Perception test - thread now open for comments

[rkawakami]

I was going to put this post into this thread: http://forum.thinkpads.com/viewtopic.php?t=67843

but thought it deserved a topic for itself. As that thread was dealing with driving and cell phone usage, this might be a good example about attention issues that was passed to me a few days ago.

Before clicking on the link below, this is your task: There are two groups of people in the video bouncing basketballs around. One team with white shirts, another team in black shirts. They are circling around each other while at the same time passing their team's ball. You are to count the number of times that the white-shirt team passes the ball. A pass can either be by bouncing the ball to their teammate or throwing it.

Please only run the video one time and then make your poll selection; no fair stopping and starting the clip!. Studies show that there may be a gender-bias in the results of this test.

http://viscog.beckman.uiuc.edu/grafs/demos/15.html

(requires a browser with Java and is about 6MB in length)

edit: I should clarify that you can submit a poll vote, but would ask that you not post any comments about the video clip here until Oct. 9th (two weeks from now) when I'll then ask for comments. Thanks!

MOD EDIT: I am sticky'ing this for a week. Since we cannot comment on it, it would roll right off. -tff

[rkawakami]

Well, I forgot to bump this on the 5th but we'll see what the 13 voters have to say about this little test.

For those of you who viewed and voted in this poll, did you notice anything strange going on while you were counting the basketball passes? And please be honest .

I got a version of this clip several weeks ago in one of those usual intra-company emails. I was extremely surprised that I missed noticing something that happened while those people were weaving around and throwing two basketballs around. It was one of the ways that the researchers were trying to point out how "inattentional blindness" can occur. It basically means that one can be concentrating on some task at hand and completely miss something else going on right under their nose. This experiment won an Ig Nobel award for Psychology in 2004.

The recent railway accident in Los Angeles is one of the latest real-world examples: A train engineer was texting on his cell phone 20-some-odd seconds before ramming head-on into another train. He apparently missed seeing at least one red light indicating that he should stop.

BTW, the actual count for basketball passes doesn't really matter. Nor was there any gender bias noted in the experiment. This was the original mis-direction provided by the researchers as a red herring.

[spuddog]

Well I noticed an ape wandering through, but I thought that was just to distract you from counting.

Scott

[basketb]

Yes, the ape was pretty obvious. At first I thought it's another player for the black-shirt team and some more new player(s) + ball(s) for the white team would follow to make it more difficult. However, I'm wondering whether there is something else I/we may have missed.

In any case, how many passes within the white-shirt team were there? I counted 15 but am in the minority apparently (I know that is secondary but I'd still like to have a definitive answer.)

[Marin85]

Nice experiment! I think (and might be wrong) that there is a tiny technical glitch in this experiment, though. People often forget things (like the ape) they have noticed, mainly because of the information (over)flow they have to deal with in similar situations. So, how could the researchers in this project really find out who had noticed what after the experiment had been completed

[Marin85]

how many passes within the white-shirt team were there? I counted 15 but am in the minority apparently

Then take a look at the majority If we assume that people generally tend to get confused (or even puzzled) in complex situations like this one, there is probably no definitive answer to your question

[sparta.rising]

Here's a better produced version:
http://www.youtube.com/watch?v=Ahg6qcgoay4

[rkawakami]

I believe that the majority did get the count correct (14). At least that's how many passes I counted after viewing the clip at least 6 or 7 times. More such clips are available here:

http://viscog.beckman.uiuc.edu/djs_lab/demos.html

but once you understand the main purpose of these tests then it loses some of their effectiveness. With some of these experiements I equate it to catching continuity errors when watching TV or movies. Invariably, something doesn't match up when cutting from one scene or view to another. On the first viewing you may not notice it, but after repeated viewings or being told what's different, then it sticks out so much that you can't believed you missed it.