Turns out, it's very hard to remove a test subject's own bias about their psychedelic test treatment.