Examining the validity of eye tracking during the computerized Wisconsin card sorting test in a sample of stroke patients and healthy controls

Abstract

ABSTRACT Introduction Each year, approximately 50,000 Canadians, one million Americans, and millions of people worldwide are hospitalized for stroke. Cognitive impairment is common after experiencing a stroke and is known to affect functioning on daily tasks. While neuropsychological assessments are often employed to assess cognitive abilities and make inferences about functional capabilities, there is growing interest in integrating contemporary technologies to augment assessment. Eye tracking allows previously overlooked information, such as overt visual attention based on fixations and saccades, to be quantified to help elucidate how responses are made during testing. Method The current study investigated the validity of eye tracking during completion of the computerized Wisconsin Card Sorting Test (cWCST), a common test of higher level cognition, in a sample of inpatients recovering from stroke (n= 41) and a control group of healthy individuals (n = 46). Results Results provided supporting evidence for the construction, criterion, and ecological validity of eye tracking on the cWCST with inpatients recovering from a stroke. Specifically, eye tracking metrics differentiated between inpatients and controls; fixations on cWCST areas of interest differed between type of response (conceptual versus non-conceptual); and average time per fixation predicted functional status early after a stroke as well as recovery during inpatient rehabilitation, above-and-beyond cWCST scores. Time spent on testing negated the effects of fixation and saccade counts for predicting cWCST performance, due to the substantial overlap in variance. Conclusion Current findings of this preliminary study provided support for the validity of eye tracking, integrated with the cWCST, for inpatients recovering from a stroke. Implications and areas for future research are discussed.