Hello, I'm working on an experiment where particpants are required to give vocal responses to auditory stimuli, and we're measuring RT's for the vocal response. I also want the experimenter to be able to press a button to code the vocal response. I have it set up so that there are three stimuli on each trial: sound file, text(just to have something on the screen to indicate that it moved from the sound stimulus to the next stimulus), text (to allow the experimenter key press). The audio RT is taken to the second stimulus. (We are actually interested in rt's to the sound file, but we can't use headphones and if we take an rt to the sound file, the sound file itself activates the rt.)

So this is all well and good, and I have it so the microphone records for 300 ms after voice rt begins: rt:aud,rec,-300 (just enough time for the participant to say "stop" or "go"). However, when I check the recorded sound files, a few of them are way longer than 300 ms, maybe a couple of seconds. They start immediately with the voice response, but then just keep going past 300 ms. There are also some that seem to start recording before the voice response. That may just be an issue of microphone sensitivity, but I don't think so. On both of those types of weird recordings, the reaction times in the log file seem to be longer than they should be.

Any insight as to why some of the recordings of the voice rt's are longer than the requested 300 ms, why some recordings start before the voice response, and if this means that reaction times aren't being recorded accurately?

Thanks.