No abstract

Background: Children development is a multidimensional construct encompassing gross motor, fine motor, adaptive behavior, speech and language, and social behavior. The China Developmental Scale of Children (CDSC) has been tested using traditional methods. This study aimed to make amendment to previous evidence by using Rasch Model and elaborate the psychometric properties of the CDSC in a new perspective. Methods: 200 children and infants were recruited and finished the CDSC. Data were collected and used for the analysis. Rasch Model was adopted to test the psychometric properties of the CDSC regard the fitness to the prediction of the Model. Unidimensionality was tested using principal component analysis. Differential item functioning (DIF) analysis was performed across gender and age group (older or younger than 6.5 months old). Result: Data from 200 children with developmental delays were obtained. The result showed that the items from the CDSC presented acceptable fitness to the Rasch Model(Person Infit MNSQ:0.83~0.9;Item Infit MNSQ:0.9~0.93). Besides, 11 unfitting(misfitting or overfitting) items were found. The impact of the unfitting items on the average fit statistics was evaluate by reanalysis, and the result showed that the person reliability and separation were lower while the infit MNSQ got higher. The unidimensionality was supported by the result of the principal component analysis(Measures explained variance:79.6%~87.8%). Further, the reliability and the separation index of the item and person indicated that the assumption of the internal consistency was confirmed (Person.Reliability:0.91~0.96; Person.Separation:4.27~3.5; Item.Reliablity:0.99~1; Item.Separation:13.39~19.28). Conclusion: Our study found that the CDSC had shown reasonable fitness to the Rasch Model. The result support that the CDSC was established based on a unidimensional structure with good internal consistency. The misfitting items still required reformulation to demonstrate corresponding content more clearly.