Psychometric evidence for assessing early childhood development in Brazil: the Child Development Diagnostic Form in the Criança Feliz Program
Ellayne Souza Cerqueira, Michael E Reichenheim, Ana Angélica Rodrigues Alves, Ricardo Rodrigues Dutra, Claudia Leite Moraes, Priscila Ribas de Farias Costa, Rita de Cássia Ribeiro-Silva, Mauricio L BarretoIntroduction
Child development is a multidimensional process influenced by socioeconomic conditions. The Programa Criança Feliz (PCF), Brazil’s largest home-visiting initiative, uses the Formulário de Observação Inicial do Desenvolvimento Infantil (FOIDI) to monitor developmental milestones. Despite being widely implemented, the FOIDI had not previously undergone formal validation. This study evaluated configural and metric validity of FOIDI in a representative sample of Brazilian children aged 0–36 months.
Methods
A total of 183 974 assessments from 82 797 children in PCF (2020–2023) were analysed. Cohorts for seven age groups were divided into three independent partitions. Initially, Horn’s parallel analysis determined the number of underlying factors. Confirmatory factor analyses followed, using weighted least-squares mean and variance adjusted estimator and polychoric correlation matrices. Model fit was evaluated using root mean square error of approximation, comparative fit index and Tucker–Lewis index. Items with factor loadings<0.40 were considered unreliable. Modification indices (MIs; univariate Lagrange multiplier) and expected parameter changes were used to identify potential cross-loadings and residual correlations. Convergent validity was estimated using average variance extracted (AVE≥0.50); discriminant validity was assessed by comparing the square root of AVE with factor correlations.
Results
Modules exhibited unidimensional (M3–M7) or bidimensional structures (M1–M2), with factor loadings ranging from 0.519 to 0.903. AVE was≥0.50 for all factors evaluated, supporting convergent validity. Discriminant validity was confirmed within the bidimensional modules. The removal of redundant items increased model parsimony without compromising informativeness.
Conclusion
The FOIDI demonstrated robust configural and metric validity, capturing core aspects of child development across different age groups in contexts of socioeconomic vulnerability. The refined instrument proved suitable for home-visiting programmes, offering a practical, reliable and scalable tool for developmental surveillance and the support of public policies.