Change blindness is a truly remarkable phenomenon. There are so many ways that the human perceptual system can be tricked into missing a change that appears right before our eyes, that it’s sometimes astonishing that we aren’t constantly running into walls or misplacing the basics of life — our car keys, wallet, our what were were planning on eating for dinner. If you’ve never seen a demonstration of change-blindness, I’d suggest checking out some of the posts we’ve written about it before (For example, here, here, and here).
So why don’t we notice these sorts of changes? This video, based on experiments conducted by Andrew Hollingworth, can help us sort through some of the issues. You’ll be shown a “scene” made out of Legos, and then the scene will disappear. When it returns, one item will by highlighted with an arrow. Next, that scene will be replaced with a slightly different version, where the highlighted object is changed in some way. Your task is to indicate which of the second two images is the same as the first image you saw:
If you’re like most people, you won’t find this task very difficult. Even though people tend not to notice these sorts of changes in a scene, when the object that changed is pointed out to them, they often are able to identify what changed. So what does make the problem more difficult? Hollingworth showed volunteers two versions of this type of video. In one version, an object was rotated or substituted with a similar object, just as above. But in another version, the entire scene except for the relevant object was removed, and again, viewers decided which object was the same as it was in the original scene. Here are the results:
In every case, the viewers performed significantly better than chance. But when the original background of the scene was intact, viewers did better still, whether they were determining if an object was rotated or substituted with a similar object. It seems that the background of the scene is helping us in some way as we look for changes changes. But how?
One possibility is that the scene is retained in our short-term memory, helping us to spot changes in the object. To test this, Hollingworth repeated the study but interleaved the scenes and tests, so viewers saw a completely new scene before they were tested on an older one (i.e. Scene 1, Test 0, Scene 2, Test 1, Scene 3, Test 2, and so on). Previous studies have shown that this exceeds the limits of visual short-term memory, so Hollingworth was testing long-term memory. Here are the results:
It’s practically identical: Even for long-term visual memory, hen the background was removed from the image, memory for the object that changed was significantly worse. So seeing objects in the context of a scene improves both long- and short-term memory for the objects’ identity and position.
It seems that the memory for the individual objects in a scene is somehow connected to the memory for the scene itself. Remove objects from that context, and it becomes significantly more difficult to remember the objects. Hollingworth designed a third experiment to try to uncover why the scene helps with memory. We’ll discuss those results tomorrow.
Andrew Hollingworth (2006). Scene and Position Specificity in Visual Memory for Objects. Journal of Experimental Psychology: Learning, Memory, and Cognition, 32 (1), 58-69 DOI: 10.1037/0278-7318.104.22.168