Coolest... Experiment... Ever

Yesterday, Steve of OmniBrain asked, "What is the coolest psychology experiment ever?" Feel free to provide your own answer in comments there and/or here. As for me, there are some that I think are really cool for theoretical reasons, but the coolest ever just for the sheer implausibility of the results has to be Daniel Simons and Daniel Levins' "real-world interaction" change blindness experiment. It simply boggles the mind. First, a little bit about change blindness.

Check out this scene, and see how long it takes you to figure out what changes when the scene flashes. I'll give you a moment.

OK, back? It might have taken you a few moments, but you probably noticed eventually that the plane's engine, on the wing near the fuselage, disappears and then comes back in successive flashes. The difficulty people have in noticing changes like this is called change blindness. There's a pretty large literature on change blindness, and I don't want to get into all of that right now, because I want to get to the coolest... experiment... ever, but I will give you a few of the conclusions from the literature, from Simons and Ambinder (p. 44)1:

  1. "Change blindness occurs whenever attention is diverted from the change signal."
  2. "Changes to objects that are central to the meaning of the scene or changes to visually distinctive objects are detected more readily than other changes, presumably because observers focus attention on important objects
  3. "Attention may be necessary for change detection, with changes to unattended objects going unnoticed."
  4. "Attention to a changing object may not be sufficient for change detection; observers frequently fail to detect changes to the central actors in motion pictures and to real-world conversation partners even though these people clearly are attended suggesting that change detection requires observers to encode the changing features before and after the change and compare them"

In short, change blindness is an issue of attention and representation. If we fail to represent an object in a scene either before or after a change, then we won't notice the change, and we tend not to represent objects that aren't important to the meaning of what we're looking at, because we're just not paying attention to them (though paying attention to them doesn't guarantee representation).

OK, now for the experiment. Number 4 above gives you a hint about what happened. Simons and Levin2 had experimenters approach a passer by with a map in hand (they were on the Cornell U. campus), and ask for directions to a campus building. After the passer by had been giving directions for about 15 seconds, a second and third experimenter, both carrying a door, passed between the first experimenter and the passer by. As the door went by, the first experimenter -- the one who'd asked for directions -- switched places with the experimenter carrying the back end of the door, who took the direction-taker's place in front of the passer by. Here are photos from the paper, so you can see exactly what happened:


After the direction-giver finished giving directions (or just stopped), the experimenter now in the place of the direction-taker asked,

We're doing a study as part of the psychology department [experimenter points to the psychology building next door] of the sorts of things people pay attention to in the real world. Did you notice anything unusual at all when that door passed by a minute ago?

If the direction-giver didn't say that he or she had noticed the person to whom he was talking changed when the door passed, the experimenter then asked,

Did you notice that I'm not the same person who approached you to ask for directions?

Eight out of fifteen direction-givers failed to notice that the person they were talking to changed in mid conversation! Eight out of fifteen! Simons and Levin also noted that all seven of the people who did notice that their interlocutor had changed were students (like the experimenters were), 20-30 years old. Older participants, between 35 and 65 years old, didn't notice the change. This led Simons and Levin to hypothesize that, if the appearance of the experimenter was relevant to the participant, they would notice the change (students are always checking out other students, ya know?), but if it wasn't relevant (all college students look the same to me!), they didn't notice the change.

To test this hypothesis, Simons and Levin conducted a second experiment in which the two experimenters who acted as direction-takers were dressed as construction workers, who are common on college campuses, and generally ignored by students and faculty alike. The rest of the procedure was the same (ask directions, guys with door walk in between them, experimenters switch, and so on). This time, only 4 of 12 young participants (college age) noticed the switch, supporting the hypothesis that people only notice the change when the appearance of the individuals was relevant.

At this point, you're thinking, "This is insane! I would notice if the stranger I was talking to suddenly looked like a completely different person!" I thought the same thing, and we're not alone. In fact, this disbelief is so common that it has its own name: change blindness blindness. In another experiment, Levin et al.3 asked participants whether they would notice changes in previously used change-blindness stimuli. In the first experiment, 83% said they would, and in the second, 64% said they would, but when asked whether other people would notice it, only 21% said they would. Both 83% and 64% are well above the percentages of people who actually noticed the changes in the stimuli (21% is closer). Thus, people consistently overestimate their ability to detect changes, but they don't think other people are as good at change detection as they are.

