Every professional describes the same number two ways in a week. A ninety-five percent success rate and a one-in-twenty failure rate. A discount for paying early and a charge for paying late. There is a substantial literature on whether that matters, and it says something more specific than the popular version.

Key Takeaway

A 1998 meta-analysis of 136 studies of risky choice framing reports a mean d = 0.31. A 2005 meta-analytic review covering risky, attribute and goal framing across 230 effect sizes reports d = 0.31[1]. But a 1998 typology paper is titled All frames are not created equal[2], and a 2002 paper's own title asserts the independence of types of effects[3].

Our Grades For These Claims

Applying the scheme from the first article in this series.

Grade A that framing effects exist. Two independent meta-analyses covering hundreds of effect sizes, converging on the same magnitude, across a literature spanning four decades.

Grade A that the average magnitude is moderate, at d = 0.31 on both.

Grade B for the three-type typology, which is a peer-reviewed framework rather than a finding, and Grade B for the independence of the types.

Grade C for the reinterpretation of risky choice framing discussed below, which rests on a single paper we obtained only in fragment.

Our position: this is one of the better-established effects in the series, and it is routinely applied as though it were three times more general than it is.

A Note On Method

Everything here is verified to August 2026.

We did not obtain any of the papers discussed in full. We have titles, citations, and in several cases quoted fragments, from reference lists and publisher records.

The two effect size figures reach us through a table in a working paper that derived a range from published meta-analyses[1]. We flag that at each use. We did not obtain either meta-analysis and cannot report their confidence intervals, moderators, or bias checks.

Several claims rest on paper titles alone, which we mark explicitly where they occur, because a title is weaker evidence than a finding.

All conversions of effect sizes are ours.

This article reviews behavioural research. It is not marketing, communications or legal advice. How prices and fees may be presented to consumers is governed by law this article does not address.

What A Framing Effect Is

The claim, in a form that makes clear why it is interesting.

A peer-reviewed source states: "Framing effects are said to indicate irrationality in decision making because they illustrate that linguistically different descriptions of equivalent options lead to inconsistent choices."[4]

Two observations, ours.

The word doing the work is equivalent. The two descriptions must convey identical information for the effect to be a failure of reasoning rather than a response to different content.

And that condition is harder to satisfy than it looks. Two descriptions that are logically equivalent may not be equivalent as communications, because the choice of framing can itself carry information about what the speaker considers salient. The twenty-seventh article in this series found exactly that reading applied to defaults, where the setting was perceived as advice.

Two Meta-Analyses, One Number

The aggregate evidence.

A table in a working paper, deriving a range from published meta-analyses, records two entries[1].

Kühberger (1998), on risky choice framing, k = 136, mean d = 0.31.

Piñon and Gambara (2005), covering risky, attribute and goal framing, 230 effect sizes, d = 0.31.

The full citations, from reference lists: Kühberger, A. (1998), The influence of framing on risky decisions: A meta-analysis, Organizational Behavior and Human Decision Processes, 75(1), 23–55, DOI 10.1006/obhd.1998.2781[4]; and Piñon, A., & Gambara, H. (2005), A meta-analytic review of framing effect: Risky, attribute and goal framing, Psicothema, 17(2), 325–331[5].

Two observations, ours.

Two teams, seven years apart, with different scopes, arriving at the same figure to two decimal places. That is unusually clean convergence for this field, and this series has more often found the opposite.

We flag firmly that both figures reach us through a third party's summary table, not from either paper. A reader depending on the number should obtain the originals.

How Large That Is

Placing the number, with our own conversion.

At d = 0.31, a framed respondent lands above the average unframed one about 62 percent of the time against 50 percent by chance, with roughly 88 percent overlap between the two distributions. Both figures are ours, computed under standard normal assumptions.

Against what this series has reported: brainstorming penalty 1.39, surface acting and wellbeing 0.85 to 1.09, the disputed nudge headline 0.43, feedback interventions 0.41, framing 0.31, growth mindset 0.08, choice overload 0.02, actor-observer asymmetry between -0.016 and 0.095.

Two observations, ours.

Framing sits squarely in the middle. It is real, replicated, and considerably smaller than the way it is usually described in business writing, where it is often presented as a lever that reverses decisions.

And 88 percent overlap is the number to keep. Most people give the same answer either way. The effect moves a minority, which is enough to matter commercially at scale and is not the same as changing everybody's mind.

The Typology

The paper that should have settled how this phrase is used.

