Using generalized additive modeling to empirically identify thresholds within the ITERS in relation to toddlers'...

Area of Science:

• Child Development
• Early Childhood Education
• Educational Measurement

Background:

• High-quality child care programs positively impact children's cognitive development.
• Quality Rating and Improvement Systems (QRIS) are increasingly used for child care benchmarking.
• There is a need for evidence-based thresholds in child care quality measures to inform policy.

Purpose of the Study:

• To explore Generalized Additive Modeling (GAM) for identifying thresholds in child care quality measures.
• To determine thresholds on the Infant/Toddler Environment Rating Scale (ITERS) related to toddler cognitive development.
• To provide guidance for policymakers on setting child care quality benchmarks.

Main Methods:

• Utilized the Early Childhood Longitudinal Study, Birth Cohort (ECLS-B) data set.
• Employed Generalized Additive Modeling (GAM) to analyze relationships between ITERS scores and Bayley Mental Development Scale Short Form-Research Edition (BMDSF-R) scores.
• Compared GAM-derived thresholds with simple linear models.

Main Results:

• GAM-derived thresholds were more effective than linear models in differentiating cognitive performance levels.
• A minimum ITERS threshold was identified, above which significant cognitive development improvements are expected.
• A potential ITERS ceiling effect was observed, with diminishing cognitive gains beyond a certain quality level.

Conclusions:

• GAM is a valuable method for identifying meaningful thresholds in child care quality measures.
• Establishing specific ITERS thresholds can inform policy decisions regarding child care quality standards.
• Understanding minimum and potential ceiling effects in child care quality is essential for optimizing early learning environments.