Few psychology studies have entered popular culture as deeply as the Stanford Prison Experiment. Conducted in August 1971 by psychologist Philip Zimbardo and colleagues Craig Haney and Curtis Banks, the study transformed the basement of Stanford University’s psychology building into a simulated prison. Healthy young men were randomly assigned to act as either prisoners or guards, and a study originally planned to last as long as two weeks was terminated after only six days. Accounts of humiliation, rebellion, emotional distress, and increasingly harsh guard behavior quickly turned the experiment into one of the most memorable illustrations of a basic social-psychological idea: powerful situations can sometimes alter behavior more dramatically than personality alone would predict. The researchers’ original article, “Interpersonal Dynamics in a Simulated Prison,” argued that the mock institution became a psychologically compelling environment that elicited unusually intense reactions from participants.
For decades, textbooks and documentaries presented the experiment as evidence that ordinary people can rapidly become abusive when given institutional power and assigned authoritarian roles. Zimbardo later developed this interpretation in The Lucifer Effect: Understanding How Good People Turn Evil, connecting the experiment to broader questions about prisons, military abuse, conformity, and institutional corruption. But that familiar story is no longer accepted uncritically. Archival investigations, participant interviews, methodological analyses, studies of demand characteristics, and the later BBC Prison Study have challenged the idea that cruelty emerged automatically from the roles themselves. The Stanford experiment remains historically important, but its modern significance lies as much in the controversy surrounding its methods and interpretation as in its original findings.
How the Stanford Prison Experiment Was Designed
Zimbardo’s team recruited participants through newspaper advertisements offering $15 per day to male college students willing to take part in a psychological study of prison life. Approximately 75 men responded. Researchers used interviews and psychological assessments to exclude applicants they considered physically or psychologically unhealthy, unusually antisocial, or otherwise inappropriate for the project. Twenty-four were selected and randomly assigned to the categories of prisoner or guard, although standby participants meant that the final published analysis was based on 10 prisoners and 11 guards. The researchers deliberately selected what they regarded as a psychologically normal and relatively homogeneous group so that dramatic differences in later behavior could be attributed more plausibly to the simulated prison environment rather than obvious preexisting pathology.
The basement of Stanford’s Jordan Hall was converted into a small prison with barred cells, an observation area, and a cramped closet used for solitary confinement. Prisoners remained inside the simulation around the clock, whereas guards worked eight-hour shifts and returned to their normal lives afterward. The prisoners were unexpectedly arrested with the cooperation of local police, processed, searched, assigned identification numbers, dressed in prison clothing, and placed in cells. Guards wore uniforms, carried batons, and used mirrored sunglasses that reduced normal eye contact. Physical violence was formally prohibited. The researchers described the project as an attempt to study how roles, institutional rules, social expectations, and unequal power might shape behavior inside a prison-like setting.
Six Days in the Basement
The first day was comparatively uneventful, but the simulation soon became more confrontational. On the second day, prisoners rebelled by barricading themselves in their cells and resisting guard authority. Guards responded by breaking the rebellion, stripping prisoners, removing beds, imposing restrictions, and developing systems of rewards and punishments. Counts that initially served the practical purpose of checking prisoner numbers evolved into lengthy exercises in authority and obedience. Some guards increasingly used humiliation, arbitrary rules, sleep disruption, denial of bathroom privileges, and forced exercises to demonstrate control. The original researchers interpreted these developments as evidence that the structure of the institution was beginning to shape behavior independently of the participants’ previous personalities.
Several prisoners displayed substantial distress, and some left the study early. The most famous case involved Douglas Korpi, identified as Prisoner 8612, whose emotional outburst was long portrayed as evidence that the prison had become psychologically real. Later accounts complicated that interpretation: Korpi subsequently said that at least part of his dramatic behavior was an attempt to obtain release from an experiment he felt he could not easily leave. The study’s procedures regarding withdrawal have themselves become a source of criticism. Documents cited by later researchers indicated that prisoners were to be discouraged from quitting, while the consent language framed release as something determined partly by medical judgment. Whatever the precise interpretation of individual breakdowns, participants clearly experienced a setting in which the boundary between voluntary research and enforced imprisonment became unusually blurred.
