Were you able to find a work around? e.g., putting both outcomes in the same stim file but in varying proportions (depending on condition) and the drawing from that list randomly? Or something along those lines?