Casual Fridays: Why most people didn't see this illusion

Just about two weeks ago, I posted this visual illusion (if you haven't seen it yet, make sure to watch it with the sound turned ON):

How many flashes do you see?

In fact the dot only flashes once, but according to the study I report on in the post, the two beeps are supposed to throw you off. If you are like Shams et al's participants, you would most likely see two flashes, not one. In fact, just 23 percent of CogDaily readers reported seeing two flashes.

Much speculation ensued in the comments, but I had a couple ideas of my own, so last Friday I developed a quick study to test two possible explanations. The first explanation is that there is a practice effect: the more you see the illusion, the more likely you are to experience it. The second explanation is that clicking "play" distracts you from the display, and you aren't focusing enough to experience the illusion. My study tested both explanations.

Participants were divided into two groups; each group would see the same 10 displays, but one group would have to press play each time, while the other group would have the displays play automatically. The actual illusion was only shown three times; the others were distractors where the dot flashed twice, or there was one or no beeps instead of two. So, was there a practice effect? Take a look at this chart of the results:

Doesn't look like a practice effect. The number of flashes seen remained quite steady whether the illusion was seen at the beginning, middle, or end of the experiment. So practicing with the illusion doesn't appear to affect whether it's seen: either you see it or you don't.

But what about pressing "play"? If you have to play the movie yourself, are you less likely to see the illusion?

Apparently you are! An average of 1.37 flashes were seen when the movie was self-played, compared to 1.55 when it played automatically. Interestingly, when we initially ran the poll on the illusion, 1.31 flashes were seen, so our slightly more formal study matches the initial poll quite well.

So why don't viewers see an average of 2 flashes, as in Shams et al found in their original study? I have a couple more thoughts. First, in the initial study, viewers were told to focus on a point some distance away from the site of the dot flashing; this is because the illusion is stronger when viewed with peripheral vision. Second, there may be some hardware / software difficulties. In the lab, Shams' team could ensure that everyone saw the same illusion, but online, it's much more difficult to control for everything.

One commenter thought that perhaps my instructions about having the sound turned up weren't clear enough, so halfway through the experiment I added more explicit instructions. Was there a difference in the results after I improved the instructions? No.

Another commenter suggested that the refresh rate of users monitors may differ. Indeed, Shams' web site offers several different movies to account for this possibility. But there could be other problems as well: distracted users, software incompatibility, or even taking the test in a noisy room could affect the results.

So what's the ideal movie for viewing the illusion? Barring hardware and software problems, it should force viewers to focus away from the center of the visual field, and it should play the movie automatically. The link below should take you to such a movie. Just make sure you focus on the cross when the movie starts to play:

Click here to try it!

And, of course, don't forget to come back next week for another Casual Fridays study.