Levin, Schneider and Gaeth published All frames are not created equal: A typology and critical analysis of framing effects in Organizational Behavior and Human Decision Processes, 76, 149–188, in 1998[2].

A later paper describes what it did: it set out three types along with their operational definitions, their typical results, and the likely underlying processes[2].

We did not obtain this paper and describe the three types from how the literature uses them, which we flag as our own characterisation rather than the paper's definitions.

Two observations, ours.

The title is an argument, in the manner of the 2004 paper this series encountered arguing the fundamental attribution error should be retired. All frames are not created equal is a claim that the field had been treating distinct things as one.

And likely underlying processes, in the plural, is the substantive point. If the three types run on different mechanisms, then evidence about one transfers to another only by argument, not automatically.

One: Risky Choice Framing

The famous one, and the one least relevant to most businesses.

This is the Asian disease problem, from Tversky and Kahneman (1981), The framing of decisions and the psychology of choice, Science, 211(4481), 453–458[5]. We did not obtain it.

The pattern, as a peer-reviewed source describes it: prospect theory suggests that people are risk-averse when facing gains, but risk-prone when facing losses, a pattern known as the framing effect[6].

Three observations, ours.

The outcome variable is risk preference: whether someone takes the sure thing or the gamble. That is a specific and unusual decision structure.

Most commercial communication is not a choice between a certainty and a gamble with matched expected values. Presenting a fee, a rate or a result is not this problem.

And note the same source calls this pattern the framing effect, in the singular, which is precisely the conflation the typology paper objects to.

Two: Attribute Framing

The one that actually applies to most businesses.

The canonical study is Levin and Gaeth (1988), How consumers are affected by the framing of attribute information before and after consuming the product, Journal of Consumer Research, 15(3), 374–378[5]. We did not obtain it and report only its title, which describes the design.

Our own characterisation of the type, flagged as ours: a single attribute of a single thing is described in positive or negative terms, and the evaluation moves. There is no risk, no gamble, and no choice between options.

Three observations.

This is structurally simpler than risky choice framing, and it is what almost every business is actually doing when it chooses wording.

The title's phrase before and after consuming the product is worth noting. The design apparently tested whether the framing survived direct experience, which is the question a business should care about most.

And we do not know what it found on that point, because we did not obtain it. We flag that as a gap in exactly the place a reader would most want an answer.

Three: Goal Framing

The third type, which concerns persuasion rather than description.

Our own characterisation: a message emphasises either the gain from acting or the loss from not acting, and the question is which produces more of the behaviour.

Reference lists identify the relevant literature: Meyerowitz and Chaiken (1987) on the effect of message framing on breast self-examination attitudes, intentions, and behavior, in JPSP; and Rothman and Salovey (1997), Shaping perceptions to motivate health behavior: The role of message framing, in Psychological Bulletin, 121, 3–19[3].

We obtained neither and report titles and citations only.

Two observations, ours.

This literature is predominantly health behaviour, which is a different setting from commercial persuasion with different stakes and different audiences.

And it is the type where the twelfth article in this series is most relevant: a message that works on one recipient may work in reverse on another, and an aggregate effect can conceal that entirely.

The Independence Finding

The result that makes the typology consequential rather than tidy.

Levin, Gaeth, Schreiber and Lauriola published A new look at framing effects: Distribution of effect sizes, individual differences, and independence of types of effects in Organizational Behavior and Human Decision Processes, 88, 411–429, in 2002[3].

We did not obtain this paper. Its title asserts three things: a distribution of effect sizes, individual differences, and the independence of types of effects. We are reporting a title, which is weaker evidence than a finding, and we say so.

Two observations, ours, conditional on the title meaning what it appears to mean.

Independence would mean susceptibility to one type does not predict susceptibility to another. A person moved by attribute framing is not thereby a person moved by risky choice framing.

And distribution of effect sizes, rather than a single average, points at the same issue this series has raised repeatedly: an average across heterogeneous studies describes a population that may not exist.

Why That Matters More Than The Size

The practical consequence. Ours.

Almost every business application of framing cites the Asian disease problem and then gives advice about wording a fee, a benefit or a result.

Three consequences if the types are independent.

That inference is invalid. Evidence about how people choose between a certainty and a gamble does not establish how they evaluate a described attribute.

The relevant evidence for a business is attribute framing, which has its own literature and its own effect sizes, and which most popular treatments never mention by name.

