The Framing Effect: An S-Tier Behavioral Designer’s Guide to Gain vs Loss Framing

Two hospitals publish the exact same survival data for a new cancer treatment. One says: “90% of patients are alive after five years.” The other says: “10% of patients are dead after five years.” Identical statistics. In study after study, patients — and even the doctors advising them — choose differently depending on which version they read first.

That gap between the information and the decision is the territory Amos Tversky and Daniel Kahneman staked out in 1981. They called it the Framing Effect, and it is the single most important idea a designer, marketer, policymaker, or product manager can steal from behavioral economics. Because the effect is this: the frame around a choice can flip the choice, even when every underlying number is identical.

If you ship experiences for human beings, you are framing whether you admit it or not. The question is whether you are doing it with intention — the way Octalysis-aware designers do — or by accident, the way most teams still do. This post is the long version of that answer: what the Framing Effect actually is, what the research really shows, where it is weaker than the pop-psych version suggests, and how to use it in Octalysis design without crossing the line into manipulation.

A frame never operates in a vacuum, though. Whenever I analyze a model like this one, I cross-reference it against the rest of my Behavioral Framework Library, where the Framing Effect sits next to Prospect Theory, Nudge Theory, and every other framework I use to pressure-test a design before it ships.

Speed Run Notes

• The Framing Effect is not a “bias” in the dismissive sense. It is a stable, reproducible feature of how humans evaluate options — preferences are constructed by the presentation, not merely revealed by it.
• Same math, reversed choice. Tversky & Kahneman’s Asian Disease Problem (1981) showed that 72% of subjects pick the certain option when a scenario is framed in terms of lives saved, while 78% flip to the risky option when the exact s…
• Gain frames push people toward risk-aversion. Loss frames push people toward risk-seeking. That asymmetry is the behavioral signature of Prospect Theory’s value function — the kinked curve that is steeper for losses than gains.
• There are three dialects of framing, not one. Attribute framing (“95% lean” vs “5% fat”), goal framing (“if you do this, you’ll gain” vs “if you don’t, you’ll lose”), and risky-choice framing (the Asian Disease pattern).
• The effect is real, but the pop-sci version overstates it. Meta-analyses (Kühberger 1998; Piñon & Gambara 2005) put the average effect in the small-to-medium range, with wide variation by domain, numeracy, and whether the decision is truly high-stakes.
• Framing maps cleanly onto Octalysis. Gain frames activate the White-Hat Core Drives — Epic Meaning, Development, Creativity.

About the Author

Yu-kai Chou is an S-Tier Behavioral Designer and the creator of the Octalysis Framework, the gamification design system now applied to products and experiences reaching over 1.5 billion users. His book Actionable Gamification is one of the most-cited works in the field, and he has been ranked the #1 Gamification Guru in the World.

He has advised MrBeast, LEGO, Microsoft, Porsche, Tesla, Stanford, Harvard, and governments including Ukraine on turning behavioral psychology into product mechanics that actually change user behavior.

Verify: Wikipedia · Google Scholar · Wikidata · LinkedIn

What Is the Framing Effect?

The Framing Effect is the finding that people’s decisions change based on how a logically equivalent piece of information is presented. Describe a choice as a potential gain and most people will reach for the safer option. Describe the identical choice as a potential loss and the same people will reach for the riskier one. The facts do not move. The decision does.

Tversky and Kahneman introduced the term in their 1981 Science paper, “The Framing of Decisions and the Psychology of Choice.” The paper did two things at once. First, it gave behavioral economics a clean laboratory demonstration that preferences are not stable objects that people “reveal” through their choices — they are constructed on the spot, shaped by context, reference points, and language. Second, it linked the effect to Prospect Theory’s value function, which the same authors had published two years earlier. The value function is concave in the gain region (so a sure gain looks sweet) and convex in the loss region (so a sure loss looks unacceptable). Change the frame and you change which region the decision lives in.

This definition is important because “framing” in everyday speech has sloppy edges. Journalists use it to mean “spin.” Politicians use it to mean “messaging.” Marketers use it to mean “positioning.” The academic construct is narrower and more useful: two descriptions are frames of the same decision when they are logically equivalent — same options, same probabilities, same outcomes — but different in surface presentation. If you can mathematically translate version A into version B without changing any of the underlying quantities, any measurable difference in behavior between them is a framing effect.

