percentile rank tables are presented for information only. Two measures of the CVWMT-I, Letter Recall and Reverse Counting, were used for analyses. Letter Recall was found to be a sensitive measure of WM ability and correlated highly with other neuropsychological measures assessing similar constructs. Inter-rater and parallel-form reliability were high for both the Letter Recall and Reverse Counting measures, as was test-retest reliability for Reverse Counting. However, test-retest reliability was low for the Letter Recall measure. Normative participants performed better than those with TBI on the CVWMT-I, indicating good construct and ecological validity.
Conclusions
The psychometric properties of the CVWMT-I indicate that it is a good measure for use within clinical settings, with potential for wide accessibility, due to the employment of a regression-based norming method. Further research is required to assess clinical effectiveness of the CVWMT-I in identifying WM deficits within other clinical populations.
| Date of Award | 29 Oct 2019 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Robin Morris (Supervisor) & Jessica Fish (Supervisor) |