find some practice you don't like which seems popular, especially if people keep reporting its success, but doesn't make you money
devise some experiment where you could show it not working. If you can't do that, find some way to show that it is only as effective as some other practice which you don't like, or which no-one has discovered the mechanism for. Popular choices here include Homeopathy, or versions of the practice which sound incredible. Another easy one is sympathetic physiological response to stimuli - where your body has observable physiological effects in response to even a simple suggestion (aka "the placebo effect")
do not, under any circumstances, ask for any review of your experiment plan before beginning from practitioners of the practice you are trying to discredit. In particular, do not question whether or not your control is a true control or not - anything which changes the method that sounds like it should make it ineffective will do. For instance, when discrediting acupuncture, picking "control points" with no channels running through them is acceptable even though according to the Chinese literature no such points exist; using lasers which are not switched on is a certain control - you don't need to first prove that it was the activation of the beam that caused the effect.
for bonus points, attach the word "sham" to your control at various points, to make it sound like it's quite common and accepted that the practice is a sham.
