12-02-2020, 12:40 PM
Yes, that bug-fix solved the problem!
Re: timing, a note: We started with 800ms stimulus duration and 200ms ISI. The 120 trial sequence now takes about 122-123 sec or ~1% error, with an average of stimulus duration of 810 ms (range 806-814ms). However, by scripting my stimuli for 790ms, I got an average duration of exactly 800ms (range 795-808ms), which is an ideal margin of error (better than I'm accustomed to seeing on locally-installed software). The presentation seems to be restricted towards changing the display _after_ the assigned stimulus duration, not at lowest-error opportunity (before or after), so setting the duration to be slightly short let's the screen update over a distribution of slightly-early to slightly-late and gives better average performance.
We'll scale this script up to our actual study design and re-test, but I have every expectation that it will work at this point.
Thanks so much for the help!
Re: timing, a note: We started with 800ms stimulus duration and 200ms ISI. The 120 trial sequence now takes about 122-123 sec or ~1% error, with an average of stimulus duration of 810 ms (range 806-814ms). However, by scripting my stimuli for 790ms, I got an average duration of exactly 800ms (range 795-808ms), which is an ideal margin of error (better than I'm accustomed to seeing on locally-installed software). The presentation seems to be restricted towards changing the display _after_ the assigned stimulus duration, not at lowest-error opportunity (before or after), so setting the duration to be slightly short let's the screen update over a distribution of slightly-early to slightly-late and gives better average performance.
We'll scale this script up to our actual study design and re-test, but I have every expectation that it will work at this point.
Thanks so much for the help!
