Head Start is the largest publically-funded preschool program in the United States, serving more than 900,000 children each year (U.S. D.H.H.S, 2010). Yet findings on the impacts of Head Start are mixed. Evidence from quasi-experimental studies suggest that Head Start programs are associated with numerous positive outcomes, including improved cognitive and social skills, higher graduation rates, and lower crime rates (for a review see Shager et al., 2013). However, recent findings from the Head Start Impact Study, a randomized control trial of over 4,000 children, suggest much smaller effects, with most impacts of Head Start fading by the time children enter into elementary school.

One explanation for the lack of persistent experimental findings may be due to the variation in classroom quality. A recent report on Head Start suggests that many children may not attend programs that provide adequate levels of instructional support (Moiduddin et al., 2012). This is of concern because classrooms that provide a cognitively stimulating environment have the potential to compensate for achievement gaps in early childhood. For example, a study of first grade children found that those with low maternal education who attended classrooms with high instructional support achieved at levels commensurate to their low-risk peers (Hamre & Pianta, 2005).

This study examines the effects of quality on child outcomes among Head Start children. One challenge in estimating this relationship is that children’s achievement may be related to a host of other factors, including genetic predispositions shared with parents, stimulating home environments, and parent motivation, all of which could be linked with parents’ selection of preschool programs. In this study, we use a family-fixed effects approach to control for family background effects and genetic similarities, and we contrast the outcomes of siblings who experience different levels of quality in Head Start programs.

We employ data from the Head Start Family and Child Experiences Survey (FACES), which collects data every three years among a nationally representative sample of Head Start programs, classrooms, and children. In 2006 and continuing in 2009, FACES included direct observations of instructional support in the classroom, measured by the CLASS^TM (Pianta et al., 2008). The inclusion of the CLASS^TM, in addition to structural dimensions (e.g. teacher education, staff-child ratio), provides a distinct opportunity to examine the effects of quality using highly-sophisticated measures. The FACES study also includes a rich assessment of children’s development, including academic, social, and language skills at the start of preschool through kindergarten.

A unique design feature of the FACES study is that it collects data on siblings if they are both randomly selected for the study. In FACES 2006, 102 sibling-pairs were selected for the study. Seventy-four of these pairs attended different classrooms, providing the opportunity for family-fixed effects. To increase our sample size, we plan to include children in the FACES 2009, which will become available to the public May 2013. Findings will be discussed in terms of the effect of classroom quality on children’s development and may elucidate a target for strengthening the impact of Head Start.