Are You Undervaluing Your Own Attention?

Narrator: Every day, people make dozens of choices about how to deploy mental effort and where to focus. Do we give our attention to work, or play? Should we exercise, or zone out in front of the TV? Understanding how we choose where to put our attention and what we’re willing to give up to maximize it could have an impact on the things we know we should do but don’t necessarily want to do.

Eric Zwick: We’re interested in whether, in recognizing that long-term goal, people actually invest in the technologies, the intermediate inputs, that will help them achieve that goal. In the case of the medication, it’s like setting the reminders or putting the pill bottle in the right place, such that you know to take it each day. In the case of going to the gym, it’s making space on your calendar or making a plan with a buddy so that you’re more likely to go to the gym. In the case of doing your homework, it’s, again, setting that calendar, making appointments so that you actually get that homework done.

Narrator: That’s Chicago Booth’s Eric Zwick. He and his coauthors ran three studies that tested the rational inattention theory. The theory assumes that people fully understand the consequences of, say, scrolling through Instagram instead of hunkering down and writing that report—and that we choose where and how best to put our attention. Their findings are a little discouraging.

Eric Zwick: One way to think about the question is, is the right model of human behavior that we’re computers that optimally trade off the costs and benefits of reminders and set exactly the right amount of reminders so that we exactly remember to go to the gym the right amount of time and get the precise workout that we want? Or are we totally lost in the sea and just randomly choosing to do activities that come to us at that particular point in time, with no ability to set up in advance the types of technology that would help us achieve those goals and go to the gym?

Narrator: For the first experiment, they asked nearly 1,400 US students enrolled in an online coding class to complete three 15-minute coding lessons each week. The students were rewarded monetarily for completing the lessons, with a randomized range of bonuses of $1 to $5. To help make completing the tasks easier, the researchers offered students a plan-making tool to block out three 15-minute windows on their online calendars.

Eric Zwick: In the coding experiment, we found that people were about 60–80 percent of the way toward computer. It is definitely the case that people recognize that these intermediate inputs help them achieve their goals. It is also the case that people are pretty far from what we think of as the optimal benchmark.

Narrator: In the second experiment, the researchers offered financial bonuses to 2,300 participants to complete an online survey. The catch? The survey wouldn’t be sent to them for anywhere from two days to six weeks later. To help participants complete the survey and earn their bonuses of $3 to $12, the researchers offered to send a set of three reminder emails, gauging the participants’ willingness to pay for them. But as in the first experiment, they find that participants undervalued the reminders.

Eric Zwick: We found that folks were more like 30–50 percent from the computer. One way to think about how to measure that is if you increase the payoff to completing the survey, you should see folks increase their take-up, or value, of the reminders. And we do see that increase in take-up, but it’s much too small in terms of the quantitative responsiveness of their take-up.

The third experiment was a bit different and focused on math problems. Basically, a bunch of equations, where it’s like 3 + 6 = 9, and then 3 + 7 = 12, and the subjects had to guess, or figure out, what share of the equations we showed them were true or false. And we gave them an intermediate input, which was the ability to shorten the prompts so that there were fewer equations, or to make it easier, such that if 90 percent of them were true, it would be easier to tell than if 60 percent of them were true, so that was like roughly 50/50, and so that was the intermediate input in that case. And we then, across each experiment, tried to measure how people valued those intermediate inputs relative to the stakes for achieving the goal of actually completing the class, getting the right answer, or completing the survey.

In some cases, they were completely lost at sea. Furthermore, when we gave them feedback about the return to these kinds of intermediates in helping them, we found that they responded to that feedback in unusual ways, in some cases actually moving further away from what would make sense from the optimal-benchmark perspective, which suggests that even learning from your experience about the value of these intermediate inputs is a pretty slow process, an unpredictable process, and folks’ ability to actually develop appropriate values for these kinds of intermediates takes a while to happen, if it ever does.

Narrator: Although the stakes in the experiments were fairly small, the findings have a compounding effect that often carries over into our daily lives. If one day you forget to take your heart medicine, you probably don’t think that’s too costly. But if you are 30 percent noncompliant with your preventative medication, that could considerably increase the risk of heart attack down the road. The researchers say that’s where the findings suggest that the stakes of inattention in today’s “distraction economy” may be much higher than you think. (upbeat music)