Child maltreatment in China: comparison of parent and child reports

Background

Reporting of the incidence of child maltreatment by parents and children might differ with implications for optimum research methodologies to determine the incidence of maltreatment. Our aim was to compare parent and child reports of child maltreatment in mainland China.

Methods

A cross-sectional study was done in two primary schools and two secondary schools in urban and rural Zhejiang Province. Children aged 10–16 years and their parents completed a questionnaire survey. The same questions about child maltreatment appeared in both parent and child questionnaires and included 38 disciplinary acts (21 physical, 12 emotional, and five, non-contact). Parent–child pairs from the same household were matched to compare parent–child reports of maltreatment. We used McNemar's χ² test and Cohen's kappa coefficient for the statistical analysis. The study was approved by University College London and Zhejiang University Research Ethics Committees. All participants gave informed consent.

Findings

Questionnaires were completed by 611 parents and 821 children, with 324 mother–child pairs and 235 father–child pairs. For mother–child pairs, the lifetime prevalences of maltreatment (mothers vs their children) were 53·4% versus 36·7% for physical acts; 76·2% versus 50·0% for emotional acts; and 19·4% versus 13·0% for non-contact acts. For father–child pairs, the lifetime prevalences of maltreatment (fathers vs their children) were 57·9% versus 39·0% for physical acts; 71·5% versus 44·3% for emotional acts; and 22·6% versus 16·2% for non-contact acts. The prevalence of emotional maltreatment in the previous year was reported more by parents than children (55·9% mothers vs 32·7% children; 54·0% fathers vs 31·5% children), with no differences for physical maltreatment and non-contact punishment. The Cohen's kappa coefficients ranged from 0·09 to 0·39, indicating low agreement between parent–child reports.

Interpretation

High levels of child maltreatment are common in China. To gain accurate figures for maltreatment, both children and caregivers should be considered in research. Consistently lower figures in children might relate to recall bias or acceptance of acts of maltreatment as normal. Parents readily admit maltreating their children, possibly indicating its normalisation in China, indicating the need for parenting education.

Funding

China Scholarship Council and Universities' China Committee in London.