Association for Public Policy Analysis & Management

Previous literature documents the pattern of treatment effect fadeout that occurs in the years following the end of early childhood interventions. The current study investigated whether certain instructional environments in kindergarten and first grade deterred treatment effect fadeout following two separate preschool interventions, Head Start and Building Blocks.  We hypothesized that the instructional environment in kindergarten and first grade could moderate the long-run impacts of preschool interventions through two separate mechanisms.  First, we hypothesized that if treatment students entered kindergarten and first-grade classrooms that emphasized advanced content taught through high-quality instruction, then these students should show less effect fadeout.  Second, we hypothesized that if instruction in kindergarten and first grade classes was specifically tailored to build upon the knowledge students learned in preschool, then this should also moderate treatment impact fadeout. 

Data were drawn from the Head Start Impact Study (HSIS) and Building Blocks (BB) scale-up evaluation.  Head Start is a comprehensive child development program that provides children with preschool education, health screenings and examinations, and nutritious meals, in a full-day, center-based setting.  BB is a preschool mathematics curriculum that was evaluated through a scale-up study that randomly assigned 42 low-income schools to one of three conditions: 1) preschool curriculum; 2) preschool curriculum with follow-through; 3) control.  The schools assigned to the follow-through condition received additional pedagogical training for kindergarten and first grade teachers designed to help teachers build upon the knowledge students learned in preschool. 

Analyses were conducted by modeling academic achievement in kindergarten and first grade as a function of treatment status, instructional environment (i.e. quality or whether advanced content emphasized) in kindergarten and first grade, and student and school background characteristics.  We interacted measures of kindergarten and first-grade instructional environment with treatment status to identify the effect of instruction on preschool effect persistence.  In order to assess the impact of tailored PD in kindergarten and first grade, we estimated the impact of assignment to the follow-through condition in the BB study on kindergarten and first-grade outcomes.

Results from models with interactions between instruction and treatment status in both the HSIS and BB studies were universally null.  Thus, our measures of instructional environment in kindergarten and first grade had no apparent effect on preschool treatment impact persistence.  However, we did find that students assigned to the follow-through condition in BB showed approximately 50% effect fadeout in kindergarten, but showed almost no additional effect fadeout in first grade.

Our results indicate that instructional environment in kindergarten and first grade can moderate treatment impact fadeout, but the instruction needs to be specifically tailored to the content taught during the preschool intervention in order to create a more sequenced progression of academic content from preschool through first grade.