Publication bias describes how the realities of research publishing can affect the quality of the science it supports, particularly in relation to any meta-analysis. Researchers are more likely to publish a paper that reports on a relatively large treatment effect, and less likely to do so when a study yields a relatively small effect. This can lead to a situation where the body of studies published on a certain topic is biased, in that it will tend to exaggerate the true magnitude of the treatment effect.