Evaluating decision-making: Validation and regression-based normative data of the judgment assessment tool

Abstract

Objective. This study presents the results of the development and validation of the Judgment Assessment Tool (JAT). The JAT measures two core aspects of judgment, namely generation of solutions (G) and assessment of options (A), the two first stages of decision-making process. Method. During the test development phase (study 1), a preliminary version of the JAT was evaluated by 14 experts and tested on 30 healthy controls (HC). One hundred and twenty HC (20 - 84 years old) and 24 participants with mild Alzheimer's disease (AD) were subsequently tested on the final version of the JAT (study 2). HC participants aged 60 and over and AD participants underwent a neuropsychological evaluation. Results. The internal consistency of the final version of the JAT assessed by Cronbach's a was 0.71 for the HC group and 0.85 for the AD group. Performance on the JAT was normally distributed both in the HC and AD groups. The test correlated with abstract reasoning, verbal fluency, and working memory. Results revealed adequate test-retest reliability and excellent interrater reliability (k coefficient was 0.92 for the G section and 0.93 for the A section). Demographically adjusted normative data were generated based on a regression analysis and results showed that AD participants performed worse than HC with a large effect size (Cohen's d ¼ 1.79). Conclusion. Overall, these results provide evidence of the reliability and strong construct validity of the JAT to evaluate judgment.