Sök artiklar i SveMed+

Tidskrift

Sammanfattning

Tests can display different psychometric properties in clinical samples compared to the normal standardization sample. However, it is particularly important that the test measures reliably and valid in the clinic. The present study examines three psychometric issues relating to the new Scandinavian version of the Wechsler Adult Intelligence Scale, version IV, in a mixed clinical sample of 226 subjects. As some verbal items are new to the Scandinavian version, and original items are rearranged according to the performance of the relatively small Scandinavian pilot study, item analyses are performed to test whether verbal items become increasingly difficult as intended. Imperfect increase in difficulty level may have detrimental consequences to the validity of the stricter Stoprules in WAIS-IV. The rank order correlations between scores with different stop-criteria are examined. The third problem relates to the extension of working memory testing in the WAIS-IV by including a digit span sequence task. A Confirmatory Factor Analysis (CFA) is performed to test whether the New Digit Span test differentiates between storage and manipulation levels of working memory (WM). All subjects are also tested with the Spatial Span subtest from the Wechlser Memory Scale-III. Results: With some few exceptions the item analysis showed the expected triangular form indicating increased difficulty level. Analyses of different stop criteria showed insufficient rank order correlations for Visual Puzzles, Information and Comprehension when applying the three zero-score stop rule prescribed in the handbook. The CFA showed best fit for a modality division of WM. Conclusion: The study shows some insufficiencies that the clinician should be aware of when using the Scandinavian WAIS-IV version in the clinic.

MeSH