Take a look at the QuickTime movie below. It will show a still image for 10 seconds, then a blank screen. Then it will show you the image again. Your job is to look for a detail that has been changed between the two images.
Most people have difficulty with this task. Even when the part that changes is central to the image, accuracy is typically no better than 50 percent. For the particular type of change depicted in this movie, accuracy averages less than 30 percent. If you didn’t notice the change, drag the slider in the movie quickly back and forth and you should be able to spot the change.
This phenomenon, “change blindness,” has been the focus of a considerable amount of research (we’ve reported on it several times before). Researchers have found that changes to “central” objects are easier to spot than “marginal” objects, but there is disagreement about what makes an object central. Is it some physical characteristic of the object itself? Or is it a cognitive process that viewers actively apply to the object?
Pauline Pearson and Evelyn Schaefer have designed an experiment to address this question. They pre-tested a set of 40 photos, asking volunteers to rate changes between the pictures as marginal or central. All the photos were street scenes, so they asked a second set of viewers to rate the changes as relevant or irrelevant to driving. (For example, moving a crosswalk might be relevant, but removing a building could be irrelevant).
For the change blindness task, they displayed a picture for 15 seconds, then a black screen for a half second, then the altered picture for another 15 seconds. This was a different procedure from other change blindness studies we’ve reported on, where the two images flashed repeatedly. Pearson and Schaefer chose this method so they could evaluate whether viewers were better at identifying objects which moved, or objects which disappeared (if pictures flash repeatedly, there is no way to know if observers are noticing the object appearing or disappearing). They found that viewers were less likely to notice central objects relocating than disappearing.
For the next part of the study, they gave different instructions to two sets of viewers. One group was asked only to look for changes between the pictures. The second group was told that they were helping to determine if the set of pictures was effective for evaluating driving ability. Here’s how the first group responded:
As expected, this group was more accurate at identifying central changes compared to marginal changes. They were also better at identifying changes related to driving. Now take a look at the second group’s responses:
For this group, as long as a change was relevant to driving, they were equally accurate at detecting central or marginal changes. Their ability to detect irrelevant changes was similar to the first group, but they were significantly better at noticing marginal, driving-related changes.
For central changes, the meaningfulness of the change does not appear to matter — viewers are no more accurate detecting the change when focused on the driving task. But for marginal changes, the cognitive focus of the viewer has a large effect on their ability to notice change.
I’d add that this result may be a partial explanation for the results of studies on driving with cell phones: as the cognitive demands of a phone conversation increase, driving performance suffers. The brain appears to have a finite capacity for multi-tasking; as the number and complexity of tasks we try to do increases, our performance must necessarily suffer.
Pearson, P.M., & Schaefer, E.G. (2005). Toupee or not toupee? The role of instructional set, centrality, and relevance in change blindness. Visual Cognition, 12(8), 1528-1543.