The reliability and validity of Bayley-III cognitive scale in China's male and female children

Background: Since publication in 2006, the Bayley-III scale has been used widely in pediatric populations worldwide; however, there have been very few studies which examined the usefulness and the potential sex differences in a Chinese context.\ud \ud Aims: To assess the reliability and validity of the Bayley-III cognitive scale, and detect possible sex differences in term children so as to provide evidence for clinical and research use in China.\ud \ud Study design: Cross-sectional study.\ud \ud Participants and outcome measures: Of the 1589 children from 3 healthcare institutions that were initially recruited, a total of 1444 children were included in the final analysis. We randomly selected 5-10% children from the total sample to evaluate the test–retest, inter-rater and criteria-related reliability in order to meet the psychometric criteria of Bayley-III scale. Inter-item consistency, test-retest and inter-rater reliability of the scale were estimated using Split-half method and Intra-class Correlation Coefficient (ICC). The content validity was evaluated by the Item-level Content Validity Index (I-CVI). The Mann-Kendall trend test was performed to assess trends of cognitive development, and post-hos Least Significant Difference test was used to detect age-appropriateness of items.\ud \ud Results: Six developmental pediatricians were trained to administer the Bayley-III cognitive scale. Inter-item consistency (n=1444) with Guttman split-half coefficient was above 0.8, while test-retest (n=144) and inter-rater reliability (n=74) had good to excellent ICCs of over 0.9. The criteria-related validity (n=74) of Bayley-III was acceptable, and associations with Gesell Developmental Schedules (GDS) were mainly above 0.8. The raw score of Bayley-III scale in total subjects (n=1444) showed an increased trend across all months of age (p