That narrowness is what gives the Framing Effect its teeth for designers. It tells you that even after you have fixed the product, the price, the probabilities, and the outcomes, you still have one more decision to make: the frame. And that last decision can move real behavior.

The Core Findings: Asian Disease and Beyond

The Asian Disease Problem

The iconic demonstration is the Asian Disease Problem. Subjects were told to imagine the United States was preparing for an outbreak of a disease expected to kill 600 people, and that two alternative response programs had been proposed. Half of subjects read the gain framing:

“If Program A is adopted, 200 people will be saved. If Program B is adopted, there is a one-third probability that 600 people will be saved and a two-thirds probability that no one will be saved.”

The other half read the loss framing:

“If Program C is adopted, 400 people will die. If Program D is adopted, there is a one-third probability that nobody will die and a two-thirds probability that 600 people will die.”

Programs A and C are statistically identical. Programs B and D are statistically identical. The only thing that changes between the two scenarios is the verbal framing of the outcomes. And yet the behavior swings dramatically: 72% of subjects chose A in the gain frame, while 78% chose D in the loss frame. In other words, the same expected value was embraced with high certainty when it was a gain and rejected in favor of a gamble when it was a loss.

That reversal is not a blip. It has been replicated thousands of times across disease scenarios, financial decisions, legal negotiations, management dilemmas, and consumer purchases. It is the behavioral fingerprint of loss aversion — but crucially, loss aversion activated by a verbal change alone, not by any change in stakes.

Three Dialects of Framing

In 1998 Irwin Levin and colleagues did the field a major favor by splitting the umbrella term into three distinct framing types. Lumping them together is a common mistake; each has its own signature, its own mechanism, and its own practical implications.

Risky-choice framing is the Asian Disease pattern: a choice between a sure outcome and a probabilistic outcome, described in gain or loss terms. This is the most famous version and the one most tightly linked to Prospect Theory. It is also the version with the largest effect sizes in the literature.

Attribute framing reframes a single attribute of an object. Ground beef described as “75% lean” is rated more favorably than ground beef described as “25% fat,” even though both are identical. A surgical procedure with “90% survival” is more acceptable than one with “10% mortality.” Here there is no risk or probability — just a single adjective flipped from positive to negative polarity. Effect sizes are typically smaller but remarkably consistent.

Goal framing shifts the consequence of acting or not acting. “If you perform breast self-examinations, you have a greater chance of catching tumors early” (gain frame for acting) versus “If you do not perform breast self-examinations, you have a reduced chance of catching tumors early” (loss frame for not acting). Meta-analyses find goal framing produces a consistent but modest edge for loss-framed messages in health communication — the domain where the stakes of inaction feel most salient.

For designers, the typology matters. If you are asking users to choose between a default plan and a riskier upgrade, you are in risky-choice territory. If you are describing a feature on a landing page, you are in attribute territory. If you are asking someone to take or skip a behavior, you are in goal territory. The lever you pull is not interchangeable.

Why the Reversal Happens

The most widely accepted explanation is that the verbal frame triggers a reference point. In the gain frame, the implicit reference is zero lives saved — anything above zero feels like progress, and risk-taking threatens that progress. In the loss frame, the implicit reference is 600 lives healthy — the scenario is now a descent from a happier baseline, and the risky option at least carries the possibility of restoring the baseline. People are not irrational here; they are anchored to different starting points and evaluating deviations from those anchors.

That mechanism — reference-dependent evaluation — is the same mechanism behind Prospect Theory, the Endowment Effect, and the Sunk Cost Fallacy. Framing is not a random quirk; it is part of a coherent family of phenomena that all point to the same underlying fact: humans evaluate outcomes relative to a shifting reference, not in absolute terms.

What Tversky and Kahneman Got Right

There are three moves in the original 1981 paper that age extraordinarily well and deserve to be honored by anyone who teaches this material.

The first is the demolition of “revealed preference” as the last word in economics. Before Tversky and Kahneman, mainstream economics treated preferences as latent properties of individuals that could be discovered by watching their choices. The Framing Effect shows that you cannot infer a stable preference from a choice if you do not also specify the frame in which the choice was made. That single insight rewired half of behavioral economics and most of consumer research.

