ABSTRACT
144The meta-analysis should contain enough studies to provide power for its test. If a meta-analysis performs moderator tests, it should also report if there are any relationships between the moderators. All meta-analyses will have at least two authors to ensure coding reliability. If there are a large number of assumed zero effect sizes, the authors should report their results both including and excluding these values from their analyses. To assess the representativeness of a particular meta-analysis one should consider theoretical boundaries, exhaustive search, secondary literature, unpublished literature, large literature, and high inference moderation. A meta-analysis should not simply be a summary of a literature but should provide a theoretical interpretation and integration. A good meta-analysis puts effort into interpreting these findings, presents how they are consistent or inconsistent with the major theories in the literature and encourages future investigations. Although it can be argued that the results of a systematic review should stand on their own, many people faced with a decision look to the discussion and authors’ conclusions for interpreting the results. Indeed, many people prefer to go directly to the conclusions before looking at the rest of the review. The discussion and conclusions should help people to understand the implications of the evidence in relationship to practical decisions. A good starting point for the discussion section of a review is to address any important methodological limitations of the included trials and the methods used in the review that might affect practical decisions about healthcare or future research. One type of evidence that can be helpful in considering the likelihood of a cause–effect relationship between an intervention and an important outcome is indirect evidence of a relationship. Decisions about applicability depend on knowledge of the particular circumstances in which decisions about healthcare are being made. Important variations include biologic, cultural, compliance, baseline risk, and results of the included studies. It is safer to report the data, with a confidence interval, as being compatible with either a reduction or an increase in the outcome. The easier way to write up a meta-analysis is to take advantage of this parallel structure by using the same sections found in primary research. In the method section, you need to describe how you collected your studies and how you obtained quantitative codes. In the results section, you describe the distribution of your effect sizes and present any moderator analyses you decided to perform. 145Discussion and conclusion section conclude with specific recommendations for the direction of future research. Reference and appendix section have a single reference section that includes both studies used in writing the paper and those included in the meta-analysis.
