ABSTRACT
I 2 is an index of heterogeneity of the effect sizes in different studies compiled for meta-analysis. Effect size could be the difference in means or in proportions, odds ratio, relative risk, or any other measurement. Meta-analysis is the term used for the method of combining evidence from various studies on the same effect. One of the requisites for studies compiled for this purpose is that they should not have widely different results from one another. Some variation is natural, and that is why we need meta-analysis, but that should be within the range generally perceived as natural across studies in different areas and in different populations with different sample sizes, and possibly with some minor variation in methodology. It is preferred that the studies chosen for meta-analysis have similar methodology, but that might be difficult to ensure.
