Rasch Analysis of the DCDDaily-Q-CN: Assessing Activities of Daily Living in Children Aged 5–10 Years

Background The DCDDaily-Questionnaire (DCDDaily-Q) is the first standardized parent-report instrument specifically developed to assess activities in daily living (ADL) performance and participation in children with suspected Developmental Coordination Disorder (DCD). Cross-cultural adaptations, including the Chinese version (DCDDaily-Q-CN), have demonstrated strong psychometric properties under Classical Test Theory (CTT). However, CTT has inherent limitations in detecting item-level bias and modeling item-ability interactions. Rasch modeling, as a modern measurement approach, addresses these limitations by providing detailed insights into item targeting, measurement invariance, and category functioning. Methods Rasch analysis was conducted on data from 1,908 typically developing children (1,020 boys and 888 girls; mean age = 7.7 ± 1.7 years) recruited from 14 mainstream schools in a migrant city. The analysis evaluated key psychometric properties, including item and person reliability and separation, item-person targeting, item fit, category functioning, dimensionality, and Differential Item Functioning (DIF). Results Rasch analysis demonstrated excellent item reliability (≥ 0.98) and strong structural validity across all domains. However, person reliability was moderate (0.50–0.72), indicating limited precision in distinguishing individuals within subdomains. The scale showed good item fit (infit and outfit MNSQ values between 0.50 and 1.70), appropriate functioning of response categories, and supported unidimensionality across all domains. Item difficulties were generally well aligned with the ability levels of most children aged 5 to 10 years, confirming that the scale is appropriately targeted for its intended screening purpose. Nevertheless, Wright maps revealed mild ceiling effects and a degree of mismatch between item difficulty and participant ability, particularly in self-care tasks. DIF analysis identified gender-related bias in several self-care and fine motor items, whereas gross motor items exhibited measurement invariance across gender. Conclusion Rasch analysis confirms that the DCDDaily-Q-CN is a psychometrically robust tool with high reliability, unidimensional structure, and generally appropriate item targeting for assessing ADL in children aged 5 to 10 years. Nonetheless, the presence of ceiling effects and gender-related DIF suggests the need for further refinement to enhance measurement precision and fairness, particularly for higher-ability children and to ensure gender equivalence.