The second is the elegance of the experimental design. A framing study is a minimum-viable cognitive experiment: two groups, one sentence different, one clean behavioral outcome. The paper barely needed statistics — the gap between 72% and 22% is visible to the naked eye. Decades of follow-up work have inherited that clean structure and made framing one of the most replicable classes of effects in psychology, which in an era of replication crises is not a small compliment.

The third is that they refused to treat the effect as an error. Tversky and Kahneman never said “people are stupid.” They said: given how our evaluation system actually works, this is the rational output. The frame changes the reference point; the value function is curved around the reference point; curved functions produce different optimal choices in different regions. The Framing Effect is what rationality looks like when your measuring stick is elastic. That framing of framing — as a feature of human cognition, not a bug — is why the finding generalizes far beyond lab problems.

Where the Framing Effect Falls Apart

The pop-psychology version of framing — “change one word and double your conversion” — is an exaggeration dressed up as science. The real literature is more cautious, and as a designer you need to know where the effect thins out before you build a product around it.

The Effect Size Is Smaller Than You’ve Heard

Anton Kühberger’s landmark 1998 meta-analysis pooled 136 framing studies and found an average effect size of about d = 0.31 — a small-to-medium effect by any honest reading. Piñon and Gambara’s 2005 follow-up meta-analysis found something similar. The Asian Disease numbers (72% vs 22%) are the upper bound of the literature, not the middle. When you hear a speaker quote a single pair of percentages as “the” Framing Effect, they are almost certainly cherry-picking the most dramatic study on the topic.

Effect sizes also vary systematically with the kind of framing. Attribute framing produces the most stable, reliable effects but usually the smallest magnitude. Risky-choice framing produces the largest magnitude but the most variance across replications. Goal framing lives in between. A designer who assumes the same lever will fire at the same strength in every context is going to be disappointed — and worse, is going to attribute the disappointment to their copy rather than to the structure of the effect.

Individual Differences Are Big

Framing is not an equal-opportunity bias. People high in numeracy — the ability to process numerical information fluently — show smaller framing effects, sometimes vanishingly small. People high in need for cognition — a trait measuring how much a person enjoys effortful thinking — show similarly reduced effects. And the effect weakens further when subjects are asked to justify their choice, because justification forces engagement with the underlying numbers rather than the surface frame.

This matters enormously for product design. Your power users, your analytical B2B buyers, your quantitative finance audience — these populations are exactly the people for whom framing fires weakly. If your growth model depends on the Framing Effect carrying the load in those segments, you are going to be outperformed by a product that competes on substance. Framing should be a tiebreaker, not a load-bearing wall.

The Frame Wears Off

Framing effects measured in cross-sectional experiments shrink in longitudinal studies. When subjects re-encounter a decision, reflect on it, or discuss it with others, the initial framing-driven preference tends to drift toward the choice they would have made under a neutral description. This is the “wear-off” problem, and it is catastrophic for any growth strategy that tries to get long-term behavior change from a one-time clever reframe.

In my own consulting work, the teams who over-indexed on clever loss-framed onboarding copy consistently saw week-one metrics beat week-twelve metrics, not because the users hated the product, but because the frame had worn off and the underlying experience was now being evaluated on its merits. If the merits weren’t there, the initial framing bump just sped up the disillusionment curve. A real behavioral system — an Octalysis-designed experience — has to earn the behavior after the frame dissolves. That’s what the Core Drives are for.

The Brain on Framing

The neuroscience of framing is some of the cleanest cognitive neuroscience you can find. In 2006, Benedetto De Martino, Dharshan Kumaran, Ben Seymour, and Ray Dolan published a now-classic fMRI study in Science titled “Frames, Biases, and Rational Decision-Making in the Human Brain.” They replicated the Asian Disease-style reversal in the scanner, and then looked at which brain regions tracked the framing-driven swing and which regions resisted it.

The amygdala lit up with the frame. When a choice was presented as a loss, amygdala activity predicted risk-seeking behavior. When the same choice was presented as a gain, amygdala activity predicted risk-aversion. The amygdala is the brain’s fast, affective, evaluative workhorse — it is doing exactly the work you’d expect the System 1 part of Dual Process Theory to do.

More interesting: subjects who showed less framing effect had greater activation in the anterior cingulate cortex and the orbitofrontal cortex — regions associated with conflict detection and deliberate evaluation. The people who resisted the frame were not cold rationalists; they were people whose deliberative circuitry was catching the amygdala in the act and over-ruling it. This is the biological basis of why need-for-cognition and numeracy reduce framing effects: those traits correlate with stronger deliberative engagement.

