NotesFAQContact Us
Search Tips
ERIC Number: ED512707
Record Type: Non-Journal
Publication Date: 2010
Pages: 11
Abstractor: ERIC
Reference Count: 16
Using Meta-Analysis to Explain Variation in Head Start Research Results: The Role of Research Design
Shager, Hilary M.; Schindler, Holly S.; Hart, Cassandra M.D.; Duncan, Greg J.; Magnuson, Katherine A.; Yoshikawa, Hirokazu
Society for Research on Educational Effectiveness
Head Start was designed as a holistic intervention to improve economically disadvantaged, preschool-aged children's cognitive and social development by providing a comprehensive set of educational, health, nutritional, and social services, as well as opportunities for parent involvement (Zigler & Valentine, 1979). Given the current interest in ECE (early childhood education) as an intervention strategy for disadvantaged children and the magnitude of public investment in Head Start ($6.9 billion in FY 2007), it is important for researchers and policy makers to be effective designers and consumers of Head Start evaluations (Office of Head Start, 2008). Although some previous meta-analyses suggest a link between evaluation characteristics and results, evidence is mixed, and recent methodological advances have not been considered. A more detailed empirical test of the contribution of particular research design characteristics is needed to enable scholars to better understand findings from prior studies, as well as to inform future studies. This study investigates the role of such factors in explaining variation in Head Start evaluation results for children's cognitive and achievement outcomes. Specifically, the authors test whether the following research design characteristics explain heterogeneity in the estimated effects of Head Start on children's cognitive and achievement outcomes: type and rigor of design, quality of dependent measure, attrition, and activity level of control group. Although the authors are still completing coding for some Head Start studies that will ultimately be included in their final analyses, preliminary examination of the available data reveals interesting variation along methods-related characteristics between contrasts. For example, they see substantial variation in the activity level of the control group; type of research design; whether baseline equivalency was tested, and if so, whether significant differences between groups were detected; and whether other sources of bias were detected by coders. Additional variation in the type and reliability of dependent measures, timing of tests, and attrition was also detected at the effect size level. (Contains 2 tables.)
Society for Research on Educational Effectiveness. 2040 Sheridan Road, Evanston, IL 60208. Tel: 202-495-0920; Fax: 202-640-4401; e-mail:; Web site:
Publication Type: Reports - Research
Education Level: Early Childhood Education
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: Society for Research on Educational Effectiveness (SREE)