I'm curious what techniques you might employ to give this a shot. There have been at least a few double-blind trials that have yielded results that are no better than random, according to the wiki entry on the topic, which I had to read to even understand what dowsing was.