For a designer, the takeaway is not “manipulate the amygdala.” It is a far more useful insight: the frame is processed before the content. By the time the user is reading your third paragraph, the amygdala has already made an affective judgment about whether the offer is gain-shaped or loss-shaped, and the rest of their reading is partially a rationalization of that initial verdict. Get the frame right and the content has a chance. Get the frame wrong and the content is fighting gravity.

Framing Effect vs Other Theories

Framing vs Prospect Theory

Prospect Theory is the engine; the Framing Effect is one of its most visible tailpipe emissions. Prospect Theory describes why the same gamble can be attractive or unattractive depending on its reference point; the Framing Effect demonstrates how language alone can shift that reference point. If you understand Prospect Theory, you will correctly predict the direction of almost every framing study. If you only understand framing, you will be reduced to memorizing individual results.

Framing vs Nudge Theory

Thaler and Sunstein’s Nudge Theory treats framing as one of many choice-architecture tools — alongside defaults, salience, simplification, and social norms. Framing is a subtype of nudge; nudging is the broader project of designing choice environments. If you are building a physical sign-up flow or a government form, nudge theory is the right container. If you are writing a single persuasion-heavy paragraph, framing is the right lens. Designers should hold both.

Framing vs Anchoring and Adjustment

Anchoring shifts the numerical starting point of an estimate; framing shifts the valence of how an outcome is evaluated. A price tag of $199 anchors the reference; labeling the same item “Save $50” frames it. These levers combine. A designer who uses an anchor price and then frames the remainder as a “gain” if the user buys, is firing both effects at once. In the research literature they are treated as cousins rather than siblings — distinct but related manifestations of reference dependence.

Framing vs Priming

Priming changes behavior by activating associated concepts (a picture of money makes people act more individualistically, for example). Framing changes behavior by reshaping the logical structure of the decision. Priming has had a rough replication decade; framing has held up far better in meta-analyses. That’s worth remembering when you see a colleague confidently cite an old priming study that hasn’t survived pre-registration — the framing literature is sturdier.

The Framing Effect in the Real World

Healthcare and Patient Decisions

The most consequential applications of framing happen in medicine. McNeil, Pauker, Sox, and Tversky’s 1982 study — a direct application of the framing effect to a clinical decision — showed that both patients and physicians preferred surgery for lung cancer significantly more often when outcomes were described in survival terms (“68% still alive after one year”) than in mortality terms (“32% dead after one year”). The information was identical. The decision was not.

That finding has been replicated across vaccine uptake, cancer screening, medication adherence, and surgical consent. Systematic reviews by Gallagher and Updegraff (2012) found that for detection behaviors (screenings, tests) loss frames tend to outperform gain frames, while for prevention behaviors (diet, exercise, sunscreen) gain frames perform at least as well if not better. Designers of patient-facing experiences should not pick a frame by instinct — the literature is specific enough that you can pick the right one for the behavior in question.

Marketing and Pricing

Every pricing page in software is a framing experiment. “Save $240/year with annual billing” is a gain frame. “Stop losing $20/month by paying monthly” is a loss frame. “Pay $20/month” is, in practice, a third frame — a neutral one that usually underperforms both. In a typical SaaS funnel I’ve audited, the gain-framed version wins in cold-traffic segments where the user hasn’t yet paid anything (no loss to anchor against). The loss-framed version wins among trial users approaching their expiry date, because the “loss” is psychologically real — their work product, their data, their progress will all become inaccessible.

Consumer packaged goods learned this decades ago. “Contains 25% fewer calories” beats “contains 75% of the calories of the original.” “Leaves your skin 80% brighter” beats “leaves 20% of the dullness behind.” The common retail promise “Save more, buy more” is a goal-framed prompt designed to activate loss-avoidance (don’t miss the savings) while maintaining the polite costume of a gain frame. You can read every good piece of direct-response copy as a framing decision wearing a suit.

Public Policy and Political Communication

Tax “relief” is a frame. Estate “tax” vs death “tax” is a frame. “Pro-choice” vs “pro-life” is a frame — two sides of the same debate that have both opted for gain frames about their own position (choice, life) and never lost a single conversation to “anti-choice” or “anti-life” framing. Entire policy debates have been lost by the side that accepted the other side’s frame. George Lakoff’s work on political framing is in many ways a popularization of Tversky and Kahneman’s cognitive mechanism, translated into long-cycle political strategy.