1Simons, D.J., & Ambinder, M.S. (2005). Change blindness: Theory and consequences. Current Directions in Psychological Science, 14(1), 44-48.
2Simons, D.J., & Levin, D.T. (1998). Failure to detect changes to people in a real-world interaction. Psychonomic Bulletin & Review, 5, 644-649.
3Levin, D.T., Momen, N., Drivdahl, S.B., & Simons, D.J. (2000). Change blindness blindness: The metacognitive error of overestimating change-detection ability. Visual Cognition, 7, 397-412.

More like this

If you're a regular reader of Cognitive Daily, you're relatively accustomed to seeing surprising things. Indeed, it's gotten to the point where you might even expect it. You've seen optical illusions and videos that baffle the imagination. Yet most participants in psychology research studies aren't…
The world of someone who has a psychological disorder is different from the world of a healthy person. For example, someone suffering from clinical anxiety will notice threatening stimuli sooner, and an alcoholic will perceive alcohol-related images quicker than healthy people. These different…
Our visual system is exceptionally good at detecting change -- as long as the change takes place while we're looking. If you glance at a scene, then look away for a moment, your ability to detect a change is substantially impaired. Changes that would be obvious when we're looking can become…
Take a look at the short movie I've linked below (Click on the picture to play. QuickTime required). The movie shows a virtual gripping device (two red balls) lifting rectangular objects and placing them on a conveyor belt. Do you notice anything unusual happening as the objects are being moved…

Well, I'm thinking you're going to get a bunch of people nominating Zimbardo's Prison Study, or Milgram's obedience experiments, but I have to say, for sheer audacity and cleverness, it's hard to beat Darley & Batson 1973

BTW, this is coming from a cognitive-neuroscientist, so I think that says a lot. The social psychologists are (or were, at least) the most willing to really mess with people's heads to make a point. It's hard not to admire that .

By boojieboy (not verified) on 14 Dec 2006 #permalink

No question, you win, this is the coolest experiment. If we so often fail to recognize a person we were just talking too it may go a long way to explain the unreliabilty of eye witnesses.

By CanuckRob (not verified) on 14 Dec 2006 #permalink

Oh, I've been so all over this one for a while now. I also like the experiment where subjects met an experimenter, who gave them a form to fill out before they went in for an experiment. The person then took the form, ducked down to file it, came back up, and gave them directions to the experiment room. But, of course, a different person came up than had ducked down. A similar percentage didn't notice that they had talked to two people. I can't remember who did this one, but it definitely gets efficiency points.

By ThePolynomial (not verified) on 14 Dec 2006 #permalink

awww... look dan still has hair in those pictures ;)

Haha... funny grad school story: I was visiting Indiana U. for their "prospective students weekend," where they wowed the people they'd admitted (it was better than Illinois' weekend, I have to admit!), and I was supposed to meet up with a certain faculty member, who was going to show me around. I'd never met this particular faculty member, but I had seen his picture on his webpage. So I looked everywhere, and actually passed him 4 or 5 times, before he finally asked me if I was looking for someone. I said I was looking for, well, him, and he said, "That's me." I was in profound disbelief. I hadn't recognize him because in his webpage picture, he had a full head of hair, but in real life, not so much.

You could put the icing on this experiment's cake by having the door guy walk by during the change blindness blindness interviews. I'm guessing a small but significant number of respondents wouldn't notice an interviewer switch even then.

It's funny when you meet people for the first time- there were a number of faculty members who i expected to be a lot taller or shorter.

what was your most entertaining grad school visit?
hopkins was pretty cool b/c they put us up in a hotel and spent lots of money on us.
illinois was pretty good - lots of drinking and parties.
nyu sucked balls
hmm.. carnegie mellon was incredibly intense.

By someone you know... (not verified) on 15 Dec 2006 #permalink

My son has two friends who are twins but not identical twins. When they are here I have to be very careful to notice which one I'm talking to.

The gentlemen in the experiment do look very similar - same height, build, hairstyle, and very similar faces. On just meeting someone, you might not notice enough details to notice that the people changed places.


