Four common data analysis mistakes, from a 'Freakonomics' economist

Data may not accurately reflect people's true behavior

People tend to make four common mistakes when looking to data to predict the future, according to "Freakonomics" co-author Stephen Dubner. 

1. They don't account for the unexpected.  

Even many economists were unable to predict the great recession, Dubner explains, "because as we advance in this age and increasingly rely on data to help predict the future, it often can't. And one of the reasons it often can't is because there are too many unforeseen variables."

Dubner demonstrates this point with an anecdote about raising turkeys for consumption. As people turned to white meat as a healthier alternative to red meat, he explains, farmers began raising turkeys with larger breasts to meet demand. However, the larger-breasted turkeys were unable to procreate, meaning that farmers had to invest in artificial insemination for female turkeys.

Increase your analytical and reporting capacity

2. They don't realize that people lie.

People don't always tell the full truth when they self-report, as seen in one study about hand washing habits at a pediatric hospital in Australia. Seventy-three percent of doctors self-reported that they washed their hands after using the restroom, but when hospital administration asked nurses to investigate whether doctors were washing their hands, they found that only 9% of doctors were doing so.

3. They forget how difficult it is to change behavior.

Inspired by the study at the Australian hospital, Cedars-Sinai Medical Center sought to boost hand-hygiene compliance at its own hospital.

A hospital committee launched a "posse patrol" to monitor doctors' hand washing habits, popping out to surprise doctors who washed their hands with a $10 Starbucks gift card. While the doctors responded well to the gift cards, hand-hygiene compliance didn't improve—perhaps because doctors knew the posse would no longer be on the floor during their rounds, or they figured out the posse's schedule.

Getting the hospital to comply 100% with hand hygiene required a slightly more drastic measure. The committee took a photo of unwashed hands covered in bacteria and used the image as a screensaver on every hospital computer. 

"It just goes to show you that incentivizing, predicting what will work and the process of trial and error is a tough road that takes time and patience," Dubner says.

4. They overthink behavior.

One study conducted at Yale University by an economist shows that sometimes, behavior is hardly that complex.

The study, which involved training capuchin monkeys to use metal washers as currency over the course of several months, sought to determine whether monkeys used the currency to purchase certain foods over others, and whether doubling the price of their favorite foods would sway the monkeys from buying them. A researcher was thrilled to see one monkey handing a washer to another monkey in what appeared to be an act of altruism. However, it soon became clear that the other monkey viewed the token as payment for intercourse (Stansbury, eCampus News, 6/23). 

How to understand academic analytics


Like what you're reading?

Keep up with the latest by following EAB on LinkedIn.

Follow us


Next in Today's Briefing

Geography determines some students' college plans

Next Briefing

  • Manage Your Events
  • Saved webpages and searches
  • Manage your subscriptions
  • Update personal information
  • Invite a colleague