Performance on the Wisconsin Card Sorting Test after Traumatic Brain Injury

Abstract

The performance of 100 patients with traumatic brain injury (TBI) on the Wisconsin Card Sorting Test (WCST) was examined to determine the underlying latent structure as well as profile subtypes. Exploratory factor analysis with oblique rotation identified three factors: response accuracy, learning, and failure to maintain set. Marker variables from each factor were then included in a two-stage cluster analysis, which identified two reliable subtypes. Level of performance differences between these subtypes were meaningfully related to both demographic background and length of coma. It is concluded that interpretation of the WCST results of patients with TBI should consider the multifactorial nature of the instrument in combination with injury severity characteristics and demographic variables.

Wisconsin Card Sorting Test in chronic severe traumatic brain injury: factor structure and performance subgroups

“…The present participants differ from those studied by Greve et al [2] and Wiegner and Donders [3] in terms of both the severity and chronicity of the injuries. This study will use two sets of raw scores obtained from two administration variants and demographically corrected t-scores to examine the similarity of factor structures across different types of scores.…”

“…The two-factor solution revealed a factor pattern very similar to the Wiegner and Donders [3] Factors 1 and 3 (since Trials to Complete 1st Category and Learning to Learn were not included in the present analyses, the factor on which they load was obviously not observed in this analysis). However, in their analysis, NPE did load significantly on Factor 1 as it did in an earlier factor analysis of WCST performance in college students [10].…”

“…Following Wiegner and Donders [3], the present analysis used both Ward and Complete Linkage clustering methods. The Ward method produced a four-cluster solution.…”

“…This analysis produced a three-factor solution which was consistent with the solution generated in their general clinical sample. In a more recent study, Wiegner and Donders [3] presented a three-factor solution in a higher functioning (based on WCST performance) TBI sample (nˆ100). Interestingly, despite the fact that rarely reported WCST scores (Learning to Learn; Trials to Complete First Category) were included, their results were consistent with those of Greve et al [2].…”

“…This study will use two sets of raw scores obtained from two administration variants and demographically corrected t-scores to examine the similarity of factor structures across different types of scores. Finally, in an attempt to further illuminate the processes underlying WCST performance, a cluster analysis following the Wiegner and Donders [3] approach was conducted. …”