Last update:


Arch Hellen Med, 30(3), May-June 2013, 340-354


The publication bias phenomenon in health sciences: Exploration of statistical methods of correction

O. Archangelidi
Clinical Epidemiology Group, Department of Epidemiology and Public Health, University College London, London, UK

Publication bias is a term used to refer to a tendency towards the preparation, submission and publication of research findings based on the nature and direction of the research results. It is a phenomenon reflected in increased publication rates among studies that show a statistically significant effect of treatment, leading to a higher prevalence of false positive published results.The subjective component of publication bias involves the "biased" opinion of the investigator and is often reflected in the tone of the presentation, while the objective component is more subtle and depends on the fact that the decision to publish is often influenced by positive results of a study. The primary causes of objective publication bias are author and editorial preferences, in addition to selective analytical maneuvers and other subjective influences. The present review explores several statistical methods that have been developed to identify and quantify the impact of publication bias on meta-analyses. Basic strategies are discussed that if widely adopted would go a long way towards alleviating publication bias, including the search for, and inclusion of "gray" literature and prospective registration of studies.

Key words: Bias, Publication, Quality, Science, Study, Validity.

© Archives of Hellenic Medicine