Gorilla psychologists: Weird stuff in plain sight

http://www.newscientist.com/article/mg20627660.900-gorilla-psychologists-weird-stuff-in-plain-sight.html


Gorilla psychologists: Weird stuff in plain sight
• 23 June 2010 by Liz Else
• Magazine issue 2766.


The "gorilla in our midst" psychology experiment is up there among the world's most famous. But as Liz Else found out from Christopher Chabris and Daniel Simons, the psychologists who devised it, exactly how it fools half of the people who take part is still a mystery

How did you come up with the experiment?

Christopher Chabris: We didn't say, "let's do a really intriguing experiment people will talk about for years". It was just a class project on visual attention in a course we were teaching 12 years ago. The gorilla suit was lying around in a lab. If it hadn't been there, who knows?

Daniel Simons: Our study revisited work from the 1970s by Ulric Neisser, where subjects had to watch a video and count the times players passed a ball. Neisser had someone with an open umbrella walk through the game, and many people didn't notice it. But his video had an odd, ghostly appearance which gave people an excuse for missing the person with the umbrella. We filmed the entire game with a single camera so that everything was fully visible - and the person in the gorilla suit was there for 9 seconds! When we showed the video, only half of the viewers saw the gorilla.

Is it the same for every group?

DS: The first subjects were Harvard University students but it worked as well with everyone we tested. For years, whenever I showed the video, I held my breath, thinking everybody would notice it. It took years before I could discount that gut instinct. Missing the gorilla is jarring. It's natural to assume that you would see it, so it's surprising and compelling when you realise what you've missed.

So why do people miss the gorilla?

CC: It's like a Rorschach (inkblot) test for cognitive abilities or personality: people think there must be something different about the people who see the gorilla compared with those who don't. Some speculate that if you are in a detail-oriented job, you are going to notice it because you notice everything, or you are not going to notice it because you are really good at focusing. But the truth is, so far no researcher has found anything that solidly predicts who is going to see it and who is not.

DS: We have looked at basic measures of attention and memory - how much you can hold in mind while doing something else, and how much you can take in with one attentional glance. These basic measures predict how well you can focus attention and count the passes, but they don't seem to predict whether you will notice the gorilla. I have just done a follow-up to the gorilla study, which I showed to about 1000 vision scientists who knew about the original video. As the gorilla enters, I introduced two changes: a backdrop curtain changes from red to gold, and one player wearing a black shirt leaves the court. The vast majority missed both, despite knowing the video was about unexpected objects. When they started looking for a gorilla, they missed other unexpected events.

Is there an evolutionary reason to miss things?

DS: I think that's the wrong way around. These failures of awareness are more a consequence of something that we need to do and that we do well - focus attention. To do any task, you need to focus on things that matter and avoid things that don't. One consequence is that sometimes you filter out things you might want to see. Intuitively we think that things that matter will catch our attention, but they don't.

CC: In our book (reviewed, right) we tell the story of a Boston police officer chasing a suspect. When you do that, you're paying careful attention, figuring out where he's going, if he's got a gun or is throwing evidence away. The officer ran past an incident of police brutality taking place close by and later claimed not to have seen it. He went to prison because the jury didn't realise the extent to which focusing on one task makes you unable to see outside that. They decided he lied to protect fellow officers.

So our picture of reality can be very wrong?

DS: Our picture of reality is correct most of the time for most of what we do. It is inherently incomplete, though, and most of the time we don't realise how much we miss. Usually, it doesn't matter - we see what's relevant to what we're doing. It's important to know that we have such limitations, though. We think we see more than we do, and that has consequences. We also think our memories are more perfect than they are, that we understand complex systems better than we do. If we were aware of our limitations, we wouldn't text and drive or think everybody who mis-remembers is lying.

Can you devise another experiment this good?

CC: The statistics are against it! But who knows what we'll find lying around tomorrow...

Profile
Christopher Chabris is a psychology professor at Union College in New York; Daniel Simons is a psychology professor at the University of Illinois at Urbana-Champaign. The "Gorillas in Our Midst" study is at bit.ly/bg2rI7