The Internet Makes You Think You’re Smarter Than You Are: An Interview with Matthew Fisher
FROM THE JULY–AUGUST 2015 ISSUE
The research: Yale doctoral candidate Matthew Fisher and his colleagues Mariel Goddu and Frank Keil asked people a series of questions that seemed answerable but were actually difficult. The questions concerned things people assume they know but actually don’t—such as why there are phases of the moon and how glass is made. Some people were allowed to look up the answers on the internet, while others were not. Then the researchers asked a second set of questions on unrelated topics. In comparison with the other subjects, the people who’d been allowed to do online searches vastly overestimated their ability to answer the new questions correctly.
The challenge: Does the internet make us overconfident? Are we unable to distinguish between what’s stored in our own heads and what’s in the cloud? Mr. Fisher, defend your research.
Fisher: We’ve zeroed in on access to this massive online database of information as the cause of an illusion of understanding. Even when people did searches and got irrelevant or no results, they were far more confident that they’d know the answers to unrelated follow-up questions.
HBR: What if those who got internet access just happened to know the answers to the follow-up questions better?
Randomly assigning participants to one of the two conditions took care of that worry. All the potential differences between the groups, such as previous knowledge, were then randomly distributed across the groups, so the only difference between the two groups was whether they used the internet to look up the answers to our initial questions.
In some ways this seems obvious. If I know I have access to a mechanic, I’ll be more confident that I can keep my car running.
We’re making a crucial distinction here. We didn’t see that people were more confident that they could find answers if they had access to search. We saw that people were more confident that they knew the answers—had the information in their heads—if they had access to search. It’s more like thinking you know how to fix a car if you have access to a mechanic.
How could you tell people thought they had the information in their heads? Couldn’t they have been more confident because they knew they could look things up?
In one experiment we simply asked them how well they could explain answers without using any outside sources. In another, instead of asking about their confidence, we told them that people who could give better answers would show more brain activity while answering. Then, instead of having them rank their confidence on a scale, we showed them a series of brain scans that depicted less to more brain activity. We asked them to indicate how much brain activity they’d use to come up with their answer. People who had been given access to search consistently chose images with more brain activity.
Yeah, I think we made that technique up.
So what’s actually going on here?
There’s a lot of research about transactive memory partners. Take an old married couple recalling their first date. In isolation neither recalls much, but if you put their memories together, they can re-create a richer memory that’s more than the sum of each person’s fragments. Now it looks like a machine can be that transactive memory partner. You plus a search is more than you or the search. It’s just that we think it’s only us.
Plus, searching the internet is almost effortless, and it’s almost always accessible. You never face your ignorance when it’s there. Because we’re so deeply plugged into it, we misattribute the connection to knowledge to actually having the knowledge ourselves. It becomes an appendage. We like to use the term “cognitive prosthesis.”
But is it so bad to have this prosthesis? It’s like a bionic arm. Bionic arms are cool!
Except what happens when it doesn’t work? Or when you can’t access the knowledge? With some professions, we want people to be truly knowledgeable, not have a false sense of their knowledge. Surgeons, for example. At the very least we have to start structuring our world so that if such people rely on this appendage, they’re never cut off from it. Look, it’s obvious the internet has benefits. We think there’s an inherent trade-off between learning about the world yourself and storing information about the world somewhere else besides your head. The more we use the internet, the harder it will be to assess what people truly know. And that includes assessments about ourselves.
How have people reacted to your findings?
They’ve resonated a lot more than I anticipated. Because there are so few places now where we can’t access the internet, we do feel it when it happens. On a plane or in a conversation where it would be rude to pull out a mobile device, you run into this roadblock. Suddenly, we don’t feel as smart. But we never were smarter, really; we just thought that what we could search for was actually something we already knew.
So the internet makes us feel like know-it-alls? Didn’t I already know that?
Psychologists have actually studied this “I knew it all along” phenomenon. When someone with credibility explains something to a layperson, a common reaction they get is “That’s obvious” or “Oh yeah, I knew that.” So one common ploy of psychologists is to describe their findings as the exact opposite of what they are, and people will react with “Yeah, that’s right—that makes sense.” I could have played this game and said to you, “We found that people feel dumb when they use the internet, that they know nothing compared to this vast resource.” And you would have said, “Yeah, duh, of course.”
Wait, how do I know you didn’t actually do that? Which is the real finding?
You have the paper.
What made you want to study this?
It was a nice real-world way to look at what I’m most interested in: metacognitive awareness, or people’s ability to assess how well they can explain things around them. Emotional investments can give people the illusion of insight. This happens in politics a lot. You end up thinking you know arguments better than you do. Our research has shown that when college students are asked to assess their knowledge of topics, they are least accurate about how much they know about their own majors. When you’re invested in something, you like to think you know a lot more than you do.
I definitely know more about Q&As than most people!
I’m sure you think you do.
A version of this article appeared in the July–August 2015 issue (pp.26–27) of Harvard Business Review.
Scott Berinato is a senior editor at Harvard Business Review.