The Price of Curiosity: Wanting Information That Changes Nothing

A lottery has finished, and its outcome cannot be improved. Finding out what happened, however, requires a short wait. What does someone buy by agreeing to that wait? Not a better decision or a second chance, but knowledge of the result. This small separation creates a useful puzzle. Information can matter to us even when it offers no control over events. Its attraction may depend both on how much remains uncertain and on whether the news concerns something pleasant or unpleasant.

A 2021 study by Lieke van Lieshout, Iris Traast, Floris de Lange and Roshan Cools investigates that puzzle through monetary lotteries. Participants encounter possible gains and losses, but learning the result does not allow them to change the draw or move to a more profitable option. The researchers therefore distinguish information that improves an outcome from information that merely reveals it. Their central question is not which gamble someone prefers. It is how much someone wants to know about an outcome that will count regardless of whether it is observed. [1]

A known vase with an unknown result

The task begins with a displayed vase containing twenty red and blue marbles. Each color carries a different monetary value. A marble will be randomly selected, and its associated amount will be gained or lost. In the first experiment, the vase can contain eighteen marbles of one color and two of the other, fifteen and five, or ten and ten. Participants know the possible outcomes and their probabilities. What remains hidden is which outcome will actually occur.

Uncertainty here is not simply a synonym for an evenly divided vase. The distance between the possible monetary outcomes also matters. In the analysis, outcome uncertainty is quantified as the variance of those outcomes, while expected value is their probability-weighted average. An average describes the center of the possibilities; variance describes their dispersion. Keeping these quantities distinct allows the researchers to examine uncertainty, expected amount, and gain versus loss context without treating all three as the same property.

Consider an illustrative calculation that is not a trial from the study. An equally likely award of twenty or sixty points has an expected value of forty. An equally likely award of thirty or fifty points has the same expected value. Yet the first pair spreads farther from that common center. Knowing the average does not resolve the difference between the lotteries. This example establishes no psychological effect by itself. It explains why someone interested in uncertainty needs a measure beyond the expected amount.

From a rating to a costly choice

Three experiments ask participants to report curiosity on a scale from one to four. Giving a higher rating does not increase the chance of seeing the result. Participants are told that the result will be revealed on half the trials, independently of their response. A curiosity rating is therefore not a disguised strategy for obtaining more feedback. The procedure separates a report about wanting information from the mechanism that grants access to it.

The first experiment presents gains and losses in separate blocks. The next two intermingle gain and loss trials, allowing the researchers to examine whether a prolonged positive or negative context might explain the result through a general change in mood or arousal. The analyzed samples contain thirty-four participants in the first experiment and thirty-three in each of the next two. Brain activity was recorded during the third experiment, but this paper does not report those neural findings. Its behavioral results cannot consequently be presented as evidence that a particular brain region caused curiosity.

The fourth experiment adds a different kind of measurement. Thirty-four included participants choose whether they want to see the result. If they request it, they spend extra time waiting. Declining the information does not cancel the monetary outcome. The methods specify a pre-outcome interval of three and a half to six and a half seconds when information is requested, compared with half a second at that stage when it is declined. The additional time cost is thus three to six seconds. This is a modest price, but it makes information seeking observable as a choice with a consequence.

Calling this a willingness-to-wait measure also preserves its particular meaning. The study does not measure how much money each participant would pay for information, and a few seconds of laboratory waiting are not interchangeable with every other kind of cost. Nevertheless, the procedure asks a sharper question than a rating alone: will the person actually accept a delay in order to know something that cannot improve the draw?

Bad news does not eliminate curiosity

Across the three rating experiments, curiosity increases with outcome uncertainty. Curiosity is also generally higher for gain lotteries than for loss lotteries. The waiting experiment produces a similar overall pattern. Participants choose to wait on an average of 45.4 percent of trials. The reported averages are 55.1 percent for gain trials and 35.6 percent for loss trials. These figures describe responses within this task, not fixed population rates of curiosity.

The lower level of information seeking about losses is not an absence of information seeking. People also become more willing to learn the outcomes of more uncertain loss lotteries. Wanting pleasant news and wanting uncertainty resolved therefore need not be identical motives. A person can want to know an unpleasant result because it answers a question, even while finding information about gains more attractive overall.

The interpretation still requires care. It would be stronger than the evidence to announce that two completely independent psychological mechanisms have been proven. The researchers find no convincing, consistent support across experiments for a reliable difference between the uncertainty effect in gains and losses. One model in the second rating experiment detects an interaction, but another analysis does not support it. The overall pattern fits largely separate influences. It does not directly expose two identical internal components in every participant.

Expected amount produces a more variable pattern across experiments. In the waiting experiment, larger expected gains encourage information seeking, whereas larger expected losses discourage it. Two influences can therefore point in opposite directions within the loss domain: greater uncertainty can increase the desire to know, while a larger expected loss can decrease it. Describing a refusal to look as simple indifference would miss that distinction.

Information for knowing and information for acting

An everyday notification provides a useful comparison only after asking what opening it would enable. If a report can change the next decision, its information serves an instrumental function that this lottery task deliberately removes. The controlled arrangement is informative because participants cannot use the news to improve their payoff. Turning its results into a general rule for research, news reading, or personal decisions would reunite two functions the experiment carefully separates.

The study's small, predominantly young samples and narrowly specified monetary task also leave room for further work on ordinary curiosity. A laboratory vase offers explicit probabilities and outcomes; everyday questions often do not. Within the experiment, however, the combination of reported interest and actual waiting decisions establishes something substantive. Information seeking is not exhausted by opportunities to improve a monetary result. Uncertainty and the prospect of gain rather than loss both help describe what participants want to learn.

The evidence-based conclusion is therefore about the value of knowing, not a command to seek or avoid every answer. Resolving uncertainty may be attractive in itself, or it may relieve an unpleasant state of not knowing; the study does not settle that explanation. Pleasant information supplies another attraction, without wiping out curiosity about unpleasant information. An outcome can remain entirely unchanged while knowledge of it changes a person's willingness to wait. That is the feature the task makes visible: even powerless information can have a price someone chooses to pay.

Sources

1. van Lieshout, L. L. F., Traast, I. J., de Lange, F. P., and Cools, R. (2021). Curiosity or savouring? Information seeking is modulated by both uncertainty and valence. PLOS ONE, 16(9), e0257011. Open primary research article.
2. Creative Commons Attribution 4.0 license.

Reviewed October 14, 2026. The research, © 2021 van Lieshout and colleagues, is available under CC BY 4.0 and is credited here. This independently written English explanation reorganizes the discussion and adds an illustrative points calculation and editorial interpretation. No direct quotations are used. Supplementary files and raw data were not independently reanalyzed.