Open data · CC BY 4.0 what each one claims, and what replicated no ads · no tracking
Bias Atlascognitive biases

Affect heuristic

also known as — Affect-as-information · Risk-as-feelings
How something feels stands in for how likely or how risky it is, so a judgement that looks like a calculation is often a reading of a mood.
Does it replicate? Mixed Finucane's risk-benefit experiment held in two preregistered repetitions, while the long-odds prediction reversed direction across 7,218 people.
VerdictMixed
First published2000
FieldDecision and judgement
Sources6
Last checked2026-09-21
What it claims

What does Affect heuristic mean?

The proposal is that people carry an affective tag on almost everything they think about, positive or negative, and consult it before they consult anything slower. Its sharpest prediction concerns risk and benefit. Finucane and colleagues argue that in the environment the two tend to be positively correlated, since an activity low in benefit would not be permitted if it were high in risk.

In people's judgements they come out inversely related: whatever feels good is rated both more beneficial and less dangerous. A second prediction, from Rottenstreich and Hsee, is that feelings should matter most exactly where probability information is hardest to feel, which is at very long odds.

Verdict

Repeats disagree, or the effect holds in some conditions and not others. The other 248.

The field
43 of the 85 entries in decision and judgement share this verdict.
Does it replicate?

Has Affect heuristic been retested?

Mixed Finucane's risk-benefit experiment held in two preregistered repetitions, while the long-odds prediction reversed direction across 7,218 people.

In the original studyd = 0.74
Pooled across the replication, aggregated = -0.08 (95% CI -0.13 to -0.03)
no effect Original Original: d = 0.74, no interval reported Replication Replication: d = -0.08, 95% CI -0.13 to -0.03
The same scale for both, so the distance between the two points is the change the repeat found. Hover a point for its interval.

Many Labs 2 put the kiss-versus-money choice in the first of its two slates, and pooled 7,218 participants. Where the original found 70% choosing the affective option under low probability against 35% under certainty, the replication found 47% against 51%, an odds ratio of 0.87 and d = -0.08, 95% confidence interval -0.13 to -0.03, p = 0.002.

The interval excludes zero, so this is a reversal rather than a null, and it is a small one. Klein and colleagues count it among three effects running opposite to their originals at p < .05, and it is not among those that clear the stricter p < .0001 line they also report. The paper prints no sample count for this effect.

It gives 64 samples for that slate as a whole, and reports this effect's individual samples as proportions: 3.33% significant in the reversed direction, 1.67% significant in the original's direction, 95.00% neither. Its heterogeneity test for this effect carries 59 degrees of freedom, which implies 60 samples, and the three proportions are exact sixtieths. That count of 60 is inferred from those figures rather than read off the paper, so it is not stored here as one of this entry's numbers.

The original d was 0.74 from 40 people, with a 95% interval Many Labs 2 gives as below 0.001 to 1.74. Finucane's own Study 2 has gone the other way. Efendić, Chandrashekar, Feldman and four colleagues repeated it on MTurk and Prolific under preregistration, analysing 1,552 participants after exclusions. Running the original's analysis, the correlation between manipulated and unmanipulated attributes was r(10) = -0.87, 95% CI [-0.96, -0.59] in the American sample and r(10) = -0.84, 95% CI [-0.95, -0.50] in the British one. The low-benefit condition was the exception, where risk ratings did not shift.

Read off Klein, R. A., Vianello, M., Hasselman, F., et al. (2018). Many Labs 2: Investigating variation in replicability across samples and settings. Advances in Methods and Practices in Psychological Science, 1(4), 443-490.

Examples

What are some examples of Affect heuristic?

The technology that feels excitingillustration

Someone taken with a new form of power generation rates it high in benefit and low in risk, while someone it unsettles rates it the other way on both counts. The two ratings move together because one feeling is answering both questions.

The food you are fond ofillustration

A cook who loves a particular cheese judges it both more nourishing and less of a problem for the heart than any evidence gives reason to. Liking has supplied the answer to a question about nutrition.

The prize you can pictureillustration

Offered a one in a million chance of a described holiday or the same chance of the cash it would cost, many people take the holiday. The odds are identical, and the difference is that one of them can be felt.

Cases marked as illustrations describe nobody in particular. They are there to make the pattern recognisable, not to report that a particular thing happened.

The experiments

What experiments is Affect heuristic based on?

The design

Finucane and colleagues ran two experiments, only the first of which used time pressure. Study 1 had 54 first-year psychology students at the University of Western Australia rate 23 hazards for risk and for benefit, 28 of them against a clock and 26 with as much time as they wanted. Correlations calculated on each person's ratings across 21 items averaged -0.45 under time pressure and -0.33 without it, and the gap between the two means gave t(52) = 1.64, p = 0.05 one-tailed, which the authors called approaching significance.

