Framing
[Last Updated January 21st 2024]
Framing is one of the most important concepts in marketing, user experience design, and decision making research. It is the idea that how information is presented will influence an individual’s perception, beliefs, assumptions, and behaviours. That is to say that the same information presented in different ways will elicit different cognitive responses. In behavioural science, there are many studies examining the framing effect, which involves decisions between choices and often involves framing in terms of gains/positives and losses/negatives (see Tversky & Kahneman, 1981). For example, Gächter et al. (2009) ran a study where they reminded researchers (mostly junior and senior economists) to register for a conference. Half the participants were told there was a $50 discount if they registered on time, while the other half were told there was a penalty of $50 if they registered late. In both situations, conference-goers would end up spending $50 more if they did not register on time. Therefore, the only difference between the two groups was how the cost was framed. For senior researchers, there did not seem to be any statistically significant difference between phrasing. But for junior researchers, there was a statistically significant difference, with only 67% of individuals in the discount condition signing up on time, while 93% in the penalty condition signed up on time. This result is consistent with loss aversion, which suggests that losses are more salient than gains. Further, this study demonstrates that framing effects differ across demographics, and are often not universal. For example, Best & Charness (2015) found generalized context-dependent age differences between younger and older adults in a meta-analysis of various framing studies.
It is important to recognize that in addition to demographic concerns, context can influence the effectiveness of framing techniques. For example, Levin et al. (1998) describe three different contexts or “types” of framing experiments and review many studies within each domain finding a specific common pattern for each separate context/type. They describe these contexts as risky choice framing, attribute framing, and goal framing. Risky choice framing (see “The Asian Disease Problem” in our Research Examples section on this page) involves a choice between a certain outcome or a risky outcome. Levin et al. (1998) found that in this context, a positive frame generally leads to preference for certainty, while a negative frame leads to preference for risk. A relatively recent meta-analysis by Steiger & Kühberger (2018) has found that this framing effect in respect to risky choice is quite robust and exists across various diverse studies. Attribute framing (see “Deciding on Surgery” in our Research Examples section on this page) is described by Levin et al. (1998) as the simplest form of framing as it involves a single direct attribute (e.g. the performance of a hockey player) being framed as either positive (e.g. shots on net scored) or negative (e.g. shots on net missed). They find that in this context, an attribute (e.g. the performance of a hockey player) is judged to be more favorable if framed as a positive (e.g. shots on net scored). A recent meta-analysis by Dolgopolova et al. (2021) found that for food products, gain/positive frames (e.g. 70% fat-free) in respect to attributes (e.g. healthiness of chocolate) do indeed seem to beneficially influence attitudes and preference more so than a loss/negative frame (e.g. 25% fat) across studies. Levin et al. (1998) describe goal framing (see “Mouth Wash Advertisements” in our Research Examples section on this page) as involving how often one chooses to carry out a behaviour, that is either justified through a positive frame that describes a positive consequence or gain (e.g. “eating healthy may prevent diabetes”) or through a negative frame that describes a negative consequence or loss (e.g. “not eating healthy may cause diabetes”). Levin et al. (1998) find that in this context, a negative frame that focuses on losses or negative consequences impacts behaviour more so than a positive frame. Consistent with this finding, a meta-analysis by Ainiwaer et al. (2021) suggests that loss-framed messages have a positive impact on cancer prevention/detection behaviours. A meta-analysis by Gallagher and Updegraff (2012) on the other hand suggest that gain framed messages are more effective towards prevention behaviours in the health domain (including cancer prevention). And a meta-analysis from O’Keefe and Jensen (2007) found little difference between gain and loss framed messages in respect to various health-related behaviours. Thus, when designing strategies that take advantage of goal framing, it may be wise to assume experimentation is necessary. This may be due to the motivation, effort, and time investment associated with active behaviour. Whereas attitudes and beliefs can form quickly and subconsciously, behaviours require cognitive effort which can be influenced by numerous context-dependent variables and motivational drives.
