Evaluating the Stroop Test With Older Adults: Construct Validity, Short Term Test-Retest Reliability, and Sensitivity to Mental Fatigue

Abstract

We conducted two studies to evaluate the construct validity, short term test-retest reliability, and sensitivity to mental fatigue of the Stroop task when used with older adults. In Study 1, 40 participants visited our lab on two separate days. On the first visit, they took five screening scales, and we measured their height and body mass. On the second visit, they completed the Stroop task twice with a 30-minute interval between assessments. In Study 2, 15 different participants took a 30-minute Flanker/Reverse Flanker task during the interval between the two administrations of the Stroop tasks and they gave subjective ratings of their mental fatigue on the Visual Analogue Scale (VAS) prior to taking either of the Stroop tasks. In Study 1, participants showed a ceiling effect on the Stroop accuracy measure, there was strong concurrent validity for the Stroop with significant score differences between the Stroop’s congruent and incongruent conditions ( p < .001), and there was excellent response time reliability (ICC = 0.926) on day two when participants took the Stroop twice within a 30-minute inter-test interval. However, there were significant test-retest performance differences with respect to cognitive inhibition ( p < .001). In Study 2, mental fatigue from the Flanker/Reverse Flanker test resulted in a significantly worse second Stroop performance ( p = .045). We concluded that the Stroop task demonstrated strong concurrent validity and response time reliability among older adults, but it showed sensitivity to mental fatigue, and repeated administrations within the short 30-minute test-retest interval revealed that the most important Stroop measure (cognitive inhibition) was unreliable. We discuss the implications of these findings.