And the d = 0.31 figure is being asked to do too much. One of the two meta-analyses covers risky choice framing only; the other pools all three. Neither gives you the attribute framing figure on its own, and we do not have it.

A Reinterpretation Of The Famous One

A finding that complicates the standard account, reported from a fragment.

A publisher record for Kühberger and Gradl (2013), Choice, rating, and ranking: Framing effects with different response modes, in the Journal of Behavioral Decision Making, 26, 109–117, contains: "The rating and ranking findings indicate that what is commonly called a risky choice framing effect is actually a framing effect that changes the evalua" and our source truncates there[7].

We did not obtain the paper, the sentence is incomplete, and we do not complete it.

Two observations, ours, offered cautiously.

The fragment says the effect changes the evaluation of something, rather than changing risk preference. If that is the argument, then the canonical framing effect is not about attitude to risk at all, and the prospect theory account of it would be wrong even though the phenomenon is real.

And it would move risky choice framing closer to attribute framing, which would be a substantial rearrangement of the typology. We flag that this is our inference from an incomplete sentence and should be treated as such.

The Monkeys

A datum worth recording for what it rules out.

A reference list identifies Lakshminarayanan, V. R., Chen, M. K., and Santos, L. R. (2011), The evolution of decision-making under risk: framing effects in monkey risk preferences, in the Journal of Experimental Social Psychology, 47, 689–693[8].

We did not obtain this paper and report only its title.

Two observations, ours.

If framing effects occur in a species without language, then the standard explanation, that people respond to linguistically different descriptions, cannot be the whole story.

And it cuts against the most common dismissal of framing research, which is that participants are simply reading a communicative intent into the experimenter's wording. That explanation is unavailable for monkeys, and its unavailability is evidence for something more basic going on.

Which One Matters To A Business

The practical sorting. Ours.

Three questions to identify which type you are actually dealing with.

Is there a gamble? If the person is choosing between a certain outcome and a probabilistic one, it is risky choice framing. Most commercial decisions are not this.

Is there one thing being described? If you are describing a single option's attribute more or less favourably, it is attribute framing. This is what a fee schedule, a success rate or a performance figure is.

Are you trying to cause a behaviour? If the message urges action by stressing gains from acting or losses from not acting, it is goal framing.

Two observations.

Almost all professional communication falls into the second category, and occasionally the third. The first is rare outside insurance, investment and litigation decisions.

And the evidence a business needs is therefore not the evidence it is usually shown.

The Surcharge And The Discount

The clearest commercial case of attribute framing. This section is our own reasoning and is not tested by anything cited.

The same price difference can be presented as a surcharge for one payment method or a discount for another. Arithmetically identical; described oppositely.

Three observations.

On the attribute framing pattern, those two presentations would not be expected to produce identical evaluations, even though they produce identical invoices.

The choice between them is not free of consequence and is usually made without thought, in the same way the twenty-seventh article found invoice defaults inherited rather than decided.

And the legal position matters here and is not addressed by this article. How a business may present price differences by payment method is governed by card network rules, contractual terms and consumer protection law, which differ by jurisdiction and by network. That constrains the choice before any behavioural consideration arises.

The Constraint That Makes This Usable

The line we would draw. Ours.

The definition quoted at the top of this article requires the two descriptions to be equivalent. That is the whole ethical content of the topic.

Three consequences.

Choosing between two accurate descriptions of the same fact is unavoidable. Every sentence has to be written somehow, and there is no neutral wording that describes a ninety-five percent rate without also being a description of a five percent rate.

So the question is not whether to frame, which is not optional, but whether the frame chosen would survive the reader learning the other one.

And that gives a test we would apply: would the client, shown both descriptions side by side, feel they had been told the same thing? If yes, the choice is presentation. If no, something has been omitted rather than framed, and the effect being relied on is not a framing effect at all.

In A Professional Firm

Where this arises daily. Ours, untested.

Four routine cases.

A success rate. Positions sustained on review, or positions challenged. Same file, two numbers.

A fee. An hourly rate, a fixed quote, or a percentage of a recovery, each of which describes the same expected cost differently.

A tax outcome. Tax saved against tax paid, or an effective rate against a marginal one.

And a risk assessment, where a position described as defensible and a position described as carrying audit exposure may be the same position.

Two observations.

In every case the professional obligation points the same way as the honest test above: the client is entitled to the description that leaves them able to make their own decision, which frequently means giving both.

And giving both is the one option the framing literature does not cover, because every study we located compares one frame against another rather than either against both.

