We're shockingly bad at noticing changes even to familiar scenes

ResearchBlogging.orgTake a look at the following pictures of U.S. dimes. As you can see, they are slightly different from one another -- the date is in the incorrect spot on one of them. Can you tell which one is "wrong"?

i-48027c9418fc26955cf7a33263aa604f-dimes.jpg

Let's make this a poll:

Don't look at your pocket change before you answer! In case you don't have a dime handy, I'll reveal the correct answer later in the post.

Even though most Americans will say they know what U.S. coins look like, a similar study in 1979 found that people can't remember the basic details of a penny. More recently, change blindness studies have shown that we are very bad at detecting changes in scenes, even those that seemingly take place before our eyes.

But Luke Rosielle and Jeffrey Scaggs point out that change blindness isn't much of a problem in the real world because things don't ordinarily disappear or change right in front of our eyes, or in the moment when we glance away. A much more common type of change happens when we've been away for a longer period of time. If you leave town for a few weeks, you might be likely to notice that your favorite coffee shop has been repainted. This is the sort of change we may be more likely to notice. Or are we?

Rosielle and Scaggs showed 48 students pictures of their own college campus, and told them that half had been photoshopped to remove or change prominent campus buildings and monuments. The students carefully observed each picture for 20 seconds, then said whether the photo was accurate or modified. After each photo, they rated their familiarity with the scene on a scale of 1 to 10. On average, the students were familiar with 97 percent of the scenes. However, they failed to identify the changes to 81 percent of the photos!

So even though the students said they recognized the scenes, they flopped at actually noticing what had been modified. Why? Rosielle and Scaggs showed the same scenes to 48 new students from the same campus, but this time they were shown the original and altered pictures side-by-side. These students were asked how difficult it would be for others to identify the changes in the pictures. Interestingly, their ratings matched the errors made by the first group of students: they could predict how good other students would be at identifying changes at a rate significantly better than chance. That said, they still weren't very good at predicting: they thought students would get about half the answers correct, when in fact they missed over 80 percent!

The researchers showed the same pairs of pictures to 48 students from a different school, who had never seen the original college campus. These students were unable to predict how well students from the original campus would do; their predictions bore no relationship to the actual results.

So it seems that while our memories of scenes aren't as good as we think they are, the memories are indeed better than nothing. The students from the different university tended to rate the larger changes (those occupying the most pixels on the screen) as easier to spot, but the students who actually attended the school recognized other features as more likely to be noticed. This makes some sense -- you'd probably be more likely to notice if your favorite coffee shop closed down than if the same thing happened at a larger place you never visit. But it's striking that even very familiar places don't actually stick very well in our memories at all.

If this is the case, then we should expect that our readers didn't do very well on the two polls above. For comparison, here's an unaltered photo of a dime:

i-8bcf5e8d15890451a19826be0f115065-dimeorig.jpg

As you can see, dime B had the date in the correct spot. But both dimes were missing a very large feature: the word "LIBERTY" to the left of Roosevelt's face. Did you notice all these changes? Let us know in the comments.

Rosielle, L., & Scaggs, W. (2008). What if they knocked down the library and nobody noticed? The failure to detect large changes to familiar scenes Memory, 16 (2), 115-124 DOI: 10.1080/09658210701787765

More like this

Got 'em both right. I realized the LIBERTY was missing - wasn't quite sure where, but realized it wasn't there.

Interesting post. Thanks

By Dave Gill (not verified) on 24 Jun 2009 #permalink

Got 'em both right. I realized the LIBERTY was missing - wasn't quite sure where, but realized it wasn't there.

Interesting post. Thanks

By Dave Gill (not verified) on 24 Jun 2009 #permalink

It's quite interesting. Having only visited the US once since infancy, I'm deeply unfamiliar with your coinage and had no idea where the date ought to be. But something looked wrong and unbalanced in both A and B. The missing "LIBERTY" explains that.

By Hilary PhD (not verified) on 24 Jun 2009 #permalink

Both right also. For the date placement, the wrong one doesn't have the date aligned with the text on the left.

Got the Liberty, missed the date.

By ctenotrish (not verified) on 24 Jun 2009 #permalink

"But Luke Rosielle and Jeffrey Scaggs point out that change blindness isn't much of a problem in the real world because things don't ordinarily disappear or change right in front of our eyes, or in the moment when we glance away."

This actually is not quite true. Things change every time we blink our eyes or a truck drives by obscuring our view. It usually isn't identity though but location.

By Steve Higgins (not verified) on 24 Jun 2009 #permalink

"But Luke Rosielle and Jeffrey Scaggs point out that change blindness isn't much of a problem in the real world because things don't ordinarily disappear or change right in front of our eyes, or in the moment when we glance away."

