Tutor Hunt Questions

Meta-Analysis is used to find a consistent conclusion from a set of similar research papers. Example, are SSRIs (a type of antidepressant) useful for severe depression? A meta-analysis would involve finding all research papers that investigate this. They will exclude poor quality papers, in order to improve the quality of the conclusion. They`ll then try to find a consistent conclusion from all the quality papers, concluding more accurately than a single paper, whether SSRIs are good for severe depression.

There are limitations to be aware of in meta-analysis, for example where people don`t publish ALL research in that area. An oft cited example is pharmaceutical companies not publishing research that says, that their products don`t work, or have severe side-effects. This is called publication bias. It is important that we find quality research, that includes various outcomes, not just the outcomes people want. That way, the conclusion will be more accurate.

Example, if a 100 studies were done on SSRIs for severe depression. If 50, showed they didn`t work, or had severe side-effects, then the pharmaceutical company would have a financial incentive to not publish! This would bias the meta-analysis.

A meta analysis is a study of studies essentially. Every study has some error associated with it and the researchers doing the meta analysis will pool all of the results and use statistics to provide a general pattern of results.

If you read up on the discovery of the Higgs Boson, you will see statistical analysis in action. It`s discovery was the results of millions of experiments where its existence was proved beyond statistical doubt.