In order to build confidence in the results of a personal experiment like this one, it's best to (double-)blind it as much as possible and log lots of concrete data every day (e.g. keep track of your bowel movements, sleep quality, alertness levels, blood pressure, etc). This should help limit the impact of any (conscious or subconscious) preconceptions or other psychosomatic effects.

Also, change as little else in your life as possible while you're performing the experiment to control for confounding variables.