This actually is not quite true. Things change every time we blink our eyes or a truck drives by obscuring our view. It usually isn't identity though but location. In most circumstances we use our knowledge about the world to guess whether things are in the same location or not.

By Steve Higgins (not verified) on 24 Jun 2009 #permalink

Cool. I knew the mint mark was above the date for a dime; it just fits better that way in the design. So I got that info correct. I *thought* there might be something wrong with the images, but not enough to say so in the poll, since I couldn't identify it.

Got'em both.

Not having seen a "dime" in my life I chose the one that looked typographically correct. That's the wrong one. Next we'll photoshop some Euros and let the world decide if our aesthetics is better than the designer's (who looked at feeling too)?

By Henk Poley (not verified) on 24 Jun 2009 #permalink

I overlooked the missing Liberty (insert political snark here), despite the fact that I used to count money for a living. I got the position of the date right, but mostly because it looked wrong to put the 4-character date above the 1-character location, where the space is narrower.

"Not having seen a "dime" in my life I chose the one that looked typographically correct. That's the wrong one."

Well, I've seen a dime, and still picked the typographically correct version. I think the question about whether there's anything wrong is biased, that most people would say yes to that question even if they didn't see anything wrong, just 'cause they figure there must be.

By neurolover (not verified) on 24 Jun 2009 #permalink

Clearly there is a fake "In God we trust" on both coins. That can't be real since it violates the separation of Church and State laws of your country.
Doesn't it?

What makes the surface of a dime a "familiar"? My dimes tend to spend most of their time in my pocket. I have no idea when the last time I looked closely at a dime was. Maybe when I was a little kid.

What makes the distant skyline from exactly that perspective "familiar"? In the study they used all photos taken from the exit of the psych building. How often are students standing in just that spot attending to distant buildings?

Now, when I fail to notice my wife's haircut... but I didn't need science to prove I'm a doofus.

By jrshipley (not verified) on 24 Jun 2009 #permalink

No idea, I can't play this week, I'm not american and we don't use your money.

The main thing that jumped out at me was that something was missing on the left side of the coin, though I couldn't think what it was without looking at one. I might easily have overlooked the wrong placement of the date if I'd just been looking at the wrong coin by itself, but as my attention was called to it, it was easy to pick out the right one. (I used to collect American coins, so I have some vague idea of what they're supposed to look like, but designs change and memories fade.)

I think there is a psychological bias to say the dime is changed. I was inclined to say something was different before I even saw the picture of the dime, just because it didn't seem likely you'd post a picture of an unchanged dime. That being said, I did have a vague feeling something should be at the top as well, although I definitely did not remember it was liberty, and felt it was closer to the top and not so far to the left...

Not American, never been to the US, I don't remember handling the money before either. I guessed B (but didn't vote) because it makes sense that the P would be above the 2003 as it fits better. It's tidier. The poll results are interesting, more people that voted are getting the coin wrong, but most people realise that something else is missing.

By Katherine (not verified) on 24 Jun 2009 #permalink

Just for the fun of it I asked my grand-nephew the questions when he came by. He picked the wrong image as correct on the date placement question, but was quite sure there was lettering missing in both pictures.

I got them both.

By curiouser_alice (not verified) on 24 Jun 2009 #permalink

This phenomenon of 'change blindness' is exactly why we are blind to shifting environmental baselines. We don't remember what the air looked like thirty years ago, other than a subjective impression. Anything quantitative (like whether you could see the glacier on the mountain from downtown, or when a certain species of migratory birds was likely to appear each year) gets all fuzzed up along with the air quality.

Recently I made a close-up photo of a dime for a post. But I still got the P wrong, and only vaguely knew that there had to have been something in the space where "Liberty" goes.

Didn't get the date right at all. I knew the Liberty was missing, but didn't realize it was all caps, and wanted to put it in the wrong place.

The A option wasn't photoshopped perfectly. There's a significant color change between the year and the open space.

I figured correctly for both, though it is harder to say whether the basis for choosing B was based on correctly remembering the detail or because the color change appears a bit too sharp in contrast (I took a semester of electronic paint, learning photoshop and such).

Last time I've looked at an up close picture of a dime as far as I can tell was when I was about 7 and we had Apple computers in the computer labs where the monitors displayed green and black and there was a program where we added denominations of money and that was when my dad volunteered at the school, so about 12 years ago.

