Selective attention video poll

[esldude]

 

 

 

This is a less than 90 second video testing how selective your attention might be.

 

There is a question at the very end of the video. Simple yes or no.

 

So please vote in the poll as to how you answered the question at the very end.

 

If you have seen this before, please hold off commenting.

[Jud]

 

 

 

This is a less than 90 second video testing how selective your attention might be.

 

There is a question at the very end of the video. Simple yes or no.

 

So please vote in the poll as to how you answered the question at the very end.

 

If you have seen this before, please hold off commenting.

 

Seems odd to me anyone could have answered differently.

[esldude]

Me too. But supposedly in a large group where this was first done, 44% did answer differently. I thought it a good thing to post here and see what numbers we get.

[Paul R]

I missed one of the basketball passes, but certainly didn't miss what the last question was asking about. In fact, I think I missed the pass count because I was laughing.

 

-Paul

[esldude]

So at this point 46 people have viewed at least this post if not the video, and only 4 votes?

 

The video in its entire length is only 1 minute 22 seconds.

 

Come on now, this is an anonymous poll. Please vote.

[The Computer Audiophile]

Interesting.

 

I answered No to the last question.

 

When counting I added an additional pass because it looked faintly like an extra pass was made to the girl but I couldn't really tell because the guy got the ball so quick. "Something" caused me to screw up at that point but I had no idea what it was until the end.

[spdif-usb]

I used the slider at the bottom to rewind the vid each time I thought I'd lost count. But I didn't see a gorilla.

[Julf]

I answered yes - but I think that was because the nature of the "something" was mentioned in one post, so I knew to look for the "something".

[Musicophile]

You'll be biased, as you'd be expecting something weird to happen, if you find such a video on this forum. I actually had read about this video beforehand, in a great book, "Thinking - fast & slow" by Daniel Kahneman, Nobel Price winner on decision theory.

[ringenesherre]

I answered yes, but it was only close to the end because I was distracted by some noise in the house and diverted attention from the white players.

 

Otherwise, it would have been a clear no :-o Nice video!

 

Cheers,

Peter

[Paul R]

Interesting replies. I wonder, were you guys sure that you would have answered no, even though you did catch it, counting just the passes from players in white T shirts? I was, I think counting all the passes and keeping a second count, mentally, of the passes by white shirts.

 

Paul

P.S. Karen answered no, and she is a lot smarter than I am. Really! Ten IQ points. She did get a really funny look on her face when it happened though, even though she swears she did not see it.

[jhwalker]

Got the number of passes right, and answered the question "yes" - don't see how you could miss it

[mav52]

I counted 16 passes darn it, but got the man dressed up like a gorilla.

[Earthenware541]

Got the number 'right' but was totally blown away to hear there was a gorilla there and couldn't believe it until I replayed it myself.

So that's it. I'm going to listen to my entire music collection again and I just know I'm now gonna hear something huge and obvious that I'd selectively removed until now. Darn it...

[wkhanna]

I lost count after the third pass........

....then got distracted watching the Singapore GP of Speed.....

....& then noticed a gorilla walk thru the scene from the corner of my eye.

[esldude]

I guess a surprise is that at least a couple people couldn't or I'll amend that to didn't stay focused on the video for what would be about 70 seconds not counting the questions at the end. So multi-tasking isn't news anymore, but makes me wonder all the same. Nor do I see how you could have a good chance of counting the passes if looking away at distractions.

 

Maybe we need a new forum.....computer_telephonainterneting_audio_video.com perhaps. The best in multi-tasked, multi-sense entertainment experience. So many potential interactions with that. All meant in good fun now, not being critical. I must admit to sometimes being distracted by F1 and MotoGp at times.

 

Come on now, would like to see at least twice as many votes. Looks to be running near the 44% in the original research from 1999. But fewer than 5% who viewed the thread have voted. Isn't nearly so bad as voting in the upcoming US political elections.

[jshawn]

my answer was "No". I wasn't even close to seeing the gorilla.

 

Then again, I got the number of passes wrong as well. Not my day.

[wgscott]

I saw it, but I was completely distracted from the assigned task.

[wgscott]

Me too. But supposedly in a large group where this was first done, 44% did answer differently. I thought it a good thing to post here and see what numbers we get.

 

Yeah, it is only 43% here. This shows objectivists can't be trusted.

[esldude]

I did of course have some ideas about how it relates to audio testing of course.

 

It supports the idea that in short focused abx testing one can miss things that would stand out in more casual viewing. I doubt many would have missed the gorilla if they had been told nothing more than watch this short video. It was the focusing of attention that used up some of your perceptual abilities and caused you to miss the obvious thing right in the middle.

 

On the other hand, if told to watch the video and nothing more, then asked how many passes occurred for the white shirts, I am pretty sure the answers would have been over a much wider range and far less accurate. That doesn't exactly go against the idea that focused blind testing might be more discriminating for some purposes. I think it perhaps offers a suggestion for improving it. That is picking something to listen to. I think most agree abx works fine and quite well for test signals. You might make it more similar with music by focusing on something. I think this also would lessen the stress and strain of blind testing.

 

When you are told listen and see if you can hear a difference, you strain to hear anything at all, with no particular thing in mind to hear different. I think this also increases the false positives. If on the other hand you were told, listen to the drums on this and see if you can hear a difference, or told listen to the trumpet and see if you hear a difference, it would be a different experience. You might focus intently, but it wouldn't be quite the strain of an amorphous task. I also think it would reduce false positives somewhat though that might be on shakier ground without testing to corroborate.

 

Like the gorilla in the video, such focused blind testing with music might let you miss something obvious in casual listening. It also could be interfered with by other parts of the music not being focused upon. In the video above, I almost missed the gorilla, it was when it got between a pass I was trying to count I noticed something and then saw it.

 

I suppose some will think I am only defending what I already thought. I think focused blind testing is ultimately more discriminating at the margins of perception. As the video demonstrates however, I think it can allow one to miss an otherwise obvious difference. I can imagine checking two components for bass guitar, voice, horns and cymbals. Picking in a blind test one product over another. Then put it altogether and listen at ease to the whole thing and finding I like the result less well than the component I didn't pick. The long term listening is better at determining comfortable preferences.

 

One does have to have genuine differences in two things for it to become apparent over time. If the two things test out the same blind testing them I would be suspicious of an alternate result long term. And yes I know blind testing can be done long term as well. Most of the work in the field suggests however that long term blind testing is less good than short term despite almost everyone feeling otherwise. A missed gorilla could be there, but one just needs to figure out the parameters to uncover it. If one had been told to watch the video and see if there is a gorilla I think the scores would be pretty much 100% correct. As usual the aligning of test results with actual use is an important variable to get right.