Immersive trail making: Construct validity of an ecological neuropsychological test

Traditional neuropsychological tests of executive function are designed to assess a range of behavioral competencies necessary for the cognitive control of behavior in the real world. Though considered the 'gold standard', these pen-and-paper tests have been criticized for questionable relevance to everyday functioning. Immersive virtual reality (VR) technologies may be a viable tool for developing more ecologically valid human performance testing and training environments. The present study provides initial data on the construct validity of a full-body 360-degree VR version of the classic Trail Making Test. Sixteen healthy volunteers completed both pen-and-paper and VR versions of the Color Trails Test (CTT). During Part A of the test, the participant connects circles containing numbers in sequential order, and in Part B, which loads more on executive function, the participant does the same but alternates between two colors. Among the eleven participants who have completed testing to date, for both pen-and-paper and VR versions, Task B completion time was longer than Task A completion time (Pen-and-paper CTT: Trails A 36.6±15.9 seconds, Trails B 64.5±17.9 seconds; VR CTT: Trails A: 137.4±52.0 seconds; Trails B: 219.2±74.4 seconds). Trails B-A completion time delta was 27.8±8.2 seconds for the pen-and-paper CTT, and 81.8±36.6 seconds for the VR CTT. The correlation (Spearman's rho; r_s) between Part A completion time on the pen-and-paper CTT and the corresponding Part A completion time on the immersive VR CTT was 0.76 (p=.007). The correlation for Part B completion time was 0.88 (p<.001). For Trails B-A delta, the correlation was 0.59 (p=.056). These findings indicate the convergent construct validity of the novel VR CTT, suggesting that it indeed measures the same cognitive construct (selective attention) as the gold standard pen-and-paper test. The longer completion times may reflect contributions of additional cognitive and motor processes. These results pave the way for larger validation studies, explorations of discriminant validity, and even more ecologically valid applications of neuropsychological test designs as implemented in the VR context.