rtp1@tank.uchicago.edu (raymond thomas pierrehumbert) (01/06/90)
We would like to do some psycholinguistic studies involving reaction time determination (the sort of thing where you flash up a picture on the screen, and then time how long it takes for the subject to respond by pressing a button. In a further refinement, there will be audio stimuli (generated externally), and it would be necessary to trigger the appropriate flash with via a sychronization signal coming in over the serial line. All this requires timing with millisecond accurace. Hardware gurus in psychology say the way to do this is with a PC clone costing about $750, running MS-DOS. While my religious denomination does not actually preclude me from using MS-DOS machines, I have a real aversion to this, as we are basically a UNIX and Mac shop (and you can't easily do realtime stuff in Unix), have no use at all otherwise for MS-DOS, don't want to learn it, and would have no other user for the PC. On the other hand, we could always use another SE or Plus around the place. So, is there any way to do what I want using a Plus? The toolbox routines only seem to give you 1/60 second accuracy on timing, which is not good enough. I know with a MacII we could get a fancy board and do it, but the cost would be out of the question for this experiment. Please reply via email, and I will post a summary if it looks like there is sufficient interest. .