Review of methodology of quantitative reviews using meta-analysis in ecology

Abstract

1. Statistical methods for combination of independent results have been well publicized in the ecological literature, and have begun to be used for reviewing research. They provide a considerable advance in scientific rigour over traditional narrative or 'vote-counting' reviews. 2. However, other methodological developments for research synthesis have not yet been widely adopted. 3. This review briefly summarizes some of the techniques used for carrying out rigorous reviewing and synthesis of results in medical science, and surveys techniques used by ecological meta-analyses. 4. Many of the methods used to reduce bias and enhance the accuracy, reliability and usefulness of reviews in medical science have not yet been widely used by ecologists. 5. The quality of ecological reviews could be improved by adoption of some of these methods, such as specifying the methods used for literature searching, stating the types of study combined in the review and the strength of evidence they provide, presenting results as a point estimate with a confidence interval, investigating bias in selection of studies using funnel plots, making a clear distinction between the main analysis and subsidiary analyses and interpreting the latter with caution, and performing sensitivity analyses.