Test-retest reliability and practice effects of the virtual environment grocery store (VEGS)

The use of virtual reality (VR) technology has been suggested as a method to increase ecological validity in neuropsychological assessments. Although validity has been a focus in VR research, little attention has been paid to other psychometric properties such as test-retest reliability and practice effects. Practice effects are common on traditional neuropsychological tests and can be impacted by novelty. Because VR is not widely used it was expected that participants would demonstrate higher practice effects on VR as compared to paper-and-pencil testing.

To compare test-retest reliability and practice effects in VR and traditional paper-and-pencil testing, the Virtual Environment Grocery Store (VEGS) and California Verbal Learning Test - Second Edition (CVLT-II) were used in healthy adults (n = 44). Participants received follow-up testing approximately 2 weeks after the initial visit.

Significant practice effects of similar magnitude were seen on memory scores (i.e., total learning, long-delay free recall, and long-delay cued recall) on the VEGS and the CVLT-II. The VEGS and CVLT-II memory scores also demonstrated strong test-retest reliability (r's > .71). Lastly, total learning scores (d = .32) and long-delay cued recall (d = .70) scores were significantly higher on the CVLT-II compared to the VEGS (p's < .01).

Results suggested similar test-retest reliability and practice effects of the VEGS and CVLT-II, although the VEGS has the benefit of being an immersive technology that simulates an everyday activity. The study replicated past findings that the VEGS is more difficult than the CVLT-II which may be a useful property for clinical assessment.