Consider starting small, with short, clear and engaging examples, like the Newcomb's problem, the PD or the Trolley problem, or the Milgram or Stanford experiments
It's awfully easy to read the Milgram or Stanford experiments as (e.g.) ammunition for anti-authoritarianism without deeply understanding what makes them tick. This seems to be a general problem with dramatic psychological results.
I'm afraid I haven't properly designed the Muggles Studies course I introduced at my local Harry Potter fan club. Last Sunday we finally had our second class (after wasted months of insistence and delays), and I introduced some very basic descriptions of common biases, while of course emphasizing the need to detect them in ourselves before trying to detect them in other people. At some point, which I didn't completely notice, the discussion changed from an explanation of the attribution bias into a series of multicultural examples in favor of moral relativity. I honestly don't know how that happened, but as more and more attendants voiced their comments, I started to fear someone would irreversibly damage the lessons I was trying to teach. They basically stopped short of calling the scientific method a cultural construct, at which point I'm sure I would have snapped. I don't know what to make of this. Some part of me tries to encourage me and make me put more effort into showing these people the need for more reductionism in their worldview, but another part of me just wants to give them up as hopeless postmodernists. What should I do?