Methodological issues in the meta‐analysis of quasi‐experiments