Zimbardo’s “Power of the Situation” Explanation
The interpretation that made the Stanford Prison Experiment famous was situationism: the idea that environments, institutions, roles, and systems of power can exert enormous influence over behavior. Zimbardo argued that focusing only on individual character can cause observers to underestimate the effect of social structures. In later discussions of abuse, including his writings about Abu Ghraib, he emphasized processes such as deindividuation, anonymity, diffusion of responsibility, conformity, group norms, and dehumanization. His broader argument was not that personality never matters, but that ordinary people placed within certain systems may behave in ways their previous lives would not predict. As he later summarized the position, “All you really need is a situation” that makes crossing moral boundaries easier.
The experiment’s enduring appeal partly comes from the disturbing simplicity of this explanation. If guards were psychologically normal before the experiment and became abusive only after receiving power, uniforms, anonymity, and institutional authority, then cruelty cannot comfortably be dismissed as something committed exclusively by unusually cruel individuals. Yet even the original data never showed that every guard became sadistic. Guard behavior varied considerably, with some acting harshly, others behaving more moderately, and some being regarded by prisoners as comparatively decent. Zimbardo himself has emphasized these individual differences in later defenses of the study. The actual results therefore never justified the strongest popular version of the claim—that assigning an ordinary person the label of “guard” automatically transforms that person into an abuser.
The Ethical Problems That Changed Psychology’s View of the Study
The Stanford experiment became a classic example of the ethical dangers involved when researchers become too immersed in the worlds they create. Zimbardo did not operate solely as an investigator; he also assumed the role of prison superintendent. He later acknowledged that this dual role compromised his capacity to remain a detached observer. Instead of independently evaluating whether the simulation had become harmful, he became involved in schedules, discipline, parole hearings, guard management, escape concerns, and everyday prison administration. In a later response to critics, he called the lack of an independent scientific observer the experiment’s “major flaw.”
The turning point came when psychologist Christina Maslach, who had recently completed her doctorate at Stanford, visited the simulated prison and saw prisoners being taken to the bathroom with bags over their heads and their legs chained. Unlike dozens of others who had encountered the study, she openly challenged what was happening. She told Zimbardo, “It’s terrible what you are doing to these boys!” Zimbardo later said the confrontation forced him to recognize how completely he had adopted the perspective of the prison administrator. The experiment was stopped on its sixth day. He subsequently apologized for failing to provide adequate oversight and for allowing participants to suffer. The episode remains a powerful illustration of why research ethics require independent monitoring, meaningful withdrawal rights, careful risk assessment, and investigators capable of suspending a study when participant welfare is threatened.
Demand Characteristics and the Problem of “Acting Like Guards”
Methodological criticism began surprisingly early. In 1975, psychologists Ali Banuazizi and Siamak Movahedi argued that participants may have entered the study already understanding what the researchers expected. This is known as demand characteristics: cues that reveal a study’s hypothesis and encourage participants to behave in ways they believe support it. When Banuazizi and Movahedi gave other students information resembling what Stanford participants would have known, more than 80 percent reportedly inferred the experiment’s hypothesis, while nearly 90 percent expected guards to behave in hostile or oppressive ways. Even Haney, Banks, and Zimbardo’s original report acknowledged that experimental demand characteristics had probably exerted some influence.
Later evidence strengthened these concerns. The guard orientation included language about creating frustration, fear, arbitrariness, and powerlessness among prisoners. Zimbardo insisted that guards were not explicitly trained to abuse prisoners, and physical violence was prohibited, but instructions may still have communicated what kind of prison environment researchers wanted. Psychologist Jared Bartels tested this possibility experimentally in three studies published in 2019. Participants exposed to wording modeled on the Stanford guard orientation expected more hostile and oppressive behavior from investigators, fellow guards, and themselves than did participants receiving more neutral instructions. The findings do not prove that Stanford’s guards merely obeyed a script, but they demonstrate that the orientation itself was capable of shifting expectations toward harsher behavior.
Archival Evidence and the Modern Reassessment
Historian and researcher Thibault Le Texier pushed the critique considerably further by examining archival materials and interviewing former participants. His 2019 paper in American Psychologist, “Debunking the Stanford Prison Experiment,” argued that the conventional account omitted important information about researcher involvement, guard instructions, data collection, and participant knowledge. According to Le Texier, guards were not simply left inside a neutral prison environment to discover their roles spontaneously; researchers shaped the institution, communicated goals, supervised behavior, and established many rules. His later book, Investigating the Stanford Prison Experiment: History of a Lie, expanded that argument using archival and interview evidence.