Of course I've seen dimes around since then, but not really examined them other than flipping them in the air to see them reflect light and be shiny. I do tend to have a very good memory for small details (though not entirely consistent - I am much more inclined to remember minute details of insignificant moments such as clothing, smells, activities, everything, and I am also inclined to remember facts of biology and physics, for instance, whereas historical fact tends to not stick so well, though I seem to remember a fair bit from AP European History and my community college US history class I took since I got a D in the high school US history, though I don't recall nearly enough of the small details of European history that the score of 4 would suggest, and that is a bit broader than the topic of distinguishing changes in familar things anyway).

I'm Australian. I thought there should have been something where the word liberty is. of course I had no idea where the date should be.

By Xeridanus (not verified) on 24 Jun 2009 #permalink

I have never been to the US so I thought the coins were fakes. Do you really use coins? We use our metro card and there is not need to carry any change at all. This might help fix your problem...

what's missing on the US Dime is something to indicate the value of the coin. (It doesn't actually say "10c" anywhere on it). Neither does the 25c coin indicate it's value, although it does give a clue by saying 'quarter'

Are they the only coins in the world that don't have the value clearly marked?

I don't know if this really says we're terrible at detecting changes or if we just don't look closely at certain common objects.

For instance, take the font, Times New Roman. One of the most popular fonts in the world, we all see it every day.

So tell me this: in Times New Roman, does the "cross" part of the number four stop at the "spine," or does it intersect it?

#27, we do indeed use coins quite often, for any exchange of physical currency involving amounts less than one dollar. Every now and then there's some discussion about phasing out the $0.01 penny and $0.05 nickel, but so far they have yet to amount to anything. We've also tried adding a $1 coin but that's never caught on either (perhaps because people would rather have a wallet full of bills than a pocket full of coins?). Pennies and $0.25 quarters have some cultural significance for us as well - finding a stray penny is considered lucky and the quarter is the definitive "coin" to us - so I'm not sure we'll get rid of either any time soon.

On the plus side, they're a cheap and reliable source of two-sided dice. :)

While this shows we don't know a couple small details I'd encourage everyone to see the amazing things that we CAN remember for many many years:

T. F. Brady, T. Konkle, G. A. Alvarez & A. Oliva (2008). Visual long-term memory has a massive storage capacity for object details. Proceedings of the National Academy of Sciences, U. S. A., 105 (38), 14325-14329.

http://visionlab.harvard.edu/Members/George/abstract-015.html

By Steve Higgins (not verified) on 25 Jun 2009 #permalink

You should have another poll at the bottom that asks if you thought there was something different, was it really the added 'Liberty' text. I thought In God had more space between the two words....

I actually got both right, but I'm not sure I would have if the photoshop job was better. Not saying I'm an expert, but obviously the top shadows in 2007 are not correct on A. Actually, now that I pulled out a tiny dime and looked at it - who would consider this a familiar object in the first place. Same goes for the college campus - I'd consider myself familiar with it, but I wouldn't be able to tell you if a building was missing that had been there two decades ago. Take a lug nut off my car that has been parked at the airport for a month, and yeah, I'm more likely to notice that than not, but perhaps less likely to notice it than if had been painted a different color instead. With no offense intended, this study/test seems kind of silly and pointless.

Ha! I got the liberty right, but not the date. I was expecting it to be like the date on the penny.

This makes me wonder how long it's been since I've last carried cash around with me, let alone pocket change. Even the image of the dime at the end of the entry didn't look familiar.

While I rarely pay attention to the coins I use (I just look at the numbers), I did notice that Coin A had a noticeable flaw in the photoshopping: where the "2" in "2003" once was, there is a small fragment of "2" left, and that's what gave it away.

And I didn't know there was even a "LIBERTY" on the coin; I just said "Yes" because I noticed the flaw earlier -- which probably isn't what you had intended.

Got em both - despite the fact that the altered one looks better. Really it was only that fact that kept me thinking for a minute because I knew the right answer as soon as I saw them. I collected coins as a kid and really loved em which I suppose gave me an advantage. Oh, and though I knew that there was another error (which I assumed to be some missing word or words) I didn't automatically come up with "liberty", numismatic hobbies notwithstanding.

mnuez

Didn't get the second one. I guess details on money aren't valuable enough for me. At least, not after I realized what denomination it is!

This is the focus of a children's birthday party game, where you take two trays of objects, and remove several of the objects from the second tray and let the children guess which objects are missing.

By Anonymous (not verified) on 08 Jul 2009 #permalink

what's missing on the US Dime is something to indicate the value of the coin. (It doesn't actually say "10c" anywhere on it). Neither does the 25c coin indicate it's value, although it does give a clue by saying 'quarter'

Are they the only coins in the world that don't have the value clearly marked