Contents
What is an effect size in a meta-analysis?
Effect size is a statistical concept that measures the strength of the relationship between two variables on a numeric scale. In Meta-analysis, effect size is concerned with different studies and then combines all the studies into single analysis.
How many studies are needed for meta regression?
Two studies
Two studies is a sufficient number to perform a meta-analysis, provided that those two studies can be meaningfully pooled and provided their results are sufficiently ‘similar’.
What is the design of a meta-analysis study?
Meta-analysis is a quantitative, formal, epidemiological study design used to systematically assess the results of previous research to derive conclusions about that body of research. Typically, but not necessarily, the study is based on randomized, controlled clinical trials.
What is the formula for a meta regression?
A standard regression equation, therefore, looks like this: In meta-regression, the variable y y we want to predict is the observed effect size ^θk θ ^ k of study k k. The formula for a meta-regression looks similar to the one of a normal regression model:
How are effect sizes used in a meta-analysis?
In Chapter 3.1, we already learned that observed effect sizes ^θ θ ^ can be more or less precise estimators of the study’s true effect, depending on their standard error. In “normal” meta-analyses, we take this into account by giving studies a smaller or higher weight.
Which is a fixed effect in a meta regression?
This means that between-study heterogeneity exists in our data, which is captured by the heterogeneity variance τ 2 τ 2. Since the equation above includes a fixed effect (the β β coefficient) as well as a random effect ( ζk ζ k ), the model used in meta-regression is often called a mixed-effects model.
What are true variances in a meta-regression?
All studies share a common τ2, i.e., they come from the same super-population of studies [7]. The observed sampling variances σi2 are the “true” variances within each study [4]. The distributional assumptions about the random effects and error terms are correct.