A psychometric analysis of a short form of the Chinese version of the ISPCAN child abuse screening tools – Children's home version (SC-ICAST-CH) using multidimensional item response theory

Background: The development of measures of child maltreatment for the Chinese population were limited until the Chinese version of the ISPCAN Child Abuse Screening Tools – Children's Home version (ICAST-CH-C) was proposed. Although the ICAST-CH-C was found to be effective in assessing the scope and prevalence of child maltreatment, it has several potential drawbacks. The time that is required to complete the ICAST-CH-C scale is longer than usual for a 36-item scale, because many of its items have one or more follow-up questions. Moreover, each item requires child victims to recall unpleasant experiences. Both phenomena can cause increases in invalid responses and in turn damage the data quality. Objective: The goal of this study was to propose a short form of the ICAST-CH-C (called the SC-ICAST-CH) to reduce the test length and response time in order to improve the measurement quality. Participants and setting: A dataset from a national survey of 5236 adolescents in Taiwan was used. Methods: A multidimensional version of the rating scale model (MRSM) was fitted to the data. The model parameters were estimated with the ConQuest software. Results: The results indicated the reliability of the SC-ICAST-CH was fairly good, with only 61 % of the original test length. Disordered thresholds were found in all five subscales; the underlying reasons for this phenomenon need further investigation. Specific cultural differences related to item retention/removal decisions were also discussed. Conclusion: The efficient, shorter SC-ICAST-CH was shown to be a valid and reliable instrument for assessing the prevalence of child maltreatment.