Multilevel Models for Meta-Analysis

Abstract

Meta-analysis concerns the statistical integration of a large number of results from empirical studies (cf. Glass, 1976). The goal is to summarize the results of a collection of independently conducted studies on one specific research question. For instance, the research question might be: What is the effect of social skills training on socially anxious children? In a meta-analysis, one would collect reports of experiments concerning this question, explicitly code the reported outcomes, and integrate the outcomes statistically into a combined “super outcome”. Often the focus is not on integrating or summarizing the outcomes, but on more detailed questions about variations in the outcomes, such as: What is the effect of different durations for the training sessions? Are there differences between different training methods? In these cases, the meta-analyst not only examines the overall study outcomes, but also codes study characteristics. These study characteristics, for example design features or type of subjects sampled, are potential explanatory variables to explain differences in the study outcomes.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 98 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.