Population surveys on child wellbeing require a brief, validated tool to measure child and adolescent maltreatment. The 7-item Short Child Maltreatment Questionnaire (SCMQ), developed by a WHO expert committee, has not been psychometrically tested. This study aimed to determine the factor structure, measurement invariance and correlates of a modified version of the SCMQ (6 of its 7 items) in a sample of adolescents attending schools in England. Data from 15,155 adolescents (aged 11-18 years) in the OxWell School Survey were used to test a modified 6-item version of the SCMQ. Information on depression, anxiety, self-harm, smoking, alcohol use, substance use, bullying victimization, peer victimization/dating abuse was collected using a structured, self-report questionnaire. The factor structure, reliability, and construct validity of the SCMQ was examined, and measurement invariance was tested across gender, age, and ethnicity. Findings from exploratory factor analyses revealed a stable one-factor model, while two- and three-factor models showed instability, suggesting that a unidimensional solution best represented the data. CFA confirmed a one-factor model χ2(df) = 92.265(12); CFI 0.992 TLI 0.987 RMSEA 0.02 90 % CI [0.020, 0.029). Around 16.3 % of the sample experienced any form of maltreatment, with exposure rates for specific types of maltreatment ranging from 0.6 to 11.6 %. Full configural, metric, scalar invariance was observed across all groups. Internal consistency (McDonald's omega) was adequate (0.71). The SCMQ demonstrated good construct validity, showing significant associations with health behaviours (substance use, alcohol use, smoking), mental health challenges (self-harm, depression, anxiety) and interpersonal violence (dating abuse, bullying) with effect sizes ranging from small to moderate. The SCMQ scores demonstrated acceptable reliability and validity among adolescents in England. It can be used to screen for maltreatment in large surveys and across diverse child populations. Opportunities to validate the SCMQ in other contexts would be beneficial.
Neelakantan et al. (Sat,) studied this question.