In order to do the experiment you need to have a copy of a game called "Gabori Attack" on your hard drive. Check to see if

you have the folder called Gabori Attack and that it contains several items, including the application itself. If you don't have it

you can download either the Macintosh or the PC version. But read this page before you do the experiment.

When you double-click on Gabori to start the program, this is the first screen that you see. Here you can see three sine wave gratings

with different spatial frequencies. The game is called Gabori Attack because a sine wave grating which has high contrast in the centre but low contrast at its edges - like these - is known as a Gabor patch. Note that the instruction below the gratings tells you what to do.

When the game starts, the screen will be an even grey. Somewhere on the screen - a random position - there will be a Gabor patch which is below threshold but is gradually increasing in contrast. As soon as you see it, you must click on it with the mouse - to kill it. The question being asked in this experiment is this:

Students get very bored doing proper psychophysical experiments, so can we use an arcade game like this to measure the contrast sensitivity function? Will it work at all; and if it does, how well will it work?

If it does work, then you should require higher contrast to see the very low and very high spatial frequencies compared with middle ones.

When you click on any Gabor (on the above screen) this next screen comes up:

Here you enter your name and the number of repetitions (Enter 10). Then, when you are ready, click "Play Gabori" and start splatting those sneaking up Gaboris.

BUT DO NOTE VERY WELL: For this game to work properly you must have your monitor set to "Millions of colours" if using a Macintosh; or True Colour (24 bit) if using a PC. To do this on a Macintosh, go to the Monitors control panel, click Monitor and choose "Millions" of colours. On a PC, choos "Start - Settings - Control Panel" , choose "Display" then "Settings" and set "Colour Palette" to "True Colour".

When you have finished playing the game, most of you will have very similar data. The tutor will give you a set of data from 10 other subjects and these will be analysed for the report. Your own data and will probably look something like this:

And below the graph you'll be given the actual data like this:

NOTE: There must be an error in the program somewhere. Each spatial frequency was done 10 times for this example but the Ns at the right say not. Note also: The standard error is a standard deviation but not of individual scores; rather, if you did the same experiment an infinite number of times and constructed the distribution of the means of those samples, the standard error would be the standard deviation of those means.

OK. When you've done all this we can go to the issues you need to consider in your report.