Linking Micro-features of Quality to Children’s Skills in a Statewide Population Study of Children’s Early Education

• In Event: Within Classroom Variation is not Noise: Predictors and Impacts of Varied ECE Experiences

Abstract

Commonly-used observational measures of quality tend to make sweeping global assessments, which provide little insight into the actionable processes that are most relevant for children’s learning. Understanding how micro-level features of quality relate to children’s outcomes can serve to inform quality improvement efforts by identifying concrete targets for improvement. In this study we examined the extent to which discrete observed teacher instructional characteristics – or micro-features of quality – were predictive of children’s skills across domains.

Data came from the first wave of a longitudinal population study of young children and early educators. The sample included a diverse range of early education and care settings, including community child care (44% of classrooms), family child care (19% of classrooms), Head Start (19% of classrooms), and public school prekindergarten (18% of classrooms). The present study included 1,773 three- or four-year old children and 1,413 educators working in 672 classrooms in 451 programs. Observers visited each classroom one time using the Teacher Observation in Preschools (TOP; Bilbrey, Vorhaus, & Farran, 2007). Direct assessments and assessor ratings from the same time period captured children’s academic, social-emotional, and executive functioning skills.

We predicted each child outcome as a function of the micro-features of quality, including a host of covariates. In addition to considering average classroom levels of micro-features of quality, we considered individual educator practices, namely the minimum or maximum TOP score for individual educators in classrooms with more than one educator (n = 541). We also examined whether differences between the maximum and minimum TOP score were related to child outcomes in this subset.

Table 1 presents descriptive statistics of micro-level practices across the sample. In addition to illustrating average trends in micro-level practices (e.g., teachers were engaged in whole group practices approximately 30% of the time), it also shows great variation in adult practices within the same classroom. For example, in classrooms with more than one teacher, the most talkative teacher tended to speak with children 69% of the time, whereas the least talkative teacher tended to speak with children less than half that amount of time (31%).

Table 2 presents unweighted results of multi-level regressions predicting child outcomes as a function of average classroom-level micro processes. The majority of micro-level processes were not associated with child outcomes. However, children’s social-emotional and executive function skills tended to be lower in classrooms where teachers disapproved of child behaviors more. Additionally, children’s academic skills tended to be higher in classrooms where teachers used a more positive tone to communicate with children. Finally, we note that children’s social-emotional functioning was higher in classrooms in which more time was focused on math.

Preliminary results highlight variation both across and within classrooms in educator practices. Results suggest that specific practices might be most relevant for child outcomes. The final presentation will include complete results and discuss the implications of the findings for quality improvement efforts.

Authors

• Emily Hanno, MDRC (Manpower Demonstration Research Corporation) - New York City
  Non-Presenting Author

• Katie Gonzalez, Mathematica
  Non-Presenting Author

• Kerry Guess Hofer, Abt Associates - Maryland
  Presenting Author

• Rosa Guzman Turco, Harvard University
  Non-Presenting Author

• Stephanie M Jones, Harvard Graduate School of Education
  Non-Presenting Author

• Nonie K Lesaux, Harvard Graduate School of Education
  Non-Presenting Author