Pei-Ying Li1, Nan-Nan Fu1, Qing-Yang Li2, Geng-Fu Wang1, Lei Gao1, Xin Zhang3. 1. Department of Maternal, Child and Adolescent Health, School of Public Health, Tianjin Medical University, No. 22 Qixiangtai Road, Tianjin 300070, China. 2. Department of Education, School of Educational Science, Tianjin Normal University, No. 393 Binshuixi Road, Tianjin 300387, China. 3. Department of Maternal, Child and Adolescent Health, School of Public Health, Tianjin Medical University, No. 22 Qixiangtai Road, Tianjin 300070, China. Electronic address: zhangxin@tmu.edu.cn.
Abstract
BACKGROUND: The reliability and validity of the Griffiths Development Scales-Chinese (GDS-C) for autistic children in China are unknown. Thus, it is urgent to verify the instrument's reliability and validity in this population. The aim of the study was to explore whether the GDS-C is reliable and valid for assessing neurodevelopment in autistic children. METHOD: This study included 296 autistic children and 141 typically developing children from 3 to 8 years of age in China. The reliability of the scale was estimated based on its internal consistency, test-retest reliability and interrater reliability. The validity of the scale was calculated based on the construct validity, discriminate validity and criterion validity. Receiver operating characteristic curves were used to calculate the general quotients (GQs) corresponding to the diagnostic classification within the Childhood Autism Rating Scale (CARS) scores. RESULTS: This study shows sufficient reliability (Cronbach's alpha = 0.957; test-retest reliability = 0.945 for the whole scale and 0.830-0.919 for the subscales; interrater reliability = 0.925 for the whole scale and 0.796-0.919 for the subscales). The results also provide good support for the validity of the GDS-C. In the discriminant analysis, 85.5% of the children in the autistic sample were correctly classified. The cutoff value for distinguishing autistic children from normal children within the CARS scale corresponds to a GQ of 84.83, and that for distinguishing severely autistic children from mild or moderately autistic children corresponds to a GQ of 66.60. CONCLUSION: Our findings suggest that the GDS-C may be a valid and reliable tool for assessing the neurodevelopment of autistic children.
BACKGROUND: The reliability and validity of the Griffiths Development Scales-Chinese (GDS-C) for autisticchildren in China are unknown. Thus, it is urgent to verify the instrument's reliability and validity in this population. The aim of the study was to explore whether the GDS-C is reliable and valid for assessing neurodevelopment in autisticchildren. METHOD: This study included 296 autisticchildren and 141 typically developing children from 3 to 8 years of age in China. The reliability of the scale was estimated based on its internal consistency, test-retest reliability and interrater reliability. The validity of the scale was calculated based on the construct validity, discriminate validity and criterion validity. Receiver operating characteristic curves were used to calculate the general quotients (GQs) corresponding to the diagnostic classification within the Childhood Autism Rating Scale (CARS) scores. RESULTS: This study shows sufficient reliability (Cronbach's alpha = 0.957; test-retest reliability = 0.945 for the whole scale and 0.830-0.919 for the subscales; interrater reliability = 0.925 for the whole scale and 0.796-0.919 for the subscales). The results also provide good support for the validity of the GDS-C. In the discriminant analysis, 85.5% of the children in the autistic sample were correctly classified. The cutoff value for distinguishing autisticchildren from normal children within the CARS scale corresponds to a GQ of 84.83, and that for distinguishing severely autisticchildren from mild or moderately autisticchildren corresponds to a GQ of 66.60. CONCLUSION: Our findings suggest that the GDS-C may be a valid and reliable tool for assessing the neurodevelopment of autisticchildren.