Page 1 of 1

4 4AFC interleaved staircases

PostPosted: Mon Oct 12, 2015 12:55 pm
by panorgiasa
Hi all, I am looking into buying Psykinematix Sharp (the edition for use with the Bits# card by CRS). I downloaded the trial version so far and I am impressed by the structure of the software. Not that it's easy, but it's well thought. I want to buy this software for a very specific experiment and want to make sure that I can do it before investing on the Sharp edition. I was unable so far to figure out how to program this experiment.

Long story short, I want to measure sensitivity in 4 different eccentricities (i.e. Top, Bottom, Left and Right). This means that I need 4 different stimuli each one presented on a different location (i.e. T, B, L, R). I want to use a staircase for each stimulus (therefore 4 different staircases) and these to be interleaved. The task for the observer will be to identify the location of the stimulus and therefore the task will be a 4AFC (T, B, L and R), however, each stimulus will only be presented in one location and the staircase should only accept this location as a correct response and all the others as wrong.

For example, the Top stimulus is presented, the subject has four alternatives (T, B, L, R) responds T (correct response) then randomly the Left stimulus is presented, the subject has 4 alternatives, responds R (wrong response), then randomly the Bottom stimulus is presented, the subject responds B (correct response) etc etc.

Any ideas whether this is doable?

Best
Thanasis

Re: 4 4AFC interleaved staircases

PostPosted: Mon Oct 12, 2015 10:50 pm
by psykinematix
Hello,

Your task can be easily created in just a couple of minutes. The experiment structure should look like this:

Image

In the top event (Experiment), you could first add an independent variable 'ecc' to specify the eccentricity of the target presentation:

Image

In the condition event, you would define the different conditions to be randomly interleaved. In your case these are the possible locations of the target specified for example by a variable 'angle' with a value defining its angular position: 0 deg for right location, 90 deg for top, 180 deg for left and 270 0 for bottom (this variable will be also used to indicate the position of the stimulus later on).

Image

In the staircase event, specify the dependent variable, for example the target contrast (this variable will be also used to indicate the stimulus contrast later on):

Image

Then in the procedure event (Location detection), indicate the number of possible responses (4) and for each of them the response key and the expression associated with the correct response (e.g. the correct response for a top-arrow key press is when the value of the 'angle' variable is 90 as defined in the condition event above):

Image

Finally link the the dependent and independent variables to the properties of the target stimulus, that is its x and y coordinates ( [ecc]*cos(pi*[angle]/180) and [ecc]*sin(pi*[angle]/180)) and contrast ( [contrast] ):

Image

Note: don't forget to put the brackets [ ] around the variables when using them in expressions !

Hope this helps!

-- The Psykinematix Support & Sales Team at KyberVision Japan LLC / Consulting, R&D in Vision Sciences

Re: 4 4AFC interleaved staircases

PostPosted: Tue Jun 13, 2017 2:24 am
by Dolin
panorgeiasa wrote:Hi all, I am looking into buying Psykinematix Sharp (the edition for use with the Bits# card by CRS). I downloaded the trial version so far and I am impressed by the structure of the software. Not that it's easy, but it's well thought. I want to buy this software for a very specific experiment and want to make sure that I can do it before investing on the Sharp edition. I was unable so far to figure out how to program this experiment.


That was really helpful. Thanks for the tutorial.