Mathematica explored the associations between quality and outcomes, examining whether certain thresholds of quality or dosage need to be met, or particular aspects of quality need to be present, before linkages appear. We also assessed interrelationships between these factors, as well as relevance for different age groups and types of care settings.

The Child Care and Early Education Quality Features, Thresholds, Dosage, and Child Outcomes Study Design project provided a basis for federal agencies and other stakeholders to field new research on the quality of early care and education, support quality improvement initiatives and practice, and inform decision making at the state and national levels. The project focused on center-based settings serving children from birth through age 5, focusing on children from low-income families, including those with risk factors affecting their potential school readiness. The research questions included:

• What aspects of quality in center-based early care and education programs serving children from birth through age 5 make a difference in child outcomes?
• What are the mechanisms by which different quality features support children’s outcomes?
• What specific features of quality lead to gains in child outcomes?
• Which measures best capture the components that make a difference?
• What thresholds of global quality or specific quality features are needed to support gains in child outcomes?
• Is there a minimum threshold of quality necessary to affect child outcomes? How well are measures of quality capturing these thresholds?
• What levels of exposure to quality (or dosage) are needed to support gains in child outcomes? How well are measures of quality capturing these levels of exposure?
• How do quality features, thresholds, and dosage relate independently to child outcomes? In relation to one another?
• How does the age of the child or the type of setting (for example, community-based or program-based) affect these relationships?

The project team conducted a literature review, analyzed secondary data, and conceptualized a logic model to prepare for the design phase. In a related task, the team identified potential sites for conducting the research and described their policy context, community characteristics, families and children served, and other factors that supported participation. Subcontractors included Child Trends, the University of North Carolina-Chapel Hill, and the University of Virginia.