Government health campaigns use framing deliberately. Australia’s “SunSmart” campaigns, Norway’s pension nudges, Singapore’s housing-subsidy messaging — all of these teams explicitly A/B test gain vs loss vs neutral frames before deployment, because a 3% swing in uptake, multiplied by a national population, is measured in lives and billions of dollars.

Product and UX Design

This is the domain I care about most. Every onboarding screen, every empty state, every renewal prompt, every confirmation dialog is a frame. “You have unused credits” vs “You have credits about to expire.” “Only 3 slots left at this price” vs “Save 20% when you reserve by Friday.” “Your account is inactive” vs “Your account is at risk of deletion.” These are not cosmetic copy differences. They are decisions about which Core Drive the product is going to lean on in that moment.

The designers who come out of my Octalysis-Prime courses often arrive thinking about “copy.” They leave thinking about frames. That shift in vocabulary is the shift from being a decorator to being a behavioral architect — and it is exactly what the Framing Effect literature has been arguing for since 1981.

Finance and Investor Behavior

Few domains expose the Framing Effect as honestly as personal finance. A retirement contribution labeled “your company will match $3,000 of free money you’re leaving on the table” outperforms “your company will match up to 6% of contributions” in enrollment rates by double digits — not because the match terms change, but because the loss frame plants a reference point ($3,000 already yours) that converts inaction into a tangible loss. The default-enrollment work that earned Richard Thaler the Nobel Prize in 2017 rests on this exact machinery.

Investor behavior during market drawdowns is the dark side of the same mechanism. A portfolio described as “down 20% from its peak” elicits panic selling. The same portfolio described as “up 45% from where you started three years ago” elicits patience. The math is identical; the frame chooses whether the investor realizes their fear or rides through it. Financial advisors who understand framing don’t just quote numbers — they choose which anchor to make salient, because that choice is what keeps clients from locking in losses at the worst possible moment.

The Elephant in the Room

Framing is the most designer-accessible cognitive bias in the literature. It costs nothing to deploy. It requires no technology, no user data, no A/B framework. You can change the frame of a product’s most important sentence in five minutes. That accessibility is exactly why framing is also the bias most likely to slide into manipulation when a team is under pressure to hit a quarterly number.

The line between persuasion and manipulation is not academic. Persuasion is reframing a real benefit so that the user can see it clearly. Manipulation is reframing a shallow benefit — or an actively harmful offer — so that the user cannot. Framing can do either, and the same paragraph of loss-framed scarcity copy can be the right thing to write for a high-integrity product and the wrong thing for a predatory one. The Framing Effect itself is neutral on the question. You are not.

This is why Octalysis splits the Core Drives into White Hat and Black Hat. Not because Black Hat drives are evil — they are some of the most powerful motivators in the framework — but because they burn trust the longer you leave them on. A product that nudges with loss-framed, Black-Hat-heavy framing can win week one and lose month six when the frame wears off and the user realizes they were optimizing against their own interest. The designer who ignores that trade-off is the one who shows up in a future Dark Patterns blog post.

My own rule: choose the frame that would still be defensible if the user read your internal design document. If the rationale is “we wanted to help them understand the upside of acting early,” loss framing is fair game. If the rationale is “we wanted to scare them into clicking before they finished thinking,” you’ve crossed a line that the Framing Effect, by itself, cannot un-cross.

There is also a second-order concern. A product that frames every interaction as a loss slowly teaches the user’s nervous system that this is a threatening environment. The amygdala research I mentioned earlier cuts both ways: if loss frames reliably activate the amygdala, then a product whose copy is 80% loss-framed is running a small, continuous stress test on its users. Short-term conversion goes up; long-term affinity for the brand craters. The framing literature does not include a “total daily loss-frame dosage” warning, but Octalysis designers should carry one in their heads.

How to Apply the Framing Effect with the Octalysis Framework

The Framing Effect tells a designer what the lever does. Octalysis tells the designer where to pull it and which Core Drive lights up when they do. The mapping between the two is one of the cleanest alignments between academic behavioral economics and practical design I’ve ever worked with, which is why I teach it in almost every corporate workshop I run.

