The rise and fall of social psychology

The field's reliance on cute studies, small samples, and political activism led to its collapse.

Written by Russell T. Warne.

After World War II ended and the full scope of Nazi atrocities became clear, scholars puzzled over their ghastly crimes. That the German people could wage a horrible war and commit genocide was disturbing to many in the West since Germany was not a backward or benighted country; rather, it was one of the most educated countries in the world, and—like other Western nations—it was a Christian nation. Those characteristics did not prevent the country from perpetrating some of the 20th century’s worst crimes against humanity.

Understanding the German situation was particularly important for Americans. In the United States, many citizens were descended from German immigrants. In 1940 U.S. Census, there were 5.2 million people born in Germany, and almost 4 million more had one or both parents born there—more than any other foreign country. This was a longstanding pattern; from 1880 to 1960, more foreign-born residents and citizens of the United States came from Germany than anywhere else. Americans were forced to grappled with the possibility that Germany was not unique, a disturbing scenario explored in such artistic works as the 1935 dystopian novel It Can’t Happen Here by Sinclair Lewis.

The issue never fully went away, and when Adolf Eichmann was tried in Israel in 1961 for his role in the Nazi regime, the world saw a defendant who looked more like an accountant than a cruel monster. In writing about Eichmann, philosopher Hannah Arendt famously coined the phrase “banality of evil” to describe how a mundane everyman could willingly perpetrate a monstrous genocide against the Jews. The fearsome implication was that beneath the surface of many—perhaps most—people lurked an Eichmann who could appear under the right circumstances.

In this zeitgeist (appropriately, a German word), social psychology flourished because it seemed to provide insight to these issues—often in dramatic fashion. Although social psychology existed in the early 20th century, the catalyst for the field’s influence came from the European experience of World War II.

Beginning in 1951, Solomon Asch announced the results of a startling series of experiments. When asked to identify which of three lines was the same length as a target line, male college students responding alone identified the correct line almost every time. However, when placed in a group of people who identified the incorrect line, the students conformed to the group over one-third of the time. Three-quarters of subjects in the study conformed at least once with the majority. Asch saw this as evidence of the susceptibility of many people to propaganda and a willingness to go along with a group—even when a person knew that the group was wrong.

In 1955, Turkish-American psychologist Muzafer Sharif published an elaborate field experiment where 12-year-old boys at a summer camp were randomly assigned to two groups that were placed in direct confrontation with one another. The boys developed an intense rivalry and prejudice against their outgroup. However, when the boys were forced to work together to solve problems with the camp’s food and water supplies and entertainment, the boys overcame their enmity and became friends.

Most famously, Stanley Milgram published an article in 1963 on the willingness of his American research participants to give a series of up to 30 electric shocks of increasing intensity to a victim when pressured to do so by an authority figure (a man in white lab coat). Depending on the conditions, up to 72% of Milgram’s participants administered the highest voltage, despite the pleas of the man to stop. This was a far higher percentage than Yale psychiatrists had predicted; they believed that just 2% of people would administer the 20th shock and that only 0.1% would administer the maximum voltage.

Social psychology’s successes were not limited to explaining the horrifying events of World War II. In 1971, Philip Zimbardo created a mock prison in the basement of a building at Stanford University. The “prison” was inhabited by male Stanford students who were randomly assigned to be guards or prisoners. The study was supposed to last two weeks, but Zimbardo called it off after six days because the results were so disturbing. The guards became cruel and sadistic, while the prisoners initiated a rebellion, and some developed psychological problems.

Zimbardo’s study, named the Stanford Prison Experiment, seemed to show that normal young people could turn into sadists when assigned a role where they were expected to behave that way. In time, Zimbardo used the study to shed light on the Attica Prison uprising, the violations of human rights at Abu Ghraib, and other headline-grabbing events. For the public, the Stanford Prison Experiment was a seductive twofer: It explained how atrocities could happen while also absolving the perpetrators. It wasn’t their fault; it was the role they were assigned. No wonder the first films inspired by the Stanford Prison Experiment were produced in Italy and Germany.

In less than 20 years, social psychology had demonstrated through these—and hundreds of other, less dramatic studies—the power of the situation on people’s behavior. Social psychology also seemed to explain how the Holocaust and other grisly war crimes could happen. Indeed, social psychologists explicitly connected their laboratory studies with the atrocities of World War II. Milgram, for example, explicitly linked his study to the Holocaust and stated that his results showed that large segments of the American population had the potential to cause great harm and suffering to others. His obedience study seemed to vindicate Hannah Arendt.

