Why Invisible Gorilla Matters

Selection attention test 'creator' Daniel Simons interview:

"Munger: So what are some real-world applications of this research?

Simons: I think there are a lot. In inattentional blindness you’re not seeing something that’s right there because your attention is engaged. The most obvious practical application of that is driving. We intuitively think that if something important happens right in front of us, we will see it.

Munger: Like a gorilla.

Simons: Like a gorilla, that’s right. Dan Levin has done studies where he just asked people, would you notice if something like this happens? He shows people the video, gives them the instructions, points out the gorilla, and then asks them “how likely would you be to notice the gorilla if you were doing this task and counting the passes?” Ninety percent of people say they’d notice. Regardless of how you ask that question, you get high confidence, and a high percentage saying “yeah, of course I’d notice that.”

That’s the intuition that’s interesting, and that’s the one that’s dangerous. If we were completely aware of these limits on attention, we wouldn’t do things like talking on cell phone while driving: We would know that it would make us just that much less likely to notice something. But we don’t have that insight into our own awareness. It’s only in that rare case where you actually have an accident that you become aware that you’ve missed something.

There’s a whole class of accidents called “looked but failed to see” accidents. Motorcyclists will swear that they’ve made eye contact with a driver right before the driver turned left in front of them. Some motorcyclists assume that drivers are out to get them. But the reality is that the driver probably never saw them, because the motorcycle is unexpected. When you’re looking for cars, motorcycles are rare; they’re not what you’re looking for, even when they’re distinctive.

Munger: We covered a study on Cognitive Daily a few years back, where participants were in a driving simulator, and a surprise motorcycle came up. If the motorcycle was the same color as some of the street signs in the simulation, people didn’t run into it, but if it was a different color, they did. And we got tons of traffic from motorcycling forums about it, because they were wondering if they should always dress in red, to look more like a stop sign.

Simons: Yes, the study was actually done by one of my former students, Steven Most. He occasionally gives talks to motorcycle safety and advocacy groups. It’s something that motorcyclists know—that drivers just don’t see them—and they start these campaigns, like “watch for motorcycles.” The problem is, those aren’t going to be terribly effective, because we’re great pattern recognizers. You can see that sign, and it might help you spot motorcycles for the next couple miles, but then you’re not seeing any motorcycles, and you stop looking for them."

"Munger: I noticed in the book The Invisible Gorilla that you start by talking about these very basic principles of human cognition and perception, and then you really expand it out to some major issues, like the anti-vaccination movement. Was that a deliberate rhetorical strategy on your part to try and get people to think more critically about what they read about science?

Simons: The whole idea of the book was to question the intuitions you have about how your mind works, and the assumptions that you make that you know how your mind works. Everybody thinks that they know the reasons they made the decisions they did.

We had to start with the attention and memory experiments because that’s the lead-in to the book, that’s the title of the book, but the idea all along was to talk about other ways in which our intuitions about thinking, reasoning, decision-making, and confidence, go wrong. I don’t know if it was was intended specifically as a rhetorical device to get from simple to complex, it just naturally follows that way. Illusions about memory and how we think about memory naturally follow from perception, discussions of confidence tie into memory, and this idea that we think we know more than we do follows fairly naturally from that. It wasn’t intended as a rhetorical device, but I think the idea was to show that a lot of the ways we think about how the mind works could be informed by science.

But we don’t often realize that they should be."