Gain Framing Activates the White Hat Core Drives

The top of the Octalysis octagon — Core Drive 1 (Epic Meaning & Calling), Core Drive 2 (Development & Accomplishment), and Core Drive 3 (Empowerment of Creativity & Feedback) — is where gain-framed design lives. When you frame an experience as “unlock new capabilities,” “progress toward mastery,” or “contribute to something bigger than yourself,” you are doing in Octalysis language what gain-framing does in behavioral economics language. The user evaluates the offer against a reference point of “nothing yet” and sees the action as pure upside.

This is the “build what users pull toward” quadrant of design. It produces long-term engagement, higher NPS, and lower burnout — but it requires more craft, because gain framing has to be earned by a genuine capability expansion. You can’t gain-frame a feature that doesn’t actually give the user anything new. The White Hat is not a free hat.

Loss Framing Activates the Black Hat Core Drives

The bottom of the octagon — Core Drive 6 (Scarcity & Impatience), Core Drive 7 (Unpredictability & Curiosity), and above all Core Drive 8 (Loss & Avoidance) — is the loss-framing quadrant. “Only 3 seats left at this price,” “Your streak is at risk,” “You’ll lose access in 24 hours” — these are not copy decisions, they are behavioral interventions explicitly designed to shift the reference point above the current state, making inaction feel like a descent.

Loss framing’s power is also its hazard. It activates faster, it converts harder, and it erodes trust the longer it stays on. The best Octalysis designs use Black Hat loss framing in pulses — during specific, ethically defensible moments like an onboarding deadline, a real promotional window, or a genuine risk of data loss — and then return the user to a White Hat, gain-framed default for the rest of the experience. Marinating users in loss framing is how you train churn.

Left-Brain vs Right-Brain Framing

The Octalysis left/right split adds a second, often-missed layer to framing decisions. The left side of the octagon — Core Drives 2, 4 (Ownership & Possession), and 6 — is the extrinsic, outcome-oriented side. Framing on this side emphasizes what you’ll get or what you’ll lose. The right side — Core Drives 3, 5 (Social Influence & Relatedness), and 7 — is intrinsic, process-oriented. Framing on this side emphasizes what you’ll experience, who you’ll become, or what you’ll discover.

Most framing research treats gain/loss as the only axis. Octalysis treats extrinsic/intrinsic as the second axis, and the interaction matters. Intrinsic gain framing (“you’ll love the craft of this”) is more durable than extrinsic gain framing (“you’ll earn points”). Extrinsic loss framing (“you’ll lose your rewards”) fires harder than intrinsic loss framing (“you’ll lose the habit”), at least in the short term. Designers who know both axes can pick the framing lever with precision the academic literature doesn’t provide.

Game Techniques That Formalize the Framing Lever

In the Octalysis Game Techniques catalog, several specific techniques are essentially framing patterns wearing a mechanical name:

• GT#21 (Countdown Timer): the purest loss frame — the reference point is “you still have time”; every tick moves you further from that baseline.
• GT#8 (Narrative): the strongest gain frame — a story reframes a series of tasks as a meaningful arc, converting effort into progress.
• GT#12 (Progress Bar): a hybrid frame — the filled portion is a gain frame (“look how far you’ve come”), the unfilled portion is a loss frame (“don’t leave it incomplete”). Good progress bars lean gain; manipulative ones lean loss.
• GT#66 (Status Points): extrinsic gain framing made visible — every point is a small positive frame over the “zero points” reference.
• GT#23 (Scarcity): extrinsic loss framing made visible — the scarcity signal plants a loss reference point and the CTA invites you to avoid it.

A team fluent in both the framing literature and the Game Techniques can decide, for any specific user moment, not just “which copy” but “which Core Drive, in which hemisphere, for how long, at what cost to trust.” That is the practical value of binding Tversky and Kahneman’s work to Octalysis: it turns a clever insight into a design operating system.

Practical Steps to Apply the Framing Effect

If you take one thing away from this piece and put it into your product next Monday, let it be a disciplined framing audit. The following steps are how I run the audit inside Octalysis-Prime consulting engagements.

Step 1: Inventory Your Frames

Pull every piece of user-facing copy into a spreadsheet — onboarding screens, empty states, CTAs, renewal prompts, paywalls, error messages, emails. For each line, label the frame type: gain, loss, or neutral. Most teams are shocked to discover that their copy is 60–80% neutral (“Sign up,” “Continue,” “Learn more”). Neutral copy is a missed opportunity — you’re not preserving choice, you’re just failing to choose.

