Extremeness aversion
Options at the edges feel risky, so the safe-seeming middle gets chosen.
What it means
Extremeness aversion is the tendency to avoid options that are extreme on the available attributes and to gravitate toward intermediate ones, so that an option's attractiveness rises merely because it occupies a middle position within the offered set. It is the mechanism underlying the compromise effect, where the central option in a range gains share, and it interacts with the attraction (decoy) effect to show that choices are governed by relative comparisons within the context rather than by stable absolute valuations. The pull toward the middle is partly a matter of reason-based choice: the compromise option is easy to justify as a sensible balance, hedging against the risk of being too cheap-and-shoddy or too dear-and-extravagant. A direct practical consequence is that adding a high, rarely chosen option to a menu can lift sales of the formerly top option by making it the new compromise. It matters because it means sellers and choice architects can steer decisions simply by manipulating the range of alternatives presented.
Why the middle feels safe
Two complementary explanations sit behind the pull. The first is reason-based: a middle option is easy to defend because it splits the difference, so a chooser who must justify the pick, even privately, reaches for it. The second, from Tversky and Simonson's later work, is that people weigh each option's attributes relative to the others in the set, and the disadvantages of an extreme option loom larger than its matching advantages, much as losses outweigh gains. A cheap laptop's weak processor feels like a bigger loss than its low price feels like a gain. The intermediate option minimizes the sum of these felt disadvantages. Both accounts share one idea: an option's value is constructed inside the set on offer, not read off a stable internal scale.
What the evidence shows
The broadest test is a meta-analysis by Neumann, Boeckenholt and Sinha covering 142 experimental observations. It confirms the basic pattern: middle options are chosen significantly more often than the average alternative, so the effect is genuinely robust rather than a lab curiosity. But the size of the pull swings by up to threefold with design. It grows with more tradeoff dimensions, non-numeric attributes and utilitarian products, and shrinks for simple price-quality tradeoffs or binary-to-trinary comparisons. The authors also warn that three common ways of measuring it, absolute share change, relative share shift and raw middle-option proportion, can disagree and even reverse which moderators look important. Anyone quoting a single headline compromise-effect share without saying how it was computed is overstating the precision of the finding.
Where it shows up
The lever is always the same: change the range, not the option itself. Retailers add a rarely-bought premium tier so the old flagship becomes the compromise, lifting its share. The effect also has a public-health edge. Sharpe, Staelin and Huber showed that dropping the smallest soft-drink size, or adding a larger one, shifts buyers toward bigger portions and raises total consumption, then ran the mechanism in reverse to design menus that cut intake without new taxes. Threat sharpens the pull further: across studies during COVID-19, Kim and colleagues found that heightened safety concern amplified travelers' preference for compromise options, more strongly for leisure than business trips. Menu design, portion policy and pricing all become steering wheels once the middle is understood as manufactured rather than found.
Where it weakens
The middle only pulls when preferences are unformed. Give a chooser a clear, well-articulated preference and the framing loses its grip; experts and repeat buyers who know which attribute they want are far less swayed. Because the effect is relative to the presented range, it can be diluted or inflated by how extreme the endpoints are, and a set with implausible extremes can read as a trick and backfire. Accountability cuts the other way: Simonson found compromise and attraction effects were stronger among people who expected to justify their choice, because the middle supplies the ready reason. Treat extremeness aversion as a default that surfaces under uncertainty and fades under expertise, strong motivation or transparent manipulation, not a constant that can be dialed up at will.
Examples
When a third, very expensive item is added to a menu of two, sales shift toward the now mid-priced item, which buyers feel they can defend as the reasonable choice.
Offered laptops with 8, 16 or 32 gigabytes, most buyers take 16. Add a 64-gigabyte model and 32 quietly becomes the sensible middle that nobody has to defend.
A donation page shows buttons for 5, 25 and 100 pounds, and most clicks land on 25. Add a 500 button and 100 starts to look like the modest, defensible choice.
Scanning a restaurant wine list, many diners skip the cheapest bottle to avoid looking cheap and the priciest to avoid overpaying, settling on a mid-range option they could not otherwise rank on taste alone.
Offered watchful waiting, medication, or surgery, patients and clinicians often settle on medication as the defensible middle course between doing nothing and doing something drastic, even when an extreme fits the case better.
First described in Itamar Simonson & Amos Tversky (1992).
Key references
- Kim, J., Park, J., Lee, J., et al. (2022). COVID-19 and extremeness aversion: The role of safety seeking in travel decision making. Journal of Travel Research, 61(4), 837-854. doi.org/10.1177/00472875211008252
- Neumann, N., Boeckenholt, U., & Sinha, A. (2016). A meta-analysis of extremeness aversion. Journal of Consumer Psychology, 26(2), 193-212. doi.org/10.1016/j.jcps.2015.05.005
- Sharpe, K. M., Staelin, R., & Huber, J. (2008). Using extremeness aversion to fight obesity: Policy implications of context dependent demand. Journal of Consumer Research, 35(3), 406-422. doi.org/10.1086/587631
- Simonson, I., & Tversky, A. (1992). Choice in context: Tradeoff contrast and extremeness aversion. Journal of Marketing Research, 29(3), 281-295. doi.org/10.1177/002224379202900301
- Simonson, I. (1989). Choice based on reasons: The case of attraction and compromise effects. Journal of Consumer Research, 16(2), 158-174. doi.org/10.1086/209205