Child Abuse Scale And Child Neglect Scale: Development, Reliability And Validity

PART â…  Child Abuse Scale: Development, Reliability and ValidityObjective: To develop a new scale as an operable instrument for assessing the abused status of child. To examine the reliability and validity of the Child Abuse Scale (CAS).Methods: Considering the cultural background of China, an item pool was established through revising items in correlative literatures and scales. Then, the first draft of the CAS was developed by reserving effective items which was graded by interrelated experts. The sample consisted of 902 students of the junior high school and the technology and profession college. Exploratory and confirmatory factor analyses were used to develop and evaluate the structure of the scale. The validity analyses included content validity, construct validity, and criterion-related validity. The reliability analyses included Cronbach's Alpha coefficients, split-half reliability, and test-retest reliability.Results: The CAS consisted of 35 items. The CAS was made of three sections which were physical abuse scale, verbal abuse scale and sexual abuse scale. The results of reliability were as follows: the cronbach's Alpha reliability of the whole child abuse scale was 0.870, the split-half reliability was 0.767, and the test-retest reliability was 0.876. The cronbach's Alpha reliability of the physical abuse scale was 0.862, the split-half reliability was 0.853, and the test-retest reliability was 0.881. The cronbach's Alpha reliability of the verbal abuse scale was 0.748, the split-half reliability was 0.696, and the test-retest reliability was 0.811. The cronbach's Alpha reliability of the sexual abuse scale was 0.725, the split-half reliability was 0.628, and the test-retest reliability was 0.861. The results of validity were as follows: the item loadings of the physical abuse scale ranged from 0.505 to 0.739, and the item loadings of the verbal abuse scale ranged from 0.402 to 0.715, and the item loadings of the sexual scale ranged from 0.332 to 0.811.The results of confirmatory factor analysis showed that the ratio of Chi-square to degrees of freedom was 3.814, the goodness of fit index was 0.904, the Tucker-Lewis index was 0.887, and the root mean square error of approximation was 0.041.