Apologies to Empirisoft staff for posting here as well as over e-mail, but it is urgent that I resolve this problem soon!

Here's my question:

I am creating a reaction time experiment in DirectRT that requires correct responses for 4 consecuctive stimuli per trial. The stimuli appear all at once, so I've programmed the trials (rows) in a kind of funky (and perhaps crude) way--for the first response, the 4 stimuli appear from left to right (loc is x,x,0, time is 0 for the first 3 stims in the row), with an RT after the fourth stimulus; this is repeated 4 times in order to get 4 responses while the same stimuli are on the screen in the same trial. (Hopefully I have explained this somewhat clearly--if not, I can forward you the .csv file to show you what I mean).

Here's the problem: I have implemented the "correct response required" feature, and it works fine UNLESS the very last response made in a trial is incorrect; in those cases, the program simply quits (when run through directrt) or crashes to a blank screen (when run through medialab) when the correct response is made after the incorrect one. I have no idea why this is happening!

An example of what's going on:

Say the task is "press the z key if the stimulus is < 3, press / if it is > 3" (this is similar to the actual task)

Here is the sequence presented:

1 2 4 5

If the responses go like this, for example, nothing goes wrong and the next trial follows:

"z", "/" (error message), "z" (correction), "/", "/" *next trial*

But if they go like this, the program will crash/quit:

"z"," z", "/", "z" (error message), "/" (correction), *crash/quit*

Please let me know how I might elaborate if this is unclear, and many thanks in advance for any help you might be able to offer,

Brian Atwood
Department of Psychology
Harvard University