Arthur J. Reynolds and Judy A. Temple

Abstract

Estimates of program effects were investigated with quasi-experimental data from a compensatory preschool intervention using several contemporary techniques of bias reduction. These included econometric simultaneous modeling, latent-variable structural modeling, and ordinary least-squares regression. The techniques were applied to longitudinal data from the Chicago Child- Parent Center Preschool Program, a Head Start-type intervention. Analyses of 806 black children followed from kindergarten to Grade 6 indicated that the techniques, by and large, produced similar estimates of the effects of preschool intervention on school achievement test scores. Effect sizes ranged from .43 to .67 at school entry to .24 to .30 at Grade 6 (7 years postprogram). These findings were robust across different model specifications and were similar to those of randomized studies. The only differences among methods occurred at Grade 3. Findings suggest that different quasi-experimental estimation methods that attempt to control for selection bias can generate similar estimates of the effects of an educational treatment.

Evaluation Review, 19, 347-379, 1995.