Neuroscience Letters 409 (2006) 234–238 Shared cognitive processes underlying performance on the Wisconsin Card Sorting Test and the Stroop Test in patients with schizophrenia: A measurement artifact? Mary H. Kosmidis a,∗ , Vasilis P. Bozikas b,c , Maria Zafiri a , Athanasios Karavatos b a Neuropsychology Group, Department of Psychology, Aristotle University of Thessaloniki, Greece b 1st Department of Psychiatry, Aristotle University of Thessaloniki, Greece c 2nd Department of Psychiatry, Aristotle University of Thessaloniki, Greece Received 27 February 2006; received in revised form 8 September 2006; accepted 21 September 2006 Abstract We explored the hypothesis that, while sensitive to different aspects of executive functioning in patients with schizophrenia, the Wisconsin Card Sorting Test (WCST) and the Stroop Test also measure the same construct, namely, inhibitory control. Specifically, our goal was to confirm and extend previous findings [A. Rossi, E. Daneluzzo, P. Mattei, M. Bustini, M. Cassachia, P. Stratta, Wisconsin Card Sorting Test and Stroop performance in schizophrenia: a shared construct, Neurosci. Lett. 226 (1997) 87–90] by demonstrating the independence of this construct from other abilities necessary to successfully perform the tasks. More importantly, we sought to improve on this previous study by eliminating the influence of the variance of speed of responding. We examined 55 patients with schizophrenia and initially found that performance on the Stroop Color-Word condition could, indeed, be predicted only by the percentage of perseverative errors on the WCST, and not variables reflective of other cognitive skills, thus replicating and extending previous findings. Once we removed the influence of speed of responding from our measure, however, thus isolating the inhibitory process, this finding disappeared. Therefore, our findings highlight the importance of isolating the individual components of interest from complex measures before drawing conclusions regarding the cognitive processes underlying particular test performance. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Wisconsin Card Sorting Test; Stroop; Executive functioning; Inhibition; Schizophrenia Impairment in executive functioning is frequently reported in patients with schizophrenia [27] and has received considerable research attention [10]. The term executive functioning, how- ever, is imbued with a multitude of definitions [27]. As such, it is used to refer to a variety of cognitive skills [12], includ- ing cognitive flexibility, establishing, maintaining and shifting cognitive sets as a response to changing environmental demands, hypothesis formation and testing, self-monitoring, goal-directed behavior, and the ability to suppress an automatic reaction in favor of a more adaptive one [10,27,32]. Consequently, tests that are considered to be sensitive to deficits in executive func- tioning may, in fact, be measuring different aspects of this broad skill. ∗ Corresponding author at: Department of Psychology, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece. Tel.: +30 2310 997308; fax: +30 2310 997384. E-mail address: kosmidis@psy.auth.gr (M.H. Kosmidis). Two widely used tests of executive functioning in schizophre- nia are the Wisconsin Card Sorting Test (WCST) [11,13] and the Stroop Color-Word Test [31]. While often used to measure executive functioning in a broad sense, in fact, both of the afore- mentioned tests are also sensitive to different aspects of this skill. Even within each test, however, different processes may be necessary for successful performance. Various cognitive processes have been implicated in WCST performance. Traditionally, this test has been used to evaluate cognitive flexibility [34] and is considered to require the contri- bution of processes such as conceptualizing, creating and testing hypotheses, as well as shifting cognitive set [15]. Since the first publication that investigated the performance of patients with schizophrenia on the WCST [9], a plethora of subsequent studies have confirmed that patients with schizophrenia perform worse on the task than healthy controls overall [8,15,16,18,22,29]. In a factor analytical study [33] of the conventionally used WCST variables [13], the investigators sought to reduce poten- tial redundancy in the variables, while assessing three groups 0304-3940/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2006.09.049