Abstract

Seeing the Forest Plot for the Trees: Using Meta-Analysis to Synthesize Research

Findings from individual studies are limited. Any one study may have inaccurate results due to small samples, methodological flaws, implementation issues or random chance. Therefore, replication is a crucial aspect of science. However, conflicting results across studies and variations in methods (or outcomes) can make it challenging to synthesize findings, especially in narrative reviews with a lot of studies. This can sometimes make it difficult to see the forest for the trees.

Meta-analysis can help synthesize quantitative research findings by statistically combining results from multiple studies. The resulting average weighted effect size may provide a more precise estimate of the outcome of interest than any individual study. Additionally, meta-analysis allows for examination of heterogeneity through moderator analyses, which can help explain potential variance in effects across studies.

This presentation will provide an overview of advancements in meta-analytic techniques, including robust variance estimation, multi-level modeling approaches and meta-analytic structural equation modeling (SEM). How these techniques impact research questions will be discussed, along with literature searches, coding, data structuring and analyses, and limitations and factors that can produce flawed interpretations.