Effect Size Calculation 

Once you've extracted the necessary information from your primary studies during the systematic review, then you will be able to use meta-analysis methods to apply statistical techniques to the data that has been collected. Typical steps of a meta-analysis include: 

  1. Calculating the effect sizes that match your research question's outcome(s) of interest

  2. Exploring the distribution of effect sizes, including estimating the mean effect size, its standard error and the heterogeneity among effect sizes in your sample

  3. Conducting a meta-regression using any relevant moderators to examine heterogeneity in effect sizes

  4. Reporting the results of your meta-analysis, including any subgroup analyses and meta-regression models​​​​​​​​​​

Our meta-analysis Video Guides provide more detailed information on how to calculate the most common effect sizes in education research, how to explore the distribution of effect sizes in your sample (estimating the mean effect size, its standard error and the heterogeneity among effect sizes), conduct meta-regression analyses, and conducting meta-analyses using R. For those who are interested in more background in R, those video guides also include an introduction to basic functions in that data analysis software.

Below is a list of the Video Guides:

Videos for getting started with R

1. Installing R and R Studio

2. Basics of R Studio

3. Importing data into R Studio

4. Transforming data in R

Introduction to the R program metafor (see Wolfgang Viechtbauer's website for more detail info)​

Videos for using escalc in metafor to compute

1. Standardized mean difference

2. Fisher's z transformation of a correlation coefficient

3. Log-odds ratio

1. Introduction to Meta-analysis​

2. Conducting a basic meta-analysis

3. Coding for meta-regression


Additionally, our meta-analysis Lecture Videos provide a deep dive into meta-analysis techniques that reflect the multilevel/hierarchical and correlated data structure of effect sizes. Below is a list of the Lectures from the Modern Meta-Analysis Research Institute (MMARI):


1. Data extraction for a meta-analysis

2. Extracting data from studies for an effect size

3. Introduction to meta-analysis using the R program metafor

4. Models for dependent effect sizes in meta-analysis (using metafor and clubSandwich)

5. Describing sources of heterogeneity in meta-analysis

6. Interpreting heterogeneity in a meta-analysis


Using R can facilitate the conduct of a meta-analysis. Several packages are used to conduct a meta-analysis in R. They include but are not limited to:

  1. metafor 

  2. robuMeta

  3. clubSandwich

  4. meta

  5. metaSEM 

  6. metaForest 

You can visit the RDocumentation page for each package to learn more.