Cognitive mechanisms underlying perseveration on the Wisconsin Card Sorting Test
Four studies were undertaken with the aim of investigating primarily task conditions associated with perseveration on the Wisconsin Card Sorting Test (WCST). In addition, the last study examined the possibility that perseveration may be associated with a reduced emotional reaction to feedback received during the WCST. The prevalent idea regarding perseveration has been that it is a result of an inability to disengage from a behaviour that is "overlearned" or that has become habitual (Milner, 1963). Following this premise, it was initially hypothesised that perseveration on a particular response would increase as reinforcement of this particular response increases. In the first study, the standard WCST as well as two variations differing in number of correct consecutive trials per category (CCT) (WCST5 and WCST15) were administered to twenty individuals with brain injury and an equal number of healthy participants. The purpose of the study was to examine a possible dif ference in perseveration when individuals were offered more (15) or less CCTs (5) for each sorting category on the WCST. Contrary to predictions, the healthy group showed significantly fewer perseverative errors and total errors on WCST15 when compared both to WCSTI0 and to WCST5 although no significant difference was found between WCST10 compared to WCST5. The brain injured group showed significantly more perseveration on CST15 compared to WCSTIO but not compared to WCST5. Also CSTIO and WCST5 were not significantly different in terms of perseveration. However, after the exclusion of individuals who did not complete any categories, the results resembled those of the healthy group. Overall the results indicated that perseveration is not likely to be a result of an inability to inhibit well-learned responses. The second study was an attempt to confirm the findings of the first study whilst eliminating the possibility of a practice effect. Twenty individuals with closed head injury were administered either WCST5 or WCST15. Perseveration was not found to be significantly different between the two groups tested. It was suggested that the results were inconclusive owing to the exceptionally high perseveration scores of the participants. The study highlighted a serious methodological issue, namely that the effect of number of CCT cannot be accurately assessed on individuals with exceptionally high perseveration scores. The third study included sixteen individuals with traumatic brain injury and four individuals with Korsakoff's disease who were tested on either WCST5 or WCST15. Apart from the number of CCT, both versions differed from the original in that ambiguous cards were removed and participants were allowed to determine the sequence of sorting categories. The aim of this study was to confirm earlier data, investigate the effect of ambiguity and alter the task to ensure that at least one sorting category being completed. Data analysis revealed that, as expected, the group that was tested on the WCST15 produced significantly less erseverative errors when compared to the group that was tested on the WCST5. The final study aimed firstly at confirming preceding findings but more importantly at further investigating a possible link between perseveration and an inefficient use of cues. Damasio's idea of the somatic marker (1991) was adapted in an attempt to examine whether perseveration is linked with a decreased emotional reaction to feedback on the WCST. High perseverators (individuals with 20% and above perseveration) were expected to show a lower Skin Conductance Response (SCR) indicating a reduced emotional response to feedback received by the examiner. In addition SCR after negative feedback was expected to be higher than SCR after positive feedback owing to the corrective nature of the negative cue. Twenty-six participants without neurological history were administered either WCST5 or WCST15. As in the third study, the sequence of sorting categories was undetermined and ambiguous cards were removed. The study supported earlier findings showing that perseveration is significantly lower on WCST15 compared to WCST5. No significant differences in SCR were found between high and low perseverators or between negative and positive feedback. In addition, SCR was not proven to be significantly lower for negative feedback compared to positive feedback. Consequently, the idea that perseveration is associated with a decreased emotional reaction to feedback, was not supported. Overall, with the exception of the second study, the findings contradicted the prevalent idea that forming a strong "response set" would increase perseveration. In fact, they indicated that performance on the WCST is facilitated by a stronger response set, and perseveration decreases. This view is inconsistent with the way perseveration is viewed in the current literature but is consistent with early conceptualisations of perseveration (Grant and Berg, 1948, Gormezanno and Grant, 1958 or Pribram, 1961) that suggest that perseveration may be linked to ambiguity or unreliability of feedback. However, the findings of the fourth study did not provide evidence for a decreased emotional reaction to cues among perseverators during the WCST. Further investigation would be helpful in deciphering the relationship between perseveration and the use of external feedback.