I have to be honest, I didn't bloody notice a thing, even after rigorously studying the hats off all the people in the picture etc...

Oh well!

There's a great example of this from British mentalist Derryn Brown, see the youtube video here:

Not a lot of similarity between the people switching either! ;-)


By Ralf Muhlberger (not verified) on 16 Dec 2006 #permalink

so according to Dan Simons - the change between people can be pretty drastic.
The thing people definitely notice is a race or sex change.

Somebody send this link to David Blaine

I had a similar experience when I was staying over at a mate's place, he got up to take a dump and I got into bed with his girlfriend.

I can't argue with numbers, what they say is what they said, but something seems odd. How would somebody NOT notice the person walking away? It is one thing to notice a difference if they were suddenly replaced, but the person had to walk away, even when visually obstructed by the door. At the very least, suddenly ducking slightly to the same stature as the door holder shown in the picture should have raised some eyebrows. The action should've made a hint, if not the appearance.

And the other problem... the voice? I guess if you got people that look and sound similar, that's okay, but even if they somehow didn't see the guy duck, they should still hear the voice as different when they talk aftwards, or even when they ask the question at the end.

By Yogurtron (not verified) on 29 Dec 2006 #permalink

Seen this before and it was interesting at first glance. Even if there was some concious change noticed it would probably be passed off as a mistake and overlooked, since the chances of two people looking for a map and still continuing the conversion are so slight and makes no sense.

The experiment now reminds me of a funny commercial where a drill sargent notices a stain on a soldiers clothing, however the soldier has a special cleansing marker that removes it so quickly that when the sargent turns around again he is shocked and proclaims there is a houdini master in front of him.

For those of you that have seen it, I guess the sargents reaction is what the experimenters were hoping to see or would want us to play out in the map change scenario :)

Yogurtron read it again, and look at the pictures, then you will understand how they did it.

Not only does this make sense to me, it is a *good* thing. That is, as I get older I get better at figuring out what matters and what does not. If you pay attention to everything, if you give it all equal attention, then you are overloaded with the trivial. Someone asks me for directions, I tell them. Why do I care who asks?

I am on the side of the people who think that is really shows what is all wet about psyche today. This so called experiment proves nothing other than that people can be fooled by "slight of hand". We don't need a psyche experiment to do that; go to a magic show. I just loved the term "change blindness". What a laugh riot. And then I wonder where the Bush adminstration gets its BS terms....

As a nit-picky detail: if the the participants were asked if they had noticed anything "unusual" then the study was flawed from the get go. "Unusual" is a loaded word. They were not testing people's awareness of change but people's defination of unusual, which may or may not fit the experimenters defination. The proper terms should have been disceprency (though most college students might not get that word) or simply "different".

It really sucks to see such poor experimental design displayed on the web as something to look up to.

Daniel -- I doubt most college students would get the word "disceprency," since it's not a word, at least according to my dictionary. I do however think most college students understand discrepancy, or inconsistency.

I actually agree with your point, at least in part. In my humble opinion, continuing a conversation after an interruption with a completely different person would count as unusual. Even the same person wearing glasses or a hat, when he wasn't prior to the interruption, would be unusual by definition. It's not usual or common for the person you're talking to to change in the space of a few seconds. It doesn't happen every day, or ever, really. Therefore, it is unusual.

Regardless, I don't think that one loaded word makes for "poor experimental design."

On the other hand, discrepancy would have been better.

In my 6-Sigma Greenbelt class they showed us a video and told us to watch for the number of times the ball was passed.

The video showed about ten guys in black and white shirts dogdging and weaving around each other, passing basketballs.

After the video, we were asked "Did you see the gorilla"?

When the video was shown again, a very obvious guy in a gorilla suit casually walks through the frame. Very few of us saw it.

Daniel: The question is how does sleight of hand work? If we truly noticed everything, we would never be fooled by sleight of hand tricks.

There was a comedy sketch on British TV in the late 80s or early 90s that showed the changing door sketch, it was very funny.

By Barry Graham (not verified) on 02 Aug 2009 #permalink

I'm really interested in this, and I see I found this article/site a little late, but all the links are broken. Can anyone update? I will if I find the stuff.