Less-Is-Better Effect: Why a Small Overflowing Cup Beats a Bigger One

The two servings above are not equal: the second holds an ounce more ice cream. Seen side by side, it is valued more on average: Hsee's participants who saw both servings at once were willing to pay more for the larger one. Seen alone, the servings are judged by a different question, not how much ice cream is this but how good a serving is this, and the cup supplies the answer. Seven ounces spilling over a 5-ounce cup looks lavish; 8 ounces sitting below the rim of a 10-ounce cup looks mean. Each serving is compared with its own container rather than with the other serving, and on that comparison the smaller one wins.

Christopher Hsee's explanation is the evaluability hypothesis, which he set out in 1996 and applied to these cases in 1998. Some attributes are easy to evaluate on their own: whether a cup is full, whether a set of dishes has broken pieces, whether a gift is at the top or bottom of its category's price range. Others are hard to evaluate without a comparison: how much 7 ounces of ice cream is, whether 24 pieces of dinnerware is a lot, whether 45 dollars is a generous gift. When an option is judged alone, the easy attributes carry the judgment, even when the hard ones matter more. Put two options together and the hard attributes become easy, because each now has a comparison, and the judgment flips.

The effect is therefore about the mode of evaluation rather than about ice cream. Hsee, Loewenstein, Blount and Bazerman reviewed in 1999 a wide set of reversals between joint and separate evaluation, in hiring, in pricing and in policy, and argued that the two modes draw on different information. Separate evaluation is the common case in life: a shopper sees one serving, a manager sees one candidate, a recipient receives one gift. The comparison that would correct the judgment is usually not available, and the easy attribute, which is often the one the seller controls, decides.

The effect replicates. A 2023 preregistered replication by Vonasch, Feldman and colleagues reran three of Hsee's four 1998 studies with 403 online participants and found the less-is-better effect in separate evaluation in all three, though the ice cream effect was less than half the size of the original. The reversal to more-is-better in joint evaluation was weaker than in the original studies and, for the dinnerware, did not reach significance. The first half of the finding, that a smaller option judged alone can be valued above a larger one, is the robust part.

Christopher Hsee of the University of Chicago published "Less is better: When low-value options are valued more highly than high-value options" in the Journal of Behavioral Decision Making in 1998. The paper reported the effect in three settings, each with options evaluated separately by different groups and, in two cases, jointly by a further group.

The first study concerned gifts. Eighty-three students imagined receiving either a 55 dollar wool coat from a store where coats run from 50 to 500 dollars, or a 45 dollar wool scarf from a store where scarves run from 5 to 50 dollars. Those given the scarf rated the giver as more generous than those given the coat, 5.63 against 5.00 on the rating scale, though the coat cost more. The scarf was at the top of its range and the coat at the bottom, and the position within the range was what people could evaluate.

The second study is the one reproduced above. Participants imagined being on a beach at Lake Michigan in summer, in the mood for ice cream, with a vendor selling a premium brand by the cup. One group read that each serving was a 5-ounce cup holding 7 ounces of ice cream; another read that it was a 10-ounce cup holding 8 ounces. Asked the most they would pay for a serving, the first group averaged 2.26 dollars and the second 1.66. A third group saw both vendors described together and valued the 8-ounce serving more. The paper does not report how many participants were in each ice cream condition.

The fourth study used dinnerware on clearance in a store where sets run from 30 to 60 dollars. One set had 24 intact pieces; the other included the same 24 intact pieces, seven more intact pieces and a few broken ones, 31 intact in all. With 104 participants, those who saw only the smaller set would pay 32.69 dollars on average and those who saw only the larger set 23.25 dollars. Shown both sets together, participants paid more for the larger, 32.03 against 29.70. Hsee concluded that the less-is-better effect occurred only when options were evaluated separately and reversed itself when they were juxtaposed, and explained the pattern by the evaluability hypothesis.

The effect appears wherever an option is judged alone against a frame that is easy to read. The rows are illustrations, not findings from the cited studies.

The effect lives in separate evaluation, so the counter is to bring in the comparison that separate evaluation lacks.

• Evaluate jointly. Before valuing an option on its own, put at least one alternative next to it. The attribute that was hard to judge alone, amount, quality, price, becomes easy with a comparison.
• Name the attribute that matters. Decide in advance whether you care about how much ice cream there is or how full the cup looks. The frame will answer the second question for you; only you can insist on the first.
• Find the reference point you were handed. An overflowing cup, a price range, a category and a broken piece are all reference points supplied by the seller or the situation. Ask what the judgment looks like against a reference point you choose.
• Convert to a common unit. Price per ounce, cost per intact piece and salary per hour strip away the frame and leave the quantity.
• Be suspicious of easy impressions. When one option simply looks better and you cannot say why in numbers, the easy-to-evaluate attribute is probably doing the work, and it may not be the one you would choose on reflection.

The less-is-better effect is a reversal between separate and joint evaluation. Its neighbors involve frames and comparisons in other ways.

• It is not distinction bias, which is the mirror problem: when options are compared side by side, small differences loom larger than they will in experience. Hsee and Zhang described it in 2004. Joint evaluation corrects the less-is-better effect but creates distinction bias, so neither mode is right for every decision.
• It is not unit bias, the tendency to treat one unit of something, a plate, a serving, a package, as the right amount to consume. Unit bias concerns how much people take; the less-is-better effect concerns how much they think a portion is worth.
• It is not the contrast effect, in which a judgment shifts because of something seen just before. Contrast is a comparison with a prior item; the less-is-better effect comes from the absence of comparison, with the option's own container or category standing in for it.