JND frequency discrimination and Direct RT

[pconner]

I am trying to set up a two-tone frequency discrimination task to find the "just noticeable difference" for a set of participants. In this task a given stimulus would be dependent upon the frequency of the previous stimulus together with the participant's response - whether it was correct or incorrect as well as how many responses were correct within 5 trials. Could anyone give me ideas about how to set up these contingencies? Thanks!

[newell]

Hm. I would conduct this experiment by creating a series of Skip events for each time a user gives their answer - whether it be incorrect or correct.

Let's say you first allowed the participant to hear the sound file sound_1. Let's say they then gave the proper response to the stimulus sound_1. We would then use the skip command to move to a trial to continue the experiment under the condition that the user had properly responded to sound_1.

Since DirectRT doesn't yet have the same variable manipulation power that MediaLab has, you may need to create a more complex matrix of skips to track exactly how many of the 5 trials the user answered correctly.

For more information on the skip feature, see the following link: http://www.empirisoft.com/directrt/h...t_skipping.htm
To see information on a similar feature (jumping), see this link: http://www.empirisoft.com/directrt/h...rt_jumping.htm

Also make sure you check out the collection of samples that comes installed with the software. They may help give you more direction with the experiment you're trying to create.