Publication bias

Publication bias is statistically distorted (English bias [ baɪəs ] ) representation of the data available in scientific journals as a result of a preferred publication of studies with " positive " or significant results. Positive results are easier to publish than those with " negative ", ie non-significant results and are also frequently published in journals with high impact factor. In medical drug research another cause for the publication bias is the preselection of negative results by pharmaceutical companies, most of which are sponsored studies.

Synonym for publication bias is also frequently used the term File Drawer problem ( " drawer problem "). So that related to the publication bias phenomenon is described that researchers increasingly their non-significant results until no longer submit for publication, but disappear in the same drawer.

Due to the increased frequency of positive results about the effectiveness of therapies may be overestimated in medicine, since studies with proven efficacy are easier to publish than those who can not prove the effectiveness. This is particularly relevant if you want to generate due to the previously published data situation on the basis of a meta-analysis of treatment recommendations. The suspicion of publication bias can be corroborated by creating a Funnel plots.

For these reasons, now require some prestigious medical journals that all the studies must be made known before. Only such studies publicized in advance are accepted for publication. This is in addition to other aspects provide an overview of the studies conducted on the subject in order to assess the publication bias can at least.

Moreover, there are already journals (mainly on the Internet, see below) that specifically studies with " negative ", ie within the meaning of the question not to publish significant results. The Cochrane Collaboration is very interested in such results in order to use them in their analyzes of the standards in medicine can.