Funnel plot and Egger test

The funnel plot shows each study by effect and standard error. Precise studies lie at the top close to the pooled effect, imprecise ones scatter more widely below: a funnel appears. If one side is missing, for example because small studies without an effect stayed unpublished, it is lopsided. The Egger test checks this asymmetry with a regression of the effect on the standard error.

Example: Egger test p = .19: no clear evidence of asymmetry (little power with few studies).

Caution: With fewer than 10 studies the test has little power (Sterne and colleagues, 2011). Asymmetry can also mean real differences between small and large studies.