wcrescioni has a great suggestion by trying to use XP emulation as a possible solution to your problem. However, I would also suggest that you compare any results regarding reaction time from the experiment using Medialab to lauch DirectRT in the Windows XP emulation to the same experiment run in an actual version of Windows XP. Because DirectRT relies on getting direct access to the OS in recording its reaction times, doing this check will allow you to see if the XP emulation creates any differences in the recorded reaction times. This would hopefully save you some heartache down the road by ensuring the XP emulation gives you reliable reaction times before you start gathering your actual data.