What To Do

Identify which of the three types you are in. Risky choice framing involves a gamble, attribute framing describes one thing, goal framing urges an action. Most business communication is the second.

Stop citing the Asian disease problem for attribute decisions. On a paper title asserting the independence of the types, evidence about one does not transfer automatically to another.

Hold the size in mind. Two meta-analyses report d = 0.31, which on our conversion leaves about 88 percent overlap. Most people answer the same way either way.

Notice that you are framing whether you decide to or not. There is no neutral way to state a rate, so the choice is which accurate description to lead with.

Apply the side-by-side test. Would the client, shown both descriptions together, feel they had been told the same thing? If not, the issue is omission rather than framing.

Consider giving both. It is the option the literature does not study, and in a professional context it is frequently the one the obligation points to.

Check the legal position before the behavioural one. How price differences by payment method may be described is governed by network rules and consumer law that vary by jurisdiction.

Do not assume framing survives experience. The canonical attribute framing study's title suggests it tested exactly that, and we could not obtain what it found.

The Limits Of This Analysis

Several caveats matter. This article reviews behavioural research and is not marketing, communications or legal advice; how prices and fees may be presented to consumers is governed by law, card network rules and contractual terms not addressed here. Everything is verified to August 2026. We did not obtain any of the papers discussed in full, and several claims in this article rest on paper titles alone, which we mark at each occurrence. Both effect size figures reach us through a summary table in a working paper, not from either meta-analysis; we cannot report their confidence intervals, moderators, or publication bias checks, and a reader depending on the figure should obtain the originals. We did not obtain the typology paper, and our descriptions of the three types are our own characterisations from how the literature uses them rather than that paper's definitions. The independence claim rests on a title. The reinterpretation of risky choice framing rests on a sentence that truncates mid-word in our source, and our reading of it is explicitly marked as inference. We did not obtain the attribute framing study, the goal framing literature, the monkey study, or the original 1981 paper, and report titles and citations only. We have no separate effect size for attribute framing, which is the type most relevant to business, and say so in the body. All conversions of effect sizes are ours. The sorting questions, the surcharge discussion, the side-by-side test and the professional firm section are our own reasoning, untested.

Behavioural Finance None Of The Economics Changed Where a presentation choice was read as a recommendation rather than a neutral setting. Behavioural Finance Ninety-Nine The other case where how a number is written turns out to be a commercial decision. No. 29 They Do Not Accumulate Expectations do shape outcomes, the definitive review concludes, but the effects are small and may be more likely to dissipate than accumulate. And a...

Frequently Asked Questions

Are framing effects real?
Yes, and unusually for this series two independent meta-analyses seven years apart, covering different scopes, both report d = 0.31. It is one of the better-established effects covered here.
How big is that in practice?
Moderate. On our own conversion it leaves roughly 88 percent overlap between the framed and unframed distributions, meaning most people answer the same way either way. It moves a minority, which matters at scale and is not the same as reversing decisions.
What are the three types?
Risky choice framing, where gains and losses shift preference between a certainty and a gamble; attribute framing, where one feature of one thing is described more or less favourably; and goal framing, where a message stresses the gain from acting or the loss from not acting. A 2002 paper's title asserts these are independent of each other.
Which applies to my business?
Almost certainly attribute framing, occasionally goal framing. Risky choice framing requires a gamble against a certainty, which is rare outside insurance, investment and litigation decisions. Yet the Asian disease problem is what business writing almost always cites.
Is choosing a frame manipulative?
Framing is not optional. There is no neutral way to state a rate, since every description of ninety-five percent is also a description of five percent. Our own test is whether the client, shown both descriptions side by side, would feel they had been told the same thing. If not, something has been omitted rather than framed.
What about surcharges and discounts?
Arithmetically identical, described oppositely, and a textbook attribute framing case. But the legal position comes first: how price differences by payment method may be presented is governed by card network rules, contractual terms and consumer law that vary by jurisdiction, none of which this article addresses.
IB

About The Insight Bureau Research Desk

The Insight Bureau is GSH Financial's research publication, written for Canadian business owners and the students who will eventually advise them. This article reports that it does not have the one effect size a business reader would most want, and says so in the body rather than only in the footnotes.

