Clicked Gallery

What is a Funnel Plot?

Highlighted from a real textbook passage. Explained by Clicked.

Used in a sentence

University Course Reader · STEM

The funnel plot was clearly asymmetric, suggesting that small negative trials had gone unpublished.

The reader highlighted one term — in an article or a paper PDF. Clicked made the statistics term “funnel plot” easy to understand:

Explained in three depths

Same facts, different vibe — Slang mode 😎

The Clicked way

●○○

Overview

A funnel plot is a chart researchers draw when pooling many studies on one question, to check whether any are missing from the published record. It is not the sales funnel chart that shows people dropping out of a process. Each study is one dot, placed by its result and its size. Big studies cluster near the top around the true answer, small ones scatter wider below, and the dots fill an upside-down funnel. One empty corner means studies probably went unpublished.
●○○

Overview

A funnel plot spots studies that should exist but don't. Not the sales funnel, where a thousand visitors shrink to ten buyers; that chart just borrowed the word. Here every study is a dot: big careful studies huddle at the top around the answer, small scrappy ones spread along the bottom on both sides. If one bottom corner is weirdly empty, the small studies that found nothing were never published, and that is publication bias with the lights on. 😎

A quick take — often all you need.

●●○

Detail

A funnel plot is a scatter chart researchers draw when they pool many studies of one question in a meta-analysis, to reveal whether studies are missing. It is not the funnel chart used in sales and marketing, which shows a group shrinking through stages; the two share a word and nothing else. Each dot is one study. Its position across shows the size of the effect that study found; its height shows how precise the study was, which mostly means how many people it included. Large studies are precise, so their dots sit high and close together, near the answer the whole set points to. Small studies are noisier, so their dots sit low and spread to both sides of that answer. An honest set of studies fills a symmetric upside-down funnel, and that symmetry is the test. If small studies that found nothing were left unpublished, the lower corner where they belong is empty and the funnel leans. A lopsided funnel is what publication bias produces, but other things produce it too. A few studies can make any shape by chance. And small trials are often run on the sickest patients, who have the most room to improve, so those trials can measure a genuinely larger effect. So a lopsided plot calls for a closer look, not a conclusion.
●●○

Detail

A funnel plot is what you draw when you have a pile of studies on the same question and want to know if the pile is complete. Forget the sales-and-marketing funnel chart; that one tracks people dropping out of a signup flow and shares only the shape. Here, one dot per study. Left to right is how big an effect the study found. Bottom to top is how big the study was. Big studies are precise, so they stack near the top in a tight clump around the answer. Small studies are jumpy, so they string out along the bottom, some overshooting the answer and some undershooting it. If every study that was run got published, the funnel would be symmetric. Now think about which studies go missing. It is the small ones that found nothing, because a small study that found nothing is the hardest thing in science to get printed. Those dots belong at the bottom on the disappointing side. Take them out and that corner is bare and the whole funnel tips. A tipped funnel is what publication bias looks like, but publication bias is not the only cause of a tip. Luck can do it with only a few studies. And small trials often pick the sickest patients, and sicker patients have further to recover, so those trials really do measure a bigger effect. So a leaning plot means go and look, not case closed. 😎

Want more? One click digs deeper.

●●●

Analogy

Thirty archers shoot at one target. The experienced ones group tightly around the centre. The beginners scatter to both sides, because a beginner has no particular direction of wrongness. Before the coach walks over, the beginners quietly pull out their worst misses, and those all happen to be on the left. The coach sees a tight core with a spray only on the right. She knows beginners miss both ways, so she knows arrows were pulled, and roughly how many. That is a funnel plot: the experienced archers are the big studies, the beginners are the small ones, and the pulled arrows are the studies nobody published.
Deepest level reached.
●●●

Analogy

A jar of sweets on the teacher's desk, and every kid writes down a guess. The careful kids who counted rows land close to the answer. The kids who glanced and guessed are all over the place, half too high, half too low, because guessing has no favourite direction. Pinned on the board, the guesses make an upside-down funnel: tight in the middle, wide and balanced at the edges. Then the kids who guessed way too low quietly bin their slips out of embarrassment. The centre still looks fine, but the spread leans high, and the teacher can see that low guesses have gone missing. That empty corner is publication bias, and the board is a funnel plot. 😎
Deepest level reached.

Unfamiliar concept? A real-world example makes it click — fresh analogies on tap.

AI explanations may contain errors · Not professional advice

Formal definition — The same term, explained the usual way

A funnel plot is a scatter plot used in meta-analysis in which each included study is plotted with its effect estimate on the horizontal axis and a measure of its precision, typically the standard error or sample size, on the vertical axis. In the absence of bias, estimates from smaller studies scatter more widely around the pooled effect than those from larger studies, producing a symmetric inverted funnel. Asymmetry, particularly a sparse region corresponding to small studies with null or unfavourable results, is taken as an indicator of publication bias or other small-study effects, and is formally assessed with tests such as Egger's regression. Asymmetry is suggestive rather than conclusive and requires an adequate number of studies to interpret. It is unrelated to the funnel chart used in sales and marketing analytics.

Want Clicked to explain terms like “funnel plot” directly in your browser — including on PDFs?

Add to Chrome — Free

50 free Explanations · No credit card required