Skip to content
BENIBIX Learn

Publication Bias

General🔬 ResearchResearch Quality

Also known as File Drawer Problem, File Drawer Effect

Exciting results get printed more than dull ones. So what we see is lopsided.

Think of it like

It is like a fishing show. It only shows the days a big fish was caught.

Example

Ten teams test the same diet. Three find it works and publish. Seven find nothing and never publish, so readers see only the three.

Why it matters

It can make a treatment look like it works when most tests found nothing. Reviews and meta-analyses carry the error forward.

Definition

Publication bias is when studies with exciting results are more likely to be printed than studies that found nothing. Studies with no effect often stay in a drawer. So what is in print makes effects look bigger and more common than they are. Reviews that only read what is in print share the problem. It helps to put a study's plan on record before it starts, so it counts either way.
Right panel: only the exciting studies get into the journal. The rest stay in a drawer.

Common mistake

Thinking that a stack of positive papers settles a question. Ask what studies may be missing.

Key takeaways

  1. 1
    Exciting gets printed. Null results often vanish.
  2. 2
    Skews the record. Effects look bigger than they are.
  3. 3
    Register first. Plans on record before results.