Specific social contexts may influence the framing effect as well. For example, Xu & Huang (2020) conducted a meta-analysis that found gain or loss frames have minimal effect on charity advertising. It may be wise to search Google Scholar for framing studies that are relevant to your industry, business, or goals, before attempting to formulate a strategy. Further, other researchers classify framing experiments in different ways, and framing is not limited to the domain of positives/gains and negatives/losses (e.g. the framing effect).
Framing in general is a technique that can be used in many different scenarios alongside other behaviour based strategies to influence how individuals perceive and act on information. For example, if you are selling a new AI chatbot to help with customer service, you could frame it as an additional helper for when human customer service representatives are busy. Or alternatively, you could frame it as a replacement for humans. Or you could frame it in terms of efficiency, and emphasize how it can help answer easy questions and only forward the user to a human agent when necessary. Additionally, you could frame it as a better service than existing competitor customer service chat bots. Or you could frame it as “the future of customer service.” All these descriptions may be accurate at the same time. But the way you frame your product/service will influence how potential customers view your brand and engage with your advertisements or websites. Often, you will find that different target markets respond in different ways to different frames. Thus, it can be helpful to use landing pages and separate ads for different target markets, which can allow you to frame your initial introduction to your brand in a way that is congruent with the needs of each group. Within this process, you want to determine if you are asking individuals to consider choices in regard to risk (e.g. if you are selling computer security software or insurance). If so, then you may want to consider looking into risky choice framing, and determining whether a loss frame or gain frame would be better. On the other hand, if you are attempting to enhance the attractiveness of your product/service, you may want to stick to a positive frame. And if you are trying to influence behaviour, you likely want to do research into your industry in respect to which type of frame might work best, and then experiment extensively. It is also helpful to consider all behavioural science strategies when developing a strategy for framing, as many of them will allow you to make decisions that significantly affect perceptions, beliefs, and behaviour. For example, if you are familiar with the default effect and importance of social norms, you might recognize that requiring individuals to opt out of organ donation on a form leads to significantly more individuals willing to donate organs than asking individuals to opt in. For example, Johnson and Goldstein (2003) provide a chart showing that countries using an opt-in method range from a 4.25% to 27% consent rate for organ donation, whereas countries with an opt-out method range from 85.9% to 99.98%. How you frame content, or even questions, can have a significant impact on your brand and user experience, and should always be kept in mind.
Framing can easily be used to mislead individuals about your product/service or brand. We are against this and believe that honesty is always important. Framing research can help you better choose what kind of messaging is ideal for conversion, growth, or a better user experience. But you should always endeavor to be truthful with the information you are providing. Put another way, framing can help you create strategies that allow you to communicate more effectively in respect to what you are offering. But you should not use it to trick people into believing something about your brand that is not true. False advertising is regulated in most countries and can lead to fines, litigation, and criminal charges. Please use framing strategies responsibly.
Imagine you are selling an imperfect product, and are required by your government to disclose its effectiveness. This would be representative of attribute framing, whereby research suggests that you would benefit most from focusing on the positives, which in this case is the success rate (Levin et al., 1998). For example, if selling condoms, it would be better to frame them as having a 90% success rate than a 10% failure rate (for unpublished data supporting this, see Linville et al., 1993). Similarly, a drug with an 80% success rate would sound more attractive than a drug with a 20% failure rate. And a car that performed above average in safety testing 95% of the time sounds much safer than a car that performed below average on 5% of tests. But even if you are required to disclose failures, you can still benefit from framing by using words like “only.” For example, you could describe a car as “only performing below average on 5% of tests” which primes consumers to interpret the 5% as a small number relative to the norm.
