Back to the directoryFind in the timeline
Framing: the Asian disease problem and price labels
Tversky and Kahneman showed in classroom questionnaires that describing the same outcomes as gains or as losses changed majority choices. The article also discussed a calculator price problem and credit-card price labels. Later work found framing effects vary by task.
- Where it sits
- Academic psychologyResearch published for other researchers and students.
- Evidence level
- DocumentedThe event or practice is documented by primary or authoritative sources.
- Sources inspected
- 4 of 43 with the relevant text, company document or archive item checked, 1 as an abstract or reference article, 0 not verified.
What happened
In January 1981 Amos Tversky and Daniel Kahneman published The Framing of Decisions and the Psychology of Choice in Science. Using brief classroom questionnaires, they showed that the same decision problem produced different majority choices when outcomes were described as gains or as losses. The best known case is the Asian disease problem. The article also discussed a calculator price problem and card surcharges.
The psychological idea
A frame is the decision-maker's conception of the acts, outcomes and contingencies in a choice. In the Asian disease problem, 600 lives are at risk, and saving 200 for certain describes the same outcome as 400 dying for certain. The authors link the reversal to the S-shaped value function of prospect theory: risk aversion for gains and risk seeking for losses.
In practice
In Problem 1 (N = 152), 72 per cent chose the program that saved 200 lives for certain. In Problem 2 (N = 155), with outcomes worded as deaths, 78 per cent chose the gamble. In a pricing problem, 68 per cent would travel to save $5 on a $15 calculator but 29 per cent on a $125 one. The authors, citing Thaler, reported that credit-card representatives asked for a price gap to be called a cash discount, not a surcharge.
Why it mattered
The article set out, with questionnaire data, a vocabulary of frames and reference points for decisions about money and lives. Levin, Schneider and Gaeth (1998) later described the risky-choice effect from this paper as one of three kinds of valence framing, with attribute and goal framing, because results differed by task.
Limits and qualifications
The data come from student questionnaires with hypothetical choices, and the authors call the effects large but by no means universal. In 151 participants, Zhen and Yu (2016) found low agreement between two classic framing tasks, though some pairings did correlate. Data Colada replicated a test that added the word exactly, after an objection that respondents read 200 as at least 200, and still found a strong effect. The price-label example is the authors' reading of an industry request, not a test of it.
Evidence behind the claims
In the Asian disease problem, with a disease expected to kill 600 people, 72 per cent of 152 respondents chose Program A (200 saved for certain) and 28 per cent chose Program B; with the deaths wording, 22 per cent of 155 chose Program C (400 die) and 78 per cent chose Program D.
The data came from students at Stanford and the University of British Columbia who answered brief classroom questionnaires, the authors link the reversal to the S-shaped value function of prospect theory, and they call the effects large and systematic but by no means universal.
In the calculator problem, 68 per cent of one group would make an extra trip to save $5 on a $15 calculator while 29 per cent of another would do so for a $125 calculator (groups of 93 and 88).
The authors, citing Thaler, say credit-card industry representatives asked that a price difference be labeled a cash discount rather than a credit-card surcharge, and they interpret the two labels as setting different reference points.
Levin, Schneider and Gaeth (1998) proposed three kinds of valence framing, risky-choice, attribute and goal framing, describing risky-choice framing as the effect introduced by Tversky and Kahneman (1981) and arguing that inconsistent results come from different operational definitions of framing.
Zhen and Yu (2016) tested 151 participants on a vignette task and a gambling task, found framing effects in both, and found low convergent validity between them (for example rs = 0.056 for the vignette loss-domain and gambling gain-domain pair), while some other pairings correlated significantly.
Data Colada reported that with the word 'exactly' added to the program descriptions, roughly 98 participants per cell still showed a strong framing effect, which they take to show the effect is not caused by a reading of 200 as at least 200.
People involved
- Amos Tversky. Psychologist and co-author
- Daniel Kahneman. Psychologist and co-author
- Richard Thaler. Economist whose credit-card example the authors cite
- Irwin Levin. Author of the 1998 typology of framing effects
In the Psy101 course
- OpenStax Psychology 2e, section 7.3: Problem SolvingChapter 7, Thinking and Intelligence. Context. Nearest related section; it does not discuss this topic by name.Covers decision biases such as anchoring, the nearest textbook material. It does not mention framing by name.
This is course relevance only. It is not evidence for the historical or corporate claims on this page.See the full course map.
Sources and inspection
4 of 4 sources were opened and read at the level shown. A source marked as not verified supports no claim on this page by itself.
- [1]The Framing of Decisions and the Psychology of ChoiceTversky, A; Kahneman, D. Science 211(4481), 453-458 (copy hosted at gwern.net), 1981. Journal article. Primary source.Relevant text checkedRead all five pages (453 to 457) of the hosted scan as images: summary, Problems 1 to 10, framing of outcomes and discussion.
- [2]All Frames Are Not Created Equal: A Typology and Critical Analysis of Framing EffectsLevin, I; Schneider, S; Gaeth, G. Organizational Behavior and Human Decision Processes 76(2), 149-188 (University of South Florida repository record), 1998. Journal article. Secondary account.Abstract or publisher record onlyRead the repository page with the abstract as returned by the fetch tool. The paper itself was not opened.
- [3]All framing effects are not created equal: Low convergent validity between two classic measurements of framingZhen, S; Yu, R. Scientific Reports (via Europe PMC, PMC4951804), 2016. Journal article. Critique or replication.Relevant text checkedRead the open-access full text from Europe PMC as returned by the fetch tool (participants, tasks, results, conclusion). The paper does not use the original Asian disease wording.
- [4][11] Exactly: The Most Famous Framing Effect Is Robust To Precise WordingSimmons, J; Nelson, L. Data Colada (blog), 2013. News report. Critique or replication.Relevant text checkedRead the blog post as returned by the fetch tool, including the replication of Mandel's test. It is a blog, not a peer-reviewed paper.
Access note: Thaler 1980 and Kahneman, Knetsch and Thaler 1986 were not opened; the credit-card example rests on the 1981 article's own account.
Independent source review: Claude review agent, independent of the research agent, 2026-10-21. Verdict: pass-with-edits.