Publication bias refers to:

• A. The tendency to publish only studies from prestigious authors.
• B. The bias in selecting which variables to measure.
• C. The fact that articles published in scientific journals tend to over-represent positive findings.
• D. The bias in sample sizes across studies.

Answer: (C)

Publication bias occurs when the published research is not representative of all the work done because studies with positive or significant results are more likely to be published. This makes journals and reviews over-represent positive findings, which can distort the overall picture when researchers summarize evidence across studies, such as in meta-analyses. The bias can arise because journals prefer novel or significant results, authors may not submit null results, and there is sometimes selective reporting within studies. For example, many small trials showing a treatment works might get published, while those with null results remain unseen, pushing the average effect upward. This isn’t about publishing only studies from famous authors, choosing which variables to measure, or differences in sample sizes across studies.