Publication bias is the tendency of journals to publish studies with statistically significant results more often than studies that find no effect. Studies with null results often go unpublished, which is called the file drawer problem.
Why it matters
If only the positive studies get printed, the published research makes an effect look stronger and more reliable than it really is. This is one reason many findings fail when other labs test them again.
How the exam tests it
Questions often describe a study with p < .05 that fails when an independent lab repeats it. The stem may then ask you to name the replication crisis, which publication bias helps cause.
Don’t confuse it with
Students confuse publication bias with the replication crisis, but publication bias is a cause, the selective printing of positive results, while the replication crisis is the result, the discovery that many findings do not hold up.