Project Activities
To achieve the goals of the project and conduct the proposed analyses, researchers will carry out the following activities. First, the team will update the current meta-analytic database with the most recent longitudinal research on school readiness. Second, the team will add additional longitudinal correlations between early childhood predictors and school readiness, which has been defined more broadly in this project than in the original meta- analysis. Third, the researchers intend to fully explore and analyze the specific relationships between different school readiness constructs and the various facets of later school performance. Finally, the research team will identify critical variables measured prior to school entry that predict school readiness and can serve as targets for intervention in preschool classrooms. Researchers will use meta-analytic procedures to search for additional studies, code the relevant information, and conduct analyses that will address the research objectives.
Structured Abstract
Setting
The data to be analyzed will be drawn from a large meta-analytic database of longitudinal studies of school outcomes including all U.S.-based studies available since 2004 that meet the eligibility criteria.
Sample
The existing meta-analysis database is comprised of longitudinal and cross-sectional correlations obtained from longitudinal panel studies. The original study focused on any predictor of school performance, school conduct, or school participation, measured at any age. That database includes 507 longitudinal studies, representing over 300,000 children. Further, 246 of those studies use a school readiness variable as a predictor and 200 use school readiness as an outcome.
There is no intervention.
Research design and methods
The first two research objectives involve determining the ability of measured school readiness to predict later school success and the ability of different risk and protective factors, observed before the start of formal schooling, to predict the elements of school readiness. The third objective involves identifying differences in observed predictive relationships that are associated with characteristics of the children in the studies. The fourth objective is to examine the correlations among different risk and protective factors at different age periods to determine which tend to co-occur and the implications for identifying broader multivariate predictive factors. Researchers will use meta-analytic procedures to search for additional studies, code the relevant information, and conduct analyses that will address the research objectives. Specifically, researchers will identify, select, and retrieve relevant U.S.-based studies available since 2004 that meet the eligibility criteria. To begin, abstracts of research reports will be reviewed for relevance and final eligibility screening will be based on the entire article. Researchers will then complete systematic coding of the characteristics and findings of the studies. The coding scheme will extract two different types of information. One type consists of effect sizes that represent the relationships of interest (i.e., relationships between risk/protective variables and school readiness, relationships among risk variables or among outcome variables, relationships of school readiness to later school success and failure). The other type of information consists of study descriptors—characteristics of study circumstances, samples, methods, procedures, and the like that may be related to differences in their findings.
Control condition
There is no control condition.
Key measures
The key outcomes are those that examined academic and social behavioral indicators of children's school readiness skills. Children's school readiness skills were measured using a range of standardized assessment tools. Across the range of studies in the meta-analytic database, data was collected using measures of children's cognitive abilities, early mathematics and literacy skills, social behavioral competence, and self-regulation.
Data analytic strategy
Multivariate analyses for meta-analysis of correctional effect sizes will be used. Conventional meta-analytic methods require that effect size estimates be statistically independent, which means only one effect size estimate per study sample for each outcome construct is used in an analysis. For this study, a statistical technique that allows inclusion of multiple effect sizes per study sample will be used. With this technique, robust standard errors are estimated for correlated effect size estimates that adjust for their lack of statistical independence. Therefore, all relevant effect sizes can be represented in the analysis and, as a result, the number available for multivariate analysis is substantially increased. Application of this technique will support differentiating the variables that predict school readiness, the facets of school readiness, and their relationship to later school outcomes.
People and institutions involved
IES program contact(s)
Products and publications
Products: Products include reports describing relationships between different school readiness constructs and later school performance. In addition, information identifying critical intervention variables prior to school readiness will be included.
Journal article, monograph, or newsletter
Wilson, S.J., Polanin, J.R., and Lipsey, M.W. (2016). Fitting Meta-Analytic Structural Equation Models With Complex Datasets. Research Synthesis Methods, 7(2), 121-139.
Supplemental information
Co-Principal Investigators: Dale Farran and Mark Lipsey
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.
