Language Programs Effect Sizes Methodology Explained
Language Program Review Meta-Analyses Procedures
For the purposes of this site, searches of studies are conducted on Education Source, Sage Pub, Google, and company websites. All studies found with a control group, sufficient reporting data, and total samples above 10 are included in analyses.
Effect sizes for studies with sample sizes above 50 per group are conducted with a Cohen’s D calculation. Effect sizes for studies with sample sizes below 50 are conducted with Hedge’s G. When effect sizes are already calculated by the peer reviewed author, these effect sizes are usually accepted as is, unless I find probable cause for an improperly calculated effect size. As of this date, the only effect sizes re-calculated were for Reading Recovery studies, as the authors did not calculate against a control group, and for LLI, as the authors broke down their calculations in a non-standard way.
Please note, that while this work is sourced from peer-reviewed material, it is not peer-reviewed itself. Moreover, the best source for this type of information should always be viewed as a peer-reviewed meta-analysis. However, as most programs either did not have a peer-reviewed meta-analysis, or a dated meta-analysis, or a poorly done meta-analysis, I began to do this work, to bridge the research gap, until better peer-reviewed meta-analysis data is available.