These findings matter because the experiment has often been treated as if it were a tightly controlled demonstration of causal psychology. In reality, it resembled an intensive simulation or case study more than a conventional controlled experiment. It lacked an ordinary control condition, involved a small and highly specific sample, depended heavily on researchers who also administered the institution, and contained numerous opportunities for participants to infer expectations. Psychologist Jared Bartels’s analysis of introductory psychology textbooks found that many texts historically presented the study’s dramatic conclusion while giving little attention to these methodological objections. The Stanford experiment therefore became an example not only of social influence but also of how a compelling scientific narrative can become stronger and simpler as it passes into textbooks and popular culture.
Did the Experiment Attract People Who Were Already Different?
Another challenge concerns self-selection. Zimbardo’s team selected psychologically healthy participants, but the participants first had to volunteer for something advertised specifically as a study of “prison life.” In 2007, psychologists Thomas Carnahan and Sam McFarland investigated whether that wording itself might attract people with particular personality characteristics. They compared volunteers responding to an advertisement for a generic psychological study with volunteers answering a nearly identical advertisement describing a psychological study of prison life.
The prison-study volunteers scored higher on measures including aggressiveness, authoritarianism, Machiavellianism, narcissism, and social dominance while scoring lower on empathy and altruism. This does not establish that the Stanford guards possessed those characteristics or that personality explains what happened in 1971. The researchers explicitly described that conclusion as conjectural. But their results undermine a simple opposition between “bad individuals” and “bad situations.” People select environments, organizations recruit particular kinds of people, leaders establish norms, and situations then interact with individual differences. A more defensible interpretation of institutional abuse is therefore person-situation interactionism rather than the idea that personality is irrelevant.
The BBC Prison Study and a Different Theory of Tyranny
Psychologists Stephen Reicher and S. Alexander Haslam offered perhaps the most important theoretical challenge with the BBC Prison Study, conducted decades later. Men were again divided into groups with unequal power inside a specially created institutional environment, but the psychological pattern differed dramatically from Stanford’s. Guards did not automatically unite around their authority. Instead, they failed to develop a strong shared identity and became reluctant to exercise power consistently. Prisoners, meanwhile, developed greater group identification, organized collectively, and eventually overcame the guards. Participants then attempted to create a more egalitarian system.
Reicher and Haslam argued that tyranny does not result simply from passive conformity to assigned roles. People must identify with groups, leaders, values, and ideological projects before acting collectively on their behalf. Under some circumstances people obey authority; under others they resist it. Their work therefore shifts the question from “Why do people automatically become their roles?” to “Under what conditions do people embrace, reject, or redefine the identities offered to them?” This interpretation also helps explain something visible in Stanford itself: not all guards acted alike, prisoners rebelled, some participants resisted expectations, and an outsider—Maslach—challenged the entire institutional frame. Social situations matter enormously, but they do not mechanically erase human agency.
What the Stanford Prison Experiment Actually Teaches Today
The Stanford Prison Experiment should no longer be described simply as proof that ordinary people inevitably become cruel when given authority. The strongest version of that conclusion exceeds what the study can establish. Too many factors were entangled: expectations communicated by researchers, existing cultural stereotypes about prisons, participant self-selection, uneven guard behavior, researcher leadership, group processes, institutional rules, and the unusual difficulty prisoners experienced when attempting to withdraw. Modern research indicates that situations can powerfully affect behavior, but those effects are filtered through personality, identity, leadership, norms, perceived expectations, and choices.
Yet rejecting the mythology surrounding Stanford does not make the experiment meaningless. Its history raises enduring questions about what happens when institutions give some people control over others, normalize degrading treatment, weaken accountability, and make resistance difficult. It also teaches a second lesson that may now be equally important: scientific stories themselves require scrutiny. Famous experiments can acquire simplified interpretations that survive long after researchers have identified serious methodological limitations. The lasting importance of the Stanford Prison Experiment is therefore more complicated than Zimbardo’s original “power of the situation” narrative. It is a case study in power, leadership, conformity, resistance, experimental ethics, demand characteristics, personality, social identity—and in psychology’s continuing responsibility to reconsider its most famous conclusions when new evidence demands it.
