Thanks for your help Jason! Your file also got me close to the task I want to use (stim is presented for 1500ms, goes away and still allows a response), but there is still one issue. When I run it, DirectRT still waits for a response for each trial (instead of just moving to the next trial after 2000ms if no response is given).

I want the task to run freely on its own, so that it's standardized across participants (whether or not they respond, they should not get a break during real trial blocks, since that could affect their performance later :] ). I tried tweaking by using "!!" in MinMax with and without -1 as a response option -- and still couldn't make that happen. This is where I'm stumped. Any other ideas?