Step 2: Map Each Frame to a Core Drive

For every gain or loss frame, ask: which Core Drive is this actually firing? A frame that doesn’t map to a Core Drive is either decoration or confusion. If your “upgrade” CTA is loss-framed but none of the Black Hat drives make sense in context — for example, nothing is actually scarce, no streak will break — you’re using the shape of a frame without its substance, and users will sense it.

Step 3: Separate One-Time Frames from Permanent Ones

Loss-framed moments (deadlines, streaks, scarcity alerts) should be temporary and narrative. Gain-framed defaults (progress bars, mastery ladders, capability expansions) should be permanent. If your interface lives in loss frame 24/7, you have turned your product into a slow-acting stress machine. Look at your user’s experience across a full week and check the ratio.

Step 4: Test Your Frames With Your Most Analytical User

Remember that framing effects are smaller in high-numeracy, high-need-for-cognition users. Your sharpest customer is not who the Framing Effect targets — they are who will see through it and tell their peers. Write frames that would not embarrass you if your smartest user printed them out and highlighted them. If a frame would only work on a distracted, low-information user, it is both weaker than you think and more corrosive to your brand than you want.

Step 5: Keep the Frame Honest

Ask the internal-memo test: if a customer FOIAed your product-design Slack channel tomorrow, would the rationale for this frame survive the headline? If yes, deploy. If not, rewrite. This test costs nothing, catches 90% of the manipulation-shaped decisions that slip through a typical design review, and will save you from the particular kind of regret that comes from shipping a growth tactic you can’t defend when a journalist calls.

Closing Thoughts

The Framing Effect is one of those findings that makes you a permanently different kind of designer once you internalize it. Before, copy was a decoration layer laid on top of the “real” product. After, copy is a reference-point-setting device that determines which part of your user’s brain will evaluate the feature in the next three seconds. That is a much bigger responsibility than the traditional “UX copy” job description captures.

And yet framing alone is not a system. It’s a verb, not a strategy. The strategy is Octalysis: a map of what motivates humans at the Core Drive level, with White Hat / Black Hat and Left Brain / Right Brain axes that tell you not just how to frame but where to frame and for how long. Designers who pair Kahneman with Octalysis stop making cosmetic choices and start making architectural ones.

That is, in the end, the whole point of behavioral design. Not to trick humans into behaviors they don’t want. Not to squeeze another 0.3 percentage points out of a funnel by swapping a word. But to build experiences where the frame and the substance reinforce each other — where what we say and what we do are the same, and the user’s best interest and our own are aligned by design. The Framing Effect is one of the oldest tools in that toolkit. It is not the whole toolkit. But used well, it is a beautifully sharp one.

If you want to go deeper on the White Hat / Black Hat axis and the full eight Core Drives that sit behind every frame you’ll ever write, the home is the Octalysis Framework. If you’d rather learn this material the way product teams do, my book Actionable Gamification walks through every Core Drive with live examples, and Octalysis-Prime engagements are how companies run a framing-and-Core-Drive audit on their own product flow. The Framing Effect is one sharp tool; Octalysis is the full bench.

Frequently Asked Questions

What is the Framing Effect in simple terms?

The Framing Effect is the finding that people’s decisions change depending on how an option is described, even when the underlying facts are identical. Describe a choice as a gain and most people play it safe; describe the same choice as a loss and most people take the risk.

Who discovered the Framing Effect?

Amos Tversky and Daniel Kahneman formally introduced the Framing Effect in their 1981 Science paper “The Framing of Decisions and the Psychology of Choice.” It grew directly out of their earlier Prospect Theory work published in 1979.

What is the Asian Disease Problem?

It is the canonical demonstration of the Framing Effect. Subjects are asked to choose between a certain outcome and a probabilistic gamble in response to a hypothetical disease outbreak; 72% picked the certain option when outcomes were described as lives saved, while 78% picked the gamble when the identical outcomes were described as lives lost.

Is the Framing Effect a cognitive bias?

It is usually classified as a cognitive bias, but it is more precisely a feature of reference-dependent evaluation. People are not making a random error; they are evaluating the choice relative to whatever reference point the frame suggests, which is entirely rational given the mechanism.

