Advanced measurement modeling of the Clock Drawing Test: A confirmatory factor analysis and item response theory-based validation of CDT-14

Abstract

Background

Early detection of cognitive impairment is vital for dementia management, especially in Alzheimer's disease. The Clock Drawing Test (CDT) is a widely used screening tool; however, many existing scoring methods lack comprehensive statistical validation and consistent factor structures.

Objective

We aimed to evaluate the validity and reliability of a novel scoring method, CDT-14, developed from items predictive of cognitive decline, and to compare its performance with three conventional scoring systems.

Methods

We enrolled 1100 outpatients undergoing cognitive assessments in a cross-sectional study. Scores from CDT-14 and the three conventional methods (Rouleau, Sunderland, Freedman) were compared. Structural validity was examined using confirmatory factor analysis (CFA), and item characteristics were analyzed via item response theory (IRT). Internal consistency was assessed with McDonald's omega. Construct validity and discriminative ability were evaluated through correlations and receiver operating characteristic (ROC) analyses. Participants were classified by Mini-Mental State Examination scores into Normal Cognition, Mild Cognitive Impairment, and Moderate-to-severe Cognitive Impairment.

Results

The CFA results supported a three-factor model (Circle, Numbering, Hands) with good fit. The IRT analysis indicated high measurement precision in the mild cognitive decline range (θ = −1 to 0). The CDT-14 score exhibited strong internal consistency (ω = 0.835). ROC analyses showed that the CDT-14 score had comparable or superior discriminative accuracy to conventional methods, with practical cut-offs and enhanced performance when adjusted for age and education.

Conclusions

The CDT-14 is a psychometrically robust scoring method providing standardized, sensitive assessment for cognitive screening and longitudinal monitoring in dementia clinical settings.