setting pointer location

[mark mitchell]

I don't think moving the cursor/pointer works very well, even if you are 
measuring RTs.  The reason is that users will run off the mousepad (or 
the desk) that way, creating bad data in the process.  You have to get 
your user to move the actual 3D mouse back to a central location.  The 
way I have done this for psycholinguistic experiments is to make the 
"next image" (next stimulus) button a small red dot in the central 
(neutral) position on the screen.  You can label this dot with an arrow 
and "press here for next question" or something similar.  This way you 
will get users returning the mouse to the central location both 
physically and virtually -- and avoid confusing or surprising them.  I 
think your data will be far more robust this way.

best of luck,

mark mitchell
Japan