What are the different types of framing?

Levin, Schneider, and Gaeth (1998) identified three distinct types: risky-choice framing (the Asian Disease pattern), attribute framing (describing a single attribute positively or negatively, such as “95% lean” vs “5% fat”), and goal framing (emphasizing the consequences of acting versus not acting).

How big is the Framing Effect in practice?

Meta-analyses put the average effect in the small-to-medium range (Kühberger 1998, d ≈ 0.31). The dramatic numbers quoted in popular articles usually come from the Asian Disease Problem itself, which is the upper end of the literature. In product design, expect real-world swings of a few percentage points on conversion, not order-of-magnitude flips.

Can the Framing Effect be resisted?

Yes. People high in numeracy, need for cognition, or those asked to justify their choice show significantly smaller framing effects. Reflection, peer discussion, and repeated exposure to the same decision also attenuate the effect over time.

How does the Framing Effect relate to Prospect Theory?

Prospect Theory is the underlying mechanism; the Framing Effect is one of its most visible behavioral consequences. Prospect Theory’s kinked value function — concave in gains, convex in losses, steeper for losses — explains why a verbal reframe can flip a choice.

What’s the difference between framing and nudging?

Framing is a tool; nudging is the practice of using tools like framing, defaults, social norms, and salience to guide choices while preserving freedom. Framing lives inside nudge theory as one of its most accessible instruments.

How do Octalysis designers use the Framing Effect?

Octalysis designers map gain framing to the White Hat Core Drives (Epic Meaning, Development, Creativity) for sustainable long-term engagement, and loss framing to the Black Hat Core Drives (Scarcity, Unpredictability, Loss & Avoidance) for short, pulsed moments of urgency. The framework prevents loss framing from becoming a permanent mode, which is how manipulation starts.

References

• Tversky, A., & Kahneman, D. (1981). The framing of decisions and the psychology of choice. Science, 211(4481), 453–458. Science link.
• Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263–291.
• Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux.
• Levin, I. P., Schneider, S. L., & Gaeth, G. J. (1998). All frames are not created equal: A typology and critical analysis of framing effects. Organizational Behavior and Human Decision Processes, 76(2), 149–188.
• Kühberger, A. (1998). The influence of framing on risky decisions: A meta-analysis. Organizational Behavior and Human Decision Processes, 75(1), 23–55.
• Piñon, A., & Gambara, H. (2005). A meta-analytic review of framing effect: risky, attribute and goal framing. Psicothema, 17(2), 325–331.
• McNeil, B. J., Pauker, S. G., Sox, H. C., & Tversky, A. (1982). On the elicitation of preferences for alternative therapies. New England Journal of Medicine, 306(21), 1259–1262.
• De Martino, B., Kumaran, D., Seymour, B., & Dolan, R. J. (2006). Frames, biases, and rational decision-making in the human brain. Science, 313(5787), 684–687.
• Gallagher, K. M., & Updegraff, J. A. (2012). Health message framing effects on attitudes, intentions, and behavior: A meta-analytic review. Annals of Behavioral Medicine, 43(1), 101–116.
• Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improving Decisions About Health, Wealth, and Happiness. Yale University Press.
• Lakoff, G. (2004). Don’t Think of an Elephant! Know Your Values and Frame the Debate. Chelsea Green Publishing.
• Rothman, A. J., & Salovey, P. (1997). Shaping perceptions to motivate healthy behavior: The role of message framing. Psychological Bulletin, 121(1), 3–19.
• Peters, E., Västfjäll, D., Slovic, P., Mertz, C. K., Mazzocco, K., & Dickert, S. (2006). Numeracy and decision making. Psychological Science, 17(5), 407–413.
• Chou, Y. (2015). Actionable Gamification: Beyond Points, Badges, and Leaderboards. Octalysis Media.
• Camerer, C. F., Loewenstein, G., & Rabin, M. (Eds.). (2004). Advances in Behavioral Economics. Princeton University Press.

Related Reading

• Prospect Theory — the engine behind the Framing Effect
• Dual Process Theory — why frames hit System 1 before System 2
• Nudge Theory — framing’s broader choice-architecture container
• Paradox of Choice — when too many frames overload the decision
• The Octalysis Framework — Core Drives as the scaffolding for every frame

WOULD YOU LIKE YU-KAI CHOU TO WORK WITH YOUR ORGANIZATION?