References

  1. Working paper deriving a human framing effect-size range from published meta-analyses, table recording Kühberger (1998) on risky choice framing with k = 136 and mean d = 0.31, and Piñon and Gambara (2005) on framing across risky, attribute and goal types with 230 effect sizes and d = 0.31; and noting that the lower bound of the derived range reflects population-level averages across hundreds of framing studies while the upper bound reflects individual high-stakes studies. Note: a working paper's summary table. This is our sole source for both effect sizes; we obtained neither meta-analysis and cannot report confidence intervals, moderators or bias checks. arxiv.org
  2. Reference lists and a hosted paper identifying 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, 149–188; and describing that paper as setting out the types along with their operational definitions, their typical results, and the likely underlying processes. Note: we did not obtain the typology paper. Our descriptions of the three types are our own characterisations from how the literature uses them, not that paper's definitions. communicationcache.com
  3. Publisher record and reference list for Levin, I. P., Gaeth, G. J., Schreiber, J., & Lauriola, M. (2002), A new look at framing effects: Distribution of effect sizes, individual differences, and independence of types of effects, Organizational Behavior and Human Decision Processes, 88, 411–429; together with citations for Meyerowitz, B. E., & Chaiken, S. (1987), on the effect of message framing on breast self-examination attitudes, intentions, and behavior, Journal of Personality and Social Psychology; and Rothman, A. J., & Salovey, P. (1997), Shaping perceptions to motivate health behavior: The role of message framing, Psychological Bulletin, 121, 3–19. Note: we did not obtain any of these papers. The independence claim in this article rests on the 2002 paper's title alone, which is weaker evidence than a finding. researchgate.net
  4. Publisher record for Kühberger, A., & Tanner, C. (2010), Risky choice framing: Task versions and a comparison of prospect theory and fuzzy-trace theory, Journal of Behavioral Decision Making, on framing effects being said to indicate irrationality in decision making because they illustrate that linguistically different descriptions of equivalent options lead to inconsistent choices; and its reference list confirming Kühberger, A. (1998), The influence of framing on risky decisions: A meta-analysis, Organizational Behavior and Human Decision Processes, 75, 23–55, and Kühberger, A., Schulte-Mecklenbeck, M., & Perner, J. (1999), The effects of framing, reflection, probability, and payoff on risk preference in choice tasks, Organizational Behavior and Human Decision Processes, 78, 204–231. Note: a publisher record; we obtained the abstract opening and reference list. onlinelibrary.wiley.com
  5. Reference list in a peer-reviewed journal article on managerial judgement, identifying Tversky, A., & Kahneman, D. (1981), The framing of decisions and the psychology of choice, Science, 211(4481), 453–458, DOI 10.1126/science.7455683; Levin, I. P., & Gaeth, G. J. (1988), How consumers are affected by the framing of attribute information before and after consuming the product, Journal of Consumer Research, 15(3), 374–378; and Piñon, A., & Gambara, H. (2005), A meta-analytic review of framing effect: Risky, attribute and goal framing, Psicothema, 17(2), 325–331. Note: citations only. We obtained none of these papers and report no findings from any of them. eujournal.org
  6. Publisher record for Mishra (2012), Framing effects and risk-sensitive decision making, British Journal of Psychology, on prospect theory suggesting that people are risk-averse when facing gains but risk-prone when facing losses, a pattern known as the framing effect. Note: a publisher record; we obtained the abstract opening only. bpspsychub.onlinelibrary.wiley.com
  7. Publisher record for Kühberger, A., & Gradl, P. (2013), Choice, Rating, and Ranking: Framing Effects with Different Response Modes, Journal of Behavioral Decision Making, 26, 109–117, containing the statement that the rating and ranking findings indicate that what is commonly called a risky choice framing effect is actually a framing effect that changes the evaluation of something, our source truncating mid-word. Note: we did not obtain the paper and the sentence is incomplete in our source. We do not complete it, and our reading of its implication is marked in the body as inference. onlinelibrary.wiley.com
  8. Reference list in a peer-reviewed journal, identifying Lakshminarayanan, V. R., Chen, M. K., & Santos, L. R. (2011), The evolution of decision-making under risk: framing effects in monkey risk preferences, Journal of Experimental Social Psychology, 47, 689–693. Note: a citation only. We did not obtain this paper and report only its title; the inference drawn from it in the body is our own. cambridge.org

This article reviews behavioural research and is not marketing, communications or legal advice. How prices and fees may be presented to consumers is governed by law, card network rules and contractual terms not addressed here. No paper discussed was obtained in full, and several claims rest on paper titles alone. Both effect size figures reach this article through a third party's summary table rather than from either meta-analysis. No separate effect size for attribute framing, the type most relevant to business, was located.