A Mokken Scale Analysis of the Bayley-4 Cognitive Scale in Greece

Abstract: The fourth edition of the Bayley Scales of Infant and Toddler Development (Bayley-4) is a standardized tool for evaluating early developmental progress. This study examined the psychometric properties of the Cognitive scale (30 items) in the Greek context using a stratified sample of 500 typically developing children aged 16 days to 42 months. Given the scale’s developmental nature, ordinal polytomous scoring, and hierarchical structure of the items, the Mokken Scale Analysis was employed. Results confirmed a strong unidimensional structure after excluding one nonscalable item ( H = .92; Hi = .78–.99). The monotonicity assumption was met, while local dependency was identified between two items. Invariant item ordering was generally supported, with a violation in one item and minor deviations in the ordering of certain adjacent items. The 28-item revised scale demonstrated excellent internal consistency (α = .97). No significant gender differences were observed. Further evaluation is warranted to enhance the diagnostic accuracy of the adapted scale, including item refinement, reordering, and assessing the usefulness of the polytomous scoring introduced in Bayley-4, as the mid-response category had low frequency across most items. Practical implications include using accumulated scores to monitor developmental progress, identify areas of cognitive delay, and guide intervention planning.