Psychometric properties of the social support scale for primary caregivers of infants and young children: Evidence from Rasch analysis

Objectives: This study aimed to examine the psychometric properties of a 12-item social support scale for primary caregivers of infants and toddlers using Rasch analysis, based on data from the Korean Early Childhood Education & Care Panel Study. Methods: A total of 1,000 participants were randomly selected from the first-wave Infant and Toddler Panel data (2022). Rasch analysis was conducted using the rating scale model. To improve category functioning, response categories 2 and 3 were combined, resulting in a 4-point scale. Unidimensionality and local independence were examined using Rasch residual principal component analysis. Item fit, item difficulty, differential item functioning (DIF) by caregiver gender and presence of social support, ceiling effect, and reliability indices (Cronbach’s α, person separation reliability, and separation index) were evaluated. Results: The results supported the unidimensionality and local independence of the scale. All items demonstrated acceptable fit to the Rasch model. Item difficulty analysis indicated that informational support items were the most difficult, whereas emotional support items were the easiest. No significant DIF was found across caregiver gender or presence of social support. The ceiling effect was 5.6%, indicating acceptable interpretability. Internal consistency was high (Cronbach’s α=.93), with a person separation reliability of 0.88 and a separation index of 2.72, suggesting that the scale effectively distinguishes levels of perceived social support among respondents. Conclusion: The 12-item social support scale demonstrates satisfactory reliability and validity for assessing perceived social support among primary caregivers of infants and toddlers. The findings support its use in research and practice related to health promotion and family support, particularly in early childhood contexts.