In my last post, I had a short quiz. If you haven’t taken it yet, do so now! Once you look at the results, your own answers won’t count.
The quiz contained several simple images:
The questions asked whether the left-hand object in each image was, respectively, longer, larger, or lighter than the right-hand object, or the other way around — or whether the two objects were actually the same.
To the uneducated eye, the answers to the quiz are obvious: (1) the line on the left is longer than the line on the right; (2) the inner circle on the left is larger than the inner circle on the right; (3) the inner circle on the left is lighter than the inner circle on the right.
However, there is a class of simple optical illusions that exploits certain contextual clues to trick our brains into making the above “obvious” conclusions, when in fact they are false. In such illusions, all of the compared objects are actually the same with respect to the specified property, but are made to look different. For instance, in image 1, the use of inward-pointing arrows serves to increase the apparent length of the first line, whereas the use of outward-pointing arrows serves to decrease the apparent length of the second line.
The first striking, and somewhat disappointing, result from the quiz is that these illusions are well-known: 27 of the 28 people who took the quiz reported having seen them before.
The second notable result is the uniformity of the answers.
- 88% of the answers were “same”
- 12% were “left”
I suppose these results would have been more interesting if more people hadn’t already seen such illusions. I also primed the quiz as follows:
If you are curious about how your brain can jump to incorrect conclusions based on what you think you know, take this 30-second quiz.
This introduction probably got people on their guard. No doubt when they recognized the questions, they realized that “what you think you know” must refer to the contextual optical illusion created by the image, and were suitably prepared to answer correctly.
That was a bit of misdirection, however, especially in the context of a post (only tangentially related, I’ll admit) that discussed the over-confidence sometimes caused by the knowledge of new information.
In reality, “what you think you know” referred to your foreknowledge of such illusions. In fact, in the above images:
- The left-hand line is 10% longer than the right-hand line.
- The left-hand inner circle is 10% larger (by area) than the right-hand circle.
- The left-hand inner circle is 10% lighter (by rough RGB estimate) than the right-hand circle.
This is actually true, as I made these images myself for this quiz. You are welcome to bust out Photoshop to check. Alternatively, a clear mind and a quick second look at the images above should convince you of the fact that the answer is “left” for each of the three questions. So only 12% of the answers were correct.
I increased the trickery a bit by adding a 7-second timer to each question. If you recognized the illusion — and assumed that the quiz was designed for people who did not recognize it — you’d likely assume that the timer was present to prevent such people from having the time to analyze the image and overrule their initial judgment (to select the left-hand object) in favor of the correct judgment (to select “same”).
In fact, its purpose was exactly the opposite: to prevent illusion-recognizers from having the time to realize that the left-hand object was clearly correct, knowledge of illusions be damned.
Some more fun facts:
- Even the one person who hadn’t seen the illusion before guessed “same” for all three questions.
- Only 3 people of 28 guessed “left” twice of their three answers.
- No one guessed “left” all three times.
Why did I do this experiment? I didn’t have much impetus other than the fact that when I thought of it I realized I’d fall for the trick and was wondering whether anyone else would, too.
I’m not sure what can be learned from it, either, given the host of confounding variables. Maybe something about how easy it is for book-learning to trump our gut feelings.
If anything, this quiz proves that there are some legitimate tests on which I, and most people who read my blog, would do worse than a first grader.