A distortion in conclusions derived from published studies because of the selective factors associated with the likelihood of publication, including whether the findings were positive and statistically significant, and the potential proprietary interests of sponsors.