ABSTRACT
A problem meta-analysts frequently face is that suitable effect size data cannot be extracted directly from all included studies.
This chapter describes formulas and R functions that allow to calculate and convert among effect sizes. It covers how to calculate a standardized mean difference (SMD) from the mean and standard error; how to convert a point-biserial correlation to SMDs; how to calculate SMDs from a one-way ANOVA, as well as from an independent two-sample t-test; how to calculate the standard error from p values; how to convert a χ 2 - https://s3-euw1-ap-pe-df-pch-content-public-p.s3.eu-west-1.amazonaws.com/9781003107347/de07b140-3dcb-4e86-8547-42a1b8d32683/content/math0_3.tif"/> statistic into an odds ratio; how the number needed to treat is defined, the limitations of this measure, and how it can be calculated for effect sizes based on continuous or binary outcome data. Lastly, the chapter also shows how to pool effect size data of multi-arm studies.