We don’t suggest you do this for ethical reasons. But we want to include this example as it is representative of a common practice in drug advertisements. When advertising a drug, you can frame effectiveness as a relative risk in order to enhance its attractiveness. Consider the fictitious detrimental condition fakeothitus. Let us pretend the average chance of getting fakeothitus in a lifetime is 4%. If you market a drug that reduces the risk of fakeothitus to 2%, rather than describing it as a “2% reduction” you can say “reduces risk by 50%.” In this situation, there is an absolute risk reduction of 2% meaning that the difference between the original risk (4%) and the risk when taking the drug (2%) is 4% - 2% = 2%. However, there is a relative risk of 50%, meaning that if you compared one group that is taking the drug (2%), to another group that is not taking the drug (4%), you can divide 0.02 by 0.04 to get 0.5 (50%). Here, the relative risk reduction is 50%, meaning that the group taking the drug died 50% less often. 50% sounds a lot more impressive than 2%. But to better understand this, lets make the absolute difference much smaller. Imagine now that the risk of fakeothitus over the lifetime is only 0.2%, and a drug reduces this to 0.1%. Here, the absolute risk reduction is a very small 0.1%. Yet, the relative risk reduction is still 50%. Now ask yourself this: Would you spend $500 on a drug that reduces your chance of getting sick by 0.1%? What about if you thought it reduced your chance by 50%? Hopefully, you can see why framing might have a large impact on consumer decision making. Especially if the consumer isn’t aware of how risk can be communicated. If you would like to read a study that tested how likely patients were to take medication based on this type of framing, please see Malenka et al. (1993). Further, Lexchin (1999) looked at how drug companies advertise to doctors in a number of Canadian medical journals in the 1990s, and found that the majority of ads presented facts in terms of relative risk reduction. Of course, the benefits of communicating relative risk compared to absolute risk are not limited to the medical field. This can be used to sell any service/product that includes risk management, such as safety equipment, security software, scientific technology, financial products, etc. However, we suggest you avoid reporting only the relative risk reduction and attempt to be honest about your products/services. Not only is this more ethical, but it creates a better relationship with your customer and helps avoid negative reviews from individuals who feel scammed. Further, there may be regulations in your country that prevent using this strategy, and even if not directly mentioned, it could constitute a form of false advertising.
(Note: this study is also described on our Loss Aversion page)
The most famous example of how loss aversion influences our decision making comes from a series of studies by Tversky and Kahneman (1981). In one of these studies they split participants into two groups. Both groups were told to imagine the U.S. was preparing for a disease outbreak which could kill 600 people. The first group was asked to choose between Program A and B. Program A would save 200 people, while program B would have a 33.33% chance that 600 would be saved, but a 66.66% chance that no one would be saved. In this scenario, 72% chose option A, where 200 people would be saved. They then presented the second group with reworded (reframed) versions of both programs. Program C was described as leading to 400 people dying, and program D was described as a 33.33% chance nobody would die, and a 66.66% chance that 600 would die. Here, 78% of people chose program D. Program A and B were identical to C and D from a mathematical perspective in respect to life and death. But because group 1 saw options framed as gains (saving lives) and group 2 saw options framed as losses (people dying) participants from each group on average made different decisions. A loss frame led to a willingness to take greater risks (to avoid a loss), whereas a gain frame led individuals to stick with certainty. Thus, if you want customers to take risks, consider framing your requests or marketing copy in terms of potential losses, rather than gains. However, keep in mind that this tactic is likely very context-dependent, and be sure to A/B test to see which frame works best for your unique situation.
Wilson et al. (1987) presented participants with a hypothetical situation whereby they had terminal liver disease that would kill them within a year, and were presented with options by a doctor. One group’s options were framed positively in terms of survival (e.g. 10% chance of surviving the surgery) while the other group’s options were framed negatively in terms of death (e.g. 90% chance of dying from the surgery). Different participants were given different values (e.g. 40% chance of surviving, 60% chance of dying). For most value combinations, those who viewed an option framed in respect to survival were more likely to risk the surgery. For example, 46% of those presented with a 10% chance of survival (the positive frame) opted for the surgery, while only 28% of those presented with a 90% of dying (the negative frame) opted for it. Thus, if attempting to influence a customer’s evaluation, judgment, or perception of an attribute, brand, or singular choice, a positive frame may be ideal.
