School failure and related problems are not only adverse events in their own right for the children who experience them, but are also related to a broad range of subsequent negative adolescent and adult outcomes. This circumstance makes it especially important to identify and understand the risk factors that predict school failure for purposes of selecting and diagnosing children in need of remedial services, designing preventive programs to offset risk factors and enhance protective ones, and guiding research on the causal processes that produce poor school outcomes. Research relevant to this issue includes a large number of longitudinal studies in which school outcomes and antecedent risk/protective variables are measured. However, no review paper, conceptual model, or systematic synthesis has yet drawn on the full depth and breadth of this research literature to compile, organize, and analyze the empirical relationships it reports. The overall aim of this project is to conduct a systematic synthesis of this body of research using advanced meta-analysis technique in order to address the following issues: 1. The capability of risk and protective variables observed during different age periods to predict school success or failure at later ages. The predictive ability of these variables will be examined in terms of the overall strength of association (product-moment correlation) and in terms of predictive accuracy (e.g., sensitivity and specificity). 2. The combinations of risk and protective variables that tend to co-occur during different age periods and the strength of their collective predictive ability when combined in multivariate factors. Particular attention will be paid to the convergence of empirical data with theoretical formulations about associations between risk/protective factors and school outcomes. 3. The extent of the co-occurrence (comorbidity) of school failure with other problem behaviors such as externalizing and internalizing behaviors and substance use at various age periods. 4. The differences in predictive relationships between risk/protective variables and school success/failure associated with the age, gender, and ethnicity of the students.
| Wilson, Sandra Jo; Polanin, Joshua R; Lipsey, Mark W (2016) Fitting meta-analytic structural equation models with complex datasets. Res Synth Methods 7:121-39 |
