Publication bias

Publication bias occurs when whether a study gets published depends on the direction or strength of its results. Positive and statistically significant findings are more likely to be written up, accepted, and cited, while null results sit in file drawers. When a systematic review pools only the published studies, the summary estimate can be inflated because the unfavorable evidence is missing. This matters most for meta-analysis. When reading a review, look for whether authors searched trial registries and grey literature, and whether they assessed asymmetry using a funnel plot or a formal test. A funnel plot that is lopsided, with small negative studies absent, is a warning sign. For example, early enthusiasm for some antidepressants shrank once unpublished trials submitted to regulators were included alongside the published ones. The question it lets you ask: does this body of evidence include the studies that found nothing, or only the ones that found something?