Homer & Yoon (1992) created five colour magazine-style advertisements for a fictitious brand of mouthwash that they called “Mintgard.” Four of these ads were meant to prevent participants from guessing the study. The fifth was used for the experiment and either involved a negative frame or a positive frame. In the negative frame, a picture of a “foul-smelling” mouth was used along with the copy “they will hate your bad breath” followed by an explanation of how poor oral hygiene has negative consequences. In the positive frame, a picture of a “fresh-looking” mouth was used, along with the copy “they will love your fresh breath” and an explanation of how good oral hygiene leads to benefits. Due to the nature of the study (which in addition to framing effects, looked at emotional and cognitive responses) the results are a bit technical. However, in respect to framing, Homer & Yoon (1992) found that the negative frame (the ad focusing on bad breath) was more influential than the positive frame (the ad focusing on good breath). Thus, when marketing a product or service that involves a behaviour or habit, it may be best to utilize a negative frame.
Ainiwaer, A., Zhang, S., Ainiwaer, X., & Ma, F. (2021). Effects of message framing on cancer prevention and detection behaviors, intentions, and attitudes: Systematic review and meta-analysis.
Journal of Medical Internet Research, 23(9), e27634.
https://doi.org/10.2196/27634
Best, R., & Charness, N. (2015). Age differences in the effect of framing on risky choice: A meta-analysis.
Psychology and Aging, 30(3), 688-698.
https://doi.org/10.1037/a0039447
Dolgopolova, I., Li, B., Pirhonen, H., & Roosen, J. (2021). The effect of attribute framing on consumers’ attitudes and intentions toward food: A meta-analysis.
Bio-based and Applied Economics, 10(4), 253-264.
https://doi.org/10.22004/ag.econ.320229
Gächter, S., Orzen, H., Renner, E., & Starmer, C. (2009). Are experimental economists prone to framing effects? A natural field experiment.
Journal of Economic Behavior & Organization, 70(3), 443-446.
https://doi.org/10.1016/j.jebo.2007.11.003
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.
https://doi.org/10.1007/s12160-011-9308-7
Homer, P. M., & Yoon, S.-g. (1992). Message framing and the interrelationships among ad-based feelings, affect, and cognition.
Journal of Advertising, 21(1), 19-32.
https://doi.org/10.1080/00913367.1992.10673357
Johnson, E. J. & Goldstein, D. (2003). Do defaults save lives?
Science, 302(5649), 1338-1339.
https://doi.org/10.1126/science.1091721
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.
https://doi.org/10.1006/obhd.1998.2804
Lexchin, J. (1999). How patient outcomes are reported in drug advertisements.
Canadian Family Physician, 45, 1213-1216.
https://pubmed.ncbi.nlm.nih.gov/10349065
Linville, P. W., Fischer, G. W., & Fischhoff, B. (1993). AIDS risk perceptions and decision biases. In J. B. Pryor & G. D. Reeder (Eds.),
The Social Psychology of HIV Infection (pp. 5-38). Lawrence Erlbaum Associates, Inc.
Malenka, D. J., Baron, J. A., Johansen, S., Wahrenberger, J. W., & Ross, J. M. (1993). The framing effect of relative and absolute risk.
Journal of General Internal Medicine, 8(10), 543-548.
https://doi.org/10.1007/bf02599636
O’Keefe, D. J., & Jensen, J. D. (2007). The relative persuasiveness of gain-framed loss-framed messages for encouraging disease prevention behaviors: A meta-analytic review.
Journal of Health Communication, 12(7), 623-644.
https://doi.org/10.1080/10810730701615198
Steiger, A., & Kühberger, A. (2018). A meta-analytic re-appraisal of the framing effect.
Zeitschrift für Psychologie, 226(1), 45-55.
https://doi.org/10.1027/2151-2604/a000321
Tversky, A., & Kahneman, D. (1981). The framing decisions and the psychology of choice.
Science, 211(4481), 453-458.
https://doi.org/10.1126/science.7455683
Wilson, D. K., Kaplan, R. M., & Schneiderman, L. J. (1987). Framing of decisions and selections of alternatives in health care.
Social Behaviour, 2(1), 51-59.
Xu, J., & Huang, G. (2020). The relative effectiveness of gain-framed and loss-framed messages in charity advertising: Meta-analytic evidence and implications.
Journal of Philanthropy and Marketing, 25(4), e1675.
https://doi.org/10.1002/nvsm.1675
Discover More Knowledge
Contact Us