Last Friday, we asked readers to identify changes in a painting by Eugčne Delacroix, Christopher Columbus and His Son at La Rábida. For the test, the picture flashed maddeningly between its original state and a doctored version Greta made in Photoshop. Here’s the original:
We admit it — we were being a little sneaky. We asked you to “identify the change” in the picture, subtly hinting that only one part of the picture would change. You can click on the picture to see the movie again and convince yourself that two things are changing in the picture: the arch in the center transforms to a squared-off doorway, and the two robed figures to the left of the doorway disappear.
Greta has run a modified version of this experiment in her lab; there the transformations were tested separately — the arch changed in one task, and in a separate task, the people disappeared. In her experiment, it took participants an average of 6.6 seconds to detect the disappearing people, and 10.4 seconds to notice the arch changing. Since designing an online environment to test reaction time requires a substantial bit more effort than we were willing to exert for a Casual Friday, we split the difference and gave you 9 seconds to see what you could spot.
So, what were the results? Our first question simply asked “What items did you notice changing in the painting? (you may choose more than one)” Aha! your first clue that we were up to something! Here’s a summary of the responses to that question:
As you might predict based on Greta’s preliminary results, significantly more participants noticed the people changing than the arch. Only 23 percent of respondents noticed both changes. Interestingly, our distractors did manage to fool people: 7.6 percent of respondents indicated they saw a change which was not present in the animation. We even got one person to indicate that the “cat” changed, even though there was no cat in the picture. However, significantly more people noticed the arch changing than were fooled by any of the distractors.
We also thought there might be age or gender effects. There were none, even though we received a disproportionate number of responses (35%) from the 20-29 age group (I think this was due to a link from BoingBoing late Friday evening).
One area where we thought we might have noticed an effect is when people noticed both the arch and the people changing versus just one of the other. This graph charts the portion of respondents who indicated that they had seen one of the distractors:
There is a trend toward people who noticed both changes also being fooled by a distractor more often than people who noticed only the arch or the people changing. However, this difference was not significant.
Where this Casual Friday was an overwhelming success was in the response rate: we collected our maximum of 250 responses by mid-day Saturday (again, I think we can thank BoingBoing for that!). Look for a new Casual Friday survey later this afternoon.
Do you have any other explanations for our results? Did you have a different experience than what the results indicate? Let us know in the comments.