The Group-Diffusion Effect

The Group-Diffusion Effect

Reviewed by Faisal Zaidi

The group-diffusion effect is the tendency to judge your own risk as lower when more people share the same threat, even when the probability is stated. Price and Matthews proposed in 2009 that it and the ratio bias are one process seen from two sides. Both come from judging ratios, and which one appears depends on whether the denominator, the total exposed, or the numerator, the number affected, is the more salient term. The account rests on one paper of four experiments, so its status here is mixed.

What it is

The effect is easiest to see against its opposite. In the ratio bias, a larger count of cases makes a risk feel bigger even when the ratio is lower. In group diffusion, a larger total makes your own risk feel smaller even when the ratio is unchanged. The two look contradictory. Price and Matthews argued in Judgment and Decision Making in 2009 that they are not, because both are judgments of a ratio and the mind leans on whichever term the presentation makes more prominent.

Their paper reports four experiments. The first replicated the group-diffusion effect in US students using health scenarios. The second found the same pattern for lottery odds, for both winning and losing. The third showed that stating the number expected to be affected removed the effect. The fourth showed that stating only that number reversed it, producing a ratio bias.

The third and fourth experiments are the ones that carry the argument. If group diffusion were a fixed feature of how people think about shared threats, changing the wording of the same probability should not undo it. It did.

In effect

The practical reading is about who chooses the presentation. A message that stresses the size of the exposed group invites the thought that the risk is spread thin. A message that stresses the number of people affected invites the opposite. Neither is false, and neither changes the probability. The choice of which term to print is a choice of which effect the reader will meet.

It also bears on how denominator neglect should be described. Yamagishi's 1997 comparison, in which a disease killing 1,286 in 10,000 was judged riskier than one killing 24.14 in 100, is a case of the numerator winning. Price and Matthews place that result and group diffusion on one dial, with salience as the control.

What it does not say

It does not say that people ignore stated probabilities. The probability was given in these experiments, and the pattern appeared anyway, which is the point.

It does not establish how large the effect is outside the laboratory. The first experiment used US students, and the page you are reading has not traced the samples of the others.

It does not credit Price and Matthews with discovering group diffusion. Their contribution, as read here, is the salience account that links it to the ratio bias. The earlier literature on the effect has not been traced for this page.

And it does not settle the wider argument. Epstein's two-system reading of the bowl experiments describes the felt conflict between intuition and rule. The salience account predicts when the pull will weaken or flip. This publication sides with the salience account as the more economical explanation and leaves the two-system account standing as a description of what the conflict feels like. One paper is thin support for either.

Sources

  1. Price, P. C., & Matthews, T. V. (2009). "From group diffusion to ratio bias: Effects of denominator and numerator salience on intuitive risk and likelihood judgments." Judgment and Decision Making, 4(6), 436-446.
  2. Yamagishi, K. (1997). "When a 12.86% mortality is more dangerous than 24.14%: Implications for risk communication." Applied Cognitive Psychology, 11, 495-506.
  3. Read for this page: the publisher's article page for Price and Matthews, which gives the abstract and a description of Yamagishi's study, on 2026-10-23. The full text was not read, and the evidence status is mixed on that account.