AUTHOR=Chen Deng , Huang Yilu , Chen Sikun , Huang Yunzhe , Swain Andrew , Yu Jinming 

TITLE=Predictive Model Construction for Social–Emotional Competence of Toddlers in Shanghai, China: A Population-Based Study

JOURNAL=Frontiers in Public Health

VOLUME=Volume 9 - 2021

YEAR=2022

URL=https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2021.797632

DOI=10.3389/fpubh.2021.797632

ISSN=2296-2565

ABSTRACT=Objective To construct a simple model containing predictors derived from Chinese Learning Accomplishment Profile (C-LAP) to better the evaluation of the social-emotional development of toddlers aged 24-36 months. Method The test results by C-LAP system and demographic information of toddlers aged 24-36 months were collected between 2013 and 2019 in Shanghai, China, whose guardians were voluntary to accept the investigation. We developed a norm with the dataset based on the study population. With the norm, stepwise regression and best subset analysis were applied to select predictors. Results Relied on the norm established and stepwise regression as well as best subset analysis, an optimal model containing only 6 indicators was finally determined. In the training and validation dataset, the AUCs of the optimal model were 0.95 (95% CI: 0.94-0.96) and 0.88 (95% CI: 0.85-0.90), respectively. When cutoff point of the model was set at 0.04, its sensitivity in training and validation dataset was 0.969 and 0.949, respectively, and the specificity in training and validation dataset is 0.802 and 0.736, respectively. Conclusion We developed a simplified predictive model which includes only 6 items derived from C-LAP to evaluate the the probabilities of being at risk of developmental problem in social-emotional development for toddlers aged 24-36 months. Meanwhile, specificity and sensitivity of the model was tested to be high enough for future fast screening.