The criticism that the participants were just doing as they were told has always seemed to me as rather missing the point. What better demonstration is there of the human capacity for cruelty than a bunch of Stanford undergrads torturing other students because a professor told them to?
Depends on what you are trying to prove. If that "ordinary men following orders and under expectation can be total bastards" then it's just Milgram's experiment all over again which result is itself obvious for any historian. And if you are trying to prove that "ordinary men in position of power in jail setting spontaneously become total bastards" then the experiment is crap for that purpose if guards were encouraged…
But as I pointed out in the other thread ( https://news.ycombinator.com/item?id=17289154 and https://news.ycombinator.com/item?id=17289386 ), Milgram has his own detractors. His conclusions were far from obvious to many, and are still controversial today. And it's true that his work was not above criticism. It seems you can't touch this subject on HN without getting personal, though. Unjustifiably-flagged comments and scores clamped to the -4 minimum are the order of the day.
I'm not a psychologist and have only a casual interest in the subject, so my experience in the last thread was like stumbling into a bar fight that I didn't know was going on. I suspect this thread will evolve along similar lines, but it's hard to be sure without conducting a formal study that will probably come under heavy dispute. /s
And if you are trying to prove that "ordinary men in position of power in jail setting spontaneously become total bastards" then the experiment is crap for that purpose if guards were encouraged to behave that way.
As everdev points out, that's the whole idea behind prison guard training! If you insist on studying these scenarios in a sterile, painstakingly-ethical academic environment, you will always be left scratching your head when an Abu Ghraib or a My Lai or a Guantanamo happens, wondering why reality stubbornly refuses to conform to your approved methodology.