These startling findings stand in contrast to the failed efforts of other social scientists to identify the causes of the Holocaust. For example, psychologist Molly Harrower examined the records of the Rorschach test for eight Nazi war criminals and compared them to a control group. She found that there was no psychological characteristic that distinguished the war criminals from the controls. In 1946, sociologists and political scientists Paul Kecskemeti and Nathan Leites published a series of articles that proposed a psychoanalytic theory that many German people supporting the Nazi regime possessed a “compulsive character” and that this tendency set the stage for genocide. This was an unsatisfying explanation because it appealed to nebulous aspects of German culture and unproven concepts in psychoanalysis. Moreover, this theory could not explain how a “compulsive character” in a culture could be recognized and contained before it led to genocide.

In contrast, social psychology’s eye-catching studies not only explained some of the darkest events of World War II, but also gave people a simple strategy to prevent atrocities: just change the environment. Milgram found that compliance with instructions to shock other were reduced when the authority figure was distant (e.g., in another room), the perpetrator had to physically put the victim’s hand on an electrified plate, and if the perpetrator knew the victim. Thus, Milgram concluded, small tweaks in the situation could make people less likely to hurt others. Asch also found different results as he varied his experiments: conformity decreased when groups were smaller and when one other person disagreed with the majority. Sharif proclaimed that his study had real-world implications and that leaders could solve intergroup conflict by focusing on cooperation and solving mutual goals.

By the end of the 20th century, social psychology was riding high, and social psychologists had a long history of using similar techniques to explore—and often explain—other social problems, including prejudice, racism, and violence. For many of these issues, experiments showed that slight adjustments to the environment produced large changes in behavior. The goal of identifying crucial environmental variables spread to education, behavioral economics, and other areas.

A consensus emerged that humans were battered about by the random fluctuations in their environment. People expect more damage from hurricanes with male names than female names. Judges give out harsher sentences just before lunch, compared to immediately after. Ovulating single women were more likely to vote for Barack Obama. Standing in a confident pose for a short time produces changes in hormone levels and a higher tolerance for risk. Signing an honesty pledge at the beginning of a form makes people report more truthfully than signing at the end. Hundreds of other examples were available, and newspaper articles, psychology textbooks, “life hack” blogs, TED talks, and other media outlets filled with surprising findings about the malleability of human behavior.

Despite the thousands of studies, much of it was a house of cards. Social psychology was the first major casualty of psychology’s replication crisis, and social psychologists have no one to blame but themselves for their field’s fall.

One of the most important triggers of psychology’s replication crisis was the discovery in 2011 of a trio of social psychologists— Lawrence Sanna, Dirk Smeesters, and Diederik Stapel—who had independently fabricated data in their studies. Of the trio, the most prominent and prolific fraudster was Stapel, who had fabricated data in over 50 studies. In the report of Stapel’s conduct, the investigators chastised the field of social psychology for creating an environment where fraud, sloppy research, and inaccurate findings flourished.

The same year, another social psychologist, Daryl Bem, published an article claiming that he had observed precognition in nine different studies. This article—with results that were, literally, not possible—cried out for explanation. Several months later, an article appeared that did just that. Though the authors did not mention Bem by name, they showed that widespread statistical practices in social psychology (and other fields) could greatly inflate the strength of the evidence and make false findings appear to be true.

Social psychology was thus subjected to scrutiny, but it was still possible to write off these situations as anomalies. That changed drastically in 2015 with the publication of an article reporting 97 replications in social and cognitive psychology. Only 25-29% of social psychology studies replicated, while cognitive psychology fared better (50-55% of studies replicating).

Over time, the characteristics of studies that did not replicate became clear. Among others, these characteristics included research procedures that were not declared in advance, small sample sizes, flexibility in data analysis, failing to report all results or analyses, researchers motivated to find a particular outcome, and results that were too “clean.” These characteristics were widespread in social psychology—and many other areas of the social sciences.

While none of the landmark studies in social psychology suffer from all these characteristics, all the studies have some of them. For example, Sharif’s study consisted of just 24 boys, and Zimbardo drew his conclusions on the basis of 18 Stanford University students. Later archival research by psychologist Gina Perry showed that Milgram did not fully report his results. Perry also found that Sharif did not report the results of a similar summer camp where two groups of boys failed to develop a rivalry. Such selective reporting of results may make for a more exciting story (and these social psychologists’ stories are exciting), but it also presents the illusion that the findings are stronger and more consistent than they really are.

Moreover, some of these famous social psychologists presented themselves as dispassionate social scientists, but behind the scenes were engineering their studies to get the results they desired. Sharif and his research assistants purposely created conflict that did not exist by creating problems or inconveniences for the boys and attributing them to the other group. Zimbardo coached his “guards” and treated his “prisoners” worse than real prisoners. In the Stanford Prison Experiment, “prisoners” had no access to fresh air or exercise; wore chains on their feet constantly; were denied access to showers, cigarettes, and even their own glasses; and were subjected to erratic sleeping and toilet schedules. Zimbardo also sent out a press release on the second day of the study describing its results, even though it was supposed to last two weeks and had barely begun