I heard that three researchers are needed to finalize the selection of articles that will be included in a meta-analysis, is this typically true in the field?
Two reviewers will satisfy the AMSTAR criterion:
http://amstar.ca/docs/AMSTARguideline.pdf
there is no condition on the number of researchers, but the meta analysis needs a lot of efforts, so 2 or more will lead to better, faster work with better quality.
Depends on the field and how many Research is already done. First I think, you should have some criteria. Along that criteria you decide which studies are to include an which you can exclude.
There is no specific number of researchers required. It should depend on the complexity of the quantitative review and the number of studies to be coded.
1 is inappropriate. 2 or more. You need cross-validation/checking when making important decisions regarding study selection and data extraction.
More specific details about reporting and protocol set up can be found in the following publication.
https://systematicreviewsjournal.biomedcentral.com/articles/10.1186/2046-4053-4-1
During the process of literature search and screening you should be two or better three researchers, so you can do inter-rater reliability analyses. In most cases a master's student and a PhD student do the work and if possible you should have an experienced researcher (e.g. Post-Doc or Prof) in the field of systematic reviews/meta-analyses, in case you have discrepancies that need to be resolved. With this amount of people you should be fine as well to do the rest of the work. I suggest contacting a librarian for a proper literature search because this is one of the crucial steps and source of systematic biases.
I personally can suggest Pim Cuijpers book "Meta-analyses in mental health research. A practical guide." It's short and a good read to get a first impression about meta-analyses.
I agree with the previous responses, at least two or three researchers for sharing the work of finding relevant studies (in some ways, the hardest part of a meta-analysis) examining inter-rater reliability, and then conducting the meta-analysis itself.
I have been a co-author on several meta-analyses, the biggest in terms of my effort involved two authors (Seto & Lalumiere, 2010, Psychological Bulletin) and most of the others involving at least three authors.
Three is a good number. Two will do selection of articles individual and compare their decisions, then for the articles that the decisions are not matching, a third researcher will make the final decision.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis