“the guards were not clearly informed that they were subjects”

I’m not a psychologist, but from what I know after studying non-consensual experimentation, this omission alone would get the proposal denied by a modern IRB.

Like the Milgram Experiment, the SPE purports to study the nature of institutional cruelty, but both experiments inflict institutional cruelty on unsuspecting subjects. I understand why Zimbardo wouldn’t want to inform his subjects of the purpose of his study: If the researcher had said something like “We want to study how ordinary people can become abusive monsters,” then the guards (or the experimenter-subjects in Milgram’s case) would be motivated by social desirability not to show abusive behavior (even if they might have displayed such behavior had they been unwitting subjects). And if you had told the “prisoners” that they could leave at any time, then any sane subject would probably have left before the conditions could reach their most abusive state.

If you cannot test a hypothesis by using a humane, scientifically valid experiment, then you cannot test the hypothesis.

Anyone who wants to show that prisons create opportunities for abusive behavior by guards can simply use Frederic Wiseman’s cinema verite technique and film actual conditions in a prison, as Wiseman did at Bridgewater.

Neither the hypothetical benefit of nor the data recorded by an inhumane and/or unscientific experiment that pretends to be scientific is adequate justification either for cruelty or for undermining the tenets of the scientific method.

The SPE should not die. It should be recorded in a chronicle of Frankensteinian hubris and taught to researchers everywhere.