Be a Supporter!

Change blindness demo

rated 2.39 / 5 stars
Share Links:
Skill - Other

Credits & Info

Mar 13, 2010 | 5:24 AM EST

If you liked this, check these out!

Author Comments

"Change-blindness" occurs when a change is masked by slow on-off flicker. Have a go at seeing how fast you can detect the changing element in the flashing array. Your results will be graphed at the end, based on the size of the feature change (shown in picture). The greater the temporal frequency (speed) change between the frames the faster you should detect it.

This demo involves detecting a change in the speed of a rippling noise pattern. An array of five randomly positioned patches slowly flash on and off. On every second presentation, one of the patches changes speed (or temporal frequency;TF). As soon as you locate the change click on it with your mouse. There are 15 trials, each includes one patch changing temporal frequency. Your data will be displayed in a graph to you at the end of the session. Oct. stands for octaves, 1 being slowest 6 being fastest. When you save your data there might be a pop-up (just ignore it) your results will be displayed here.



Rated 3 / 5 stars


Was too easy and made no sense


Rated 0 / 5 stars

I totally don't understand this.

The idea seems to be interesting, but in the current form, I can't allow this to hit the portal, as I have no idea what this is about, even though I really want to. I just clicked random objects till the end.

Hey, know what? Put some in-flash explanation that's easy to understand, and re-submit. I'm sure that once we (or I, at least) get to know what this is about (well, I'm guessing some eye test...) we'll like it :)

People find this review helpful!


Rated 4 / 5 stars

some presentation flaws

although premise is scientifically valid and concept is sound...

instructions are unintuitive at first, with several initial trials involving random clicking, and with several initial "trials" invalid due to lack of actual understanding, the end data is relatively flawed.

similarly (in pespective of an average user), end-result data presentation can be difficult to relate to if unfamiliar with the academic mannerism. Perhaps tooltips that "simplify" the jargon a notch could help?