We found a match
Your institution may have rights to this item. Sign in to continue.
- Title
Comparison of Two Methods to Detect Publication Bias in Meta-analysis.
- Authors
Peters, Jaime L.; Sutton, Alex J.; Jones, David R.; Abrams, Keith R.; Rushton, Lesley
- Abstract
The study discusses Egger's regression tests that help detect publication bias in meta-analyses. The performance of the Egger's test is compared with another regression test, which is based on sample size, is a modification of Macaskill's test, and can be used when the summary estimate is the natural log of the odds ratio (InOR). Study methods include the simulation of scenarios with and without publication bias and between-study heterogeneity, as well as an outcome measure of Type 1 error rates. The authors conclude that the alternative test can be used in place of the Egger's regression test when the estimates are InORs.
- Subjects
REGRESSION analysis; META-analysis; STATISTICAL bias; MEASUREMENT errors; STATISTICAL reliability; STATISTICAL sampling; RESEARCH methodology; QUALITY control; MEDICAL publishing
- Publication
JAMA: Journal of the American Medical Association, 2006, Vol 295, Issue 6, p676
- ISSN
0098-7484
- Publication type
Article
- DOI
10.1001/jama.295.6.676