Future Tests of Medical Dowsing

I still need to have more tests of the medical dowsing experience. The test with the fungal infected seeds in petri dishes showed results that were perfectly consistent with what random chance would predict, ie. no ability there. Why then, since statistics obviously works, does my medical dowsing always produce results that are better than that precisely dictated by statistics?

I want to have the following types of tests:
A. Lactobacillus: subjects are given either a Lactobacillus supplement or a sugar pill. Subjects are unaware of which one they were given. I am to try to dowse whether they have taken a Lactobacillus supplement or not.
B. General medical dowsing: general readings where I list any medical information I detect, which is then checked for accuracy.
C. Dowsing for specific information: such as matching a diagnose to a person.
D. Migraine healing: I stll need to try migraine healing on at least one more person to see whether it works or not.

Note that all of my tests occur within safe, ethical, and legal constraints. I would not have it any otherwise. And that I make no money off my claim or testing, rather this is a sometimes expensive hobby of mine. Subjects are preferably skeptics, or persons who do not find out what I might have seen in them and come to no harm. I do not practice psychic readings, I only investigate the experience of medical dowsing. So there is no reason to try to stop my work. :P

/