•  
  •  
 

CONSTRUCT VALIDITY OF A NEW MEDICAL APP AND TRADITIONAL TESTING

Abstract

Ryley Skinner1, Victoria Carter1, Jacob Armstrong1, Robert J. Britton1, David Wibowo1, Heidi A. VanRavenhorst-Bell1, C. Brendan Clark1

1Wichita State University, Wichita, Kansas

PURPOSE: This presentation shows the preliminary results of a study comparing participants’ performance of a new medical app, that was designed to test cognitive performance, with performance on a battery of traditional cognitive tests. This was done to demonstrate the construct validity of this new medical app. METHODS: This study’s data collection is still in progress. The current N for this study is 25, though it has an expected N of 100. The medical app was administered on a smartphone. Cognitive assessments were administered in-person by licensed master’s level clinicians. Study participants were administered all assessments in the same order, with the app assessments (i.e., Impulse Control Assessment and Delayed Recall/Working Memory Assessment) being followed by the traditional cognitive tests (i.e., the Wechsler Adult Memory Scale-IV, the Wechsler Adult Intelligence Scale-IV, the Delis–Kaplan Executive Function System, and the California Verbal Learning Test – 3rd Edition). Correlations were conducted based off of overlap in the constructs being assessed. RESULTS: Associations between the medical app and the traditional cognitive tests were found by using Pearson’s r correlational analyses. Findings showed the app's Impulse Control Assessment demonstrated significant correlations with D-KEFS, Trail Making Condition 1, r = -.41, p = .040. In addition, it was also found that app’s Impulse Control Assessment had correlations, r = -.34, p = .098, with WAIS-IV Cancellation and, r = -.25, p = .235, with D-KEFS, Trail Making Condition 3. Memory constructs of the app’s Delayed Recall and Working Memory Assessment had no significant correlation with traditional testing scales (p > 0.05, respectively); however, a clinically relevant with, r = -.31, p = .131 was demonstrated against WMS-IV Symbol Span. CONCLUSION: Though our analyses were largely underpowered, the results demonstrated significant correlations between constructs as measured by the medical app and the traditional cognitive assessments. Clinically relevant associations were also found within the data, which may suggest significance with a sufficient sample size. In conclusion, the new medical app demonstrated correlations with gold-standard cognitive assessments that support further research into the apps’ validity and reliability.

ACKNOWLEDGEMENTS: This study was funded by Sway Medical.

This document is currently not available here.

Share

COinS