Does Having a Choice Provide an Illusion of Control?

The theory has held for decades: give people a choice and you’re also giving them an illusory sense of control. For example, a much-cited 1975 study found that if people are allowed to pick the numbers on a lottery ticket rather than having them randomly assigned, they’re more likely to think their ticket will win—even though every ticket carries the same odds.

“We didn’t start our project questioning this assumption,” says Joowon Klusowski, an assistant professor of marketing at Yale SOM. She and Deborah Small, also at SOM, along with Wharton’s Joseph Simmons, were simply looking at how different choice processes affected various outcomes. “But as we started running studies to verify that having a choice in a lottery makes us believe we’re more likely to win, we weren’t replicating the effects. So we dove deeper.”

The researchers began running one test after another, tweaking the setup to see if adjustment to different variables would finally reproduce the expected results. After the first 11 experiments they still hadn’t found support for the original findings. “The data was speaking to us,” Small says. In a study published in Psychological Science, the researchers conclude that having choice on its own doesn’t make people believe they are more likely to win in a random drawing. But if someone believes—incorrectly—that one option is more likely to be correct, choice allows them to act on that belief.

In all of the experiments, the participants—more than 10,000 in total—were entered into a lottery of some sort. In every case, one group of participants was allowed to choose their lottery options while the other group was assigned their options. In the first experiment, for example, select participants were allowed to choose a number between one and six; other participants were assigned a number.

Prior to revealing the winning draw, the researchers then surveyed the participants about how likely they thought they were to win the lottery and how confident they felt. According to the established theory, those who were given a choice among functionally identical options should have expressed a greater likelihood of winning, or at least greater confidence in their selection, than those without choice.

“Based on the first 11 experiments, it didn’t look like the effect existed,” Klusowski says. The researchers did, however, find uncommon cases in which people believed the option they chose had a higher probability of winning. Was this result, they wondered, the direct consequence of giving people a choice, or due to something else? The remaining experiments were designed to determine what was behind this pattern.

In this second set of experiments, participants had the chance to win a $1 bonus contained in a virtual box in a computer game. Again, one group was able to choose a box; the other group was assigned a box. Participants were also asked to estimate the probability that each box contained the dollar. In some of these experiments, there was also a third condition, in which participants estimated these probabilities before even choosing or being assigned a box. If participants rated their box to be more likely to win only after they were given a choice, this would suggest instead that choice provided them a sense of control. However, if they rated one of the boxes to be more likely to win even before choosing or being assigned a box, this would suggest they had a preexisting belief about which box was “more likely” to win.

The researchers found the latter to be true: having a choice was allowing the participants to act on preexisting beliefs. Of course, many of us harbor irrational notions about equivalent choices—we feel that the prize sits behind the middle door, or the number 3 happens to be lucky, or the lotto card with the blue background must be the winner. Choice does not create this feeling, but it provides agency to act on it. “In short, choice can be helpful because it allows you to get what you want,” Klusowski says.

For Klusowski, the results raise two considerations. She hopes, she says, that the findings encourage researchers who plan to study the subject of choice to be careful in determining whether effects they find are related directly to choice itself or choice is the conduit for something else.

More generally, Klusowski says that this project reoriented her convictions about assumed truths. “One important takeaway for me was that, even when faced with the most influential and widely accepted ideas, it’s often our job as scientists to continue questioning them,” she says. “We shouldn’t take anything as pure truth. That’s a point I’ve taken to heart.”