Do Children in Their First and Second Year of Head Start Make Different Cognitive Gains Across a Year of Head Start?

• In Event: Sequencing of Early Learning Experiences: Dosage, Instructional Alignment, and Children’s Developmental Outcomes

Abstract

Children who attended Head Start for more than one year had stronger cognitive skills at the end of pre-K or when entering kindergarten than children who attended for one year or less, on average (Lee 2011; Wen et al. 2013; Xue et al. 2016). However, children’s growth in their second year of Head Start has been found to be lower than in their first year (Klein et al. 2018). Using nationally representative data from the Head Start Family and Child Experiences Survey 2014, we explore potential characteristics of children, families, and classrooms that could explain why children who complete a second year of Head Start (second-year children) experience fewer cognitive gains than children who complete their first year (first-year children).

We first examine differences between first- and second-year children who attended Head Start between fall 2014 and spring 2015 on their: characteristics, family background, and classroom experiences (part-day enrollment, classroom quality and teaching practices, and teacher experience and education). We then conduct multilevel regressions with children nested within classrooms, centers, and programs to predict cognitive gains based on whether children were in their second year of Head Start. We predict gain scores for five cognitive outcomes measured using standardized assessments: English receptive vocabulary, English expressive vocabulary, letter-word knowledge, early writing, and early math. We successively add sets of variables representing child, family, and classroom characteristics in five models, with each model including the variables from the prior models, to determine whether controlling for these characteristics explains why second year children make fewer gains than first year children. Model 1 is an uncontrolled model that only includes the time between the fall and spring assessments, model 2 includes child’s age in months at the spring assessment, model 3 includes a broad set of other child and family background characteristics, model 4 includes fall scores on the cognitive outcome being predicted, and model 5 includes teacher experience because it was the only classroom variable that differed in a way that could explain the fewer gains for second-year children.

First- and second-year children had similar family background characteristics, enrollment in part-day Head Start, and classroom characteristics, but second-year children were older, less likely to be White, and had higher fall cognitive skills. When we examined uncontrolled differences between first- and second-year children’s gains (model 1), we found that second-year children showed fewer gains in four of the five cognitive outcomes (Table 1). After controlling for age (model 2), the difference in gains between first- and second-year children were substantially reduced and no longer statistically significant for early writing and early math. However, after controlling for all variables, second-year children still showed fewer gains in expressive vocabulary and letter-word knowledge (at trend-level) than first-year children.

Finding that age is negatively associated with cognitive children’s growth is consistent with prior research (Wen et al. 2012). We will discuss implications for early childhood education such as providing higher-level content to older children and second-year children to increase their cognitive gains.

Authors

• Jessica Harding, Mathematica Policy Research - New Jersey
  Presenting Author

• Philip Gleason, Mathematica
  Non-Presenting Author

• Nikki Aikens, Mathematica Policy Research - District of Columbia
  Non-Presenting Author

• Lizabeth Malone, Mathematica Policy Research - District of Columbia
  Non-Presenting Author

• Nina M Philipsen, Office of Planning Research and Evaluation
  Non-Presenting Author

• Louisa Banks Tarullo, Mathematica - Massachusetts
  Non-Presenting Author