Study 2 gave undergraduates at the University of Oregon vignettes designed to raise or lower the perceived risk or benefit of nuclear power, natural gas and food preservatives. The attribute nobody had manipulated moved in the affectively matching direction, and the correlation across the resulting 12 data points was -0.75. That paper's method section reports 213 participants and its results section counts 219.

Rottenstreich and Hsee supplied the probability prediction, offering 40 people a choice between a kiss from a favourite film star and fifty dollars, either as a certainty or at a 1% chance. Many Labs 2's account of that study gives 35% taking the kiss under certainty and 70% under the long shot.

2 primary research3 replication and reanalysis1 background and reference

Origin

Who first described Affect heuristic, and when?

Year2000
WhoMelissa Finucane, Ali Alhakami, Paul Slovic and Stephen Johnson
WhereJournal of Behavioral Decision Making

Their paper is "The affect heuristic in judgments of risks and benefits". Slovic and colleagues set out the general account seven years later in the European Journal of Operational Research under the bare title "The affect heuristic".

The claim, and the replication this entry reads its figures from.

Where it runs out

When does Affect heuristic not apply?

What failed is one prediction from one paper, and the affect heuristic is broader than that paper. The inverse relationship between perceived risk and perceived benefit has survived repetition. Skagerlund, Forsblad, Slovic and Västfjäll report it in 2020 at r = -0.85 among the 187 people in their sample of 575 who rated both attributes, and at r = -0.77 among 41 students rating 64 activities in the laboratory.

The time-pressure claim rests on less than that, a one-tailed t of 1.64 in 54 students, and Efendić and colleagues repeated Finucane's Study 2 rather than Study 1. Read the failure as a verdict on the probability-weighting prediction rather than on the idea as a whole. The original kiss study is worth looking at before treating the reversal as a surprise: with 40 participants its interval reached 1.74, so it was compatible with almost any effect size including none.

Often confused with

Which biases is Affect heuristic confused with?

Each of these is named in this entry's own prose as something it gets taken for. Where the two came out differently, that is worth knowing before you quote either.

Commonly misread as

How is Affect heuristic commonly misread?

The affect heuristic is often invoked as though any emotional influence on a judgement confirms it. That makes the idea unfalsifiable and is not what its authors proposed; the claim has specific predictions and at least one of them has now failed at scale. It is also frequently merged with the availability heuristic, since vivid things are both easy to recall and easy to feel, but those are separate mechanisms making separate predictions.

And a heuristic is a shortcut rather than a mistake. Finucane and colleagues offered affect as a way of making judgement efficient when a decision is complex or mental resources are short, not as a source of error.

Sources

Sources for Affect heuristic

  1. 1Finucane, M. L., Alhakami, A., Slovic, P., & Johnson, S. M. (2000). The affect heuristic in judgments of risks and benefits. Journal of Behavioral Decision Making, 13(1), 1-17.primary
  2. 2Rottenstreich, Y., & Hsee, C. K. (2001). Money, kisses, and electric shocks: On the affective psychology of risk. Psychological Science, 12(3), 185-190.primary
  3. 3Klein, R. A., Vianello, M., Hasselman, F., et al. (2018). Many Labs 2: Investigating variation in replicability across samples and settings. Advances in Methods and Practices in Psychological Science, 1(4), 443-490. Figures read from the accepted manuscript in the LSE repository, including its Tables 2, 3 and 5.replication
  4. 4Efendić, E., Chandrashekar, S. P., Lee, C. S., Yeung, L. Y., Kim, M. J., Lee, C. Y., & Feldman, G. (2022). Risky therefore not beneficial: Replication and extension of Finucane et al.'s (2000) affect heuristic experiment. Social Psychological and Personality Science, 13(7), 1173-1184.replication
  5. 5Skagerlund, K., Forsblad, M., Slovic, P., & Västfjäll, D. (2020). The affect heuristic and risk perception: Stability across elicitation methods and individual cognitive abilities. Frontiers in Psychology, 11, 970.replication
  6. 6Slovic, P., Finucane, M. L., Peters, E., & MacGregor, D. G. (2007). The affect heuristic. European Journal of Operational Research, 177(3), 1333-1352.background

Every claim on this page was held against these sources on 2026-09-21. Nothing here is written from memory. Written and checked by Krishna Chagti; how entries are written and corrected. Found a mistake? Report it, with the sentence and a source.

About Affect heuristic

Is Affect heuristic real?

Finucane's risk-benefit experiment held in two preregistered repetitions, while the long-odds prediction reversed direction across 7,218 people. The effect sizes from the original study and from the replication are printed side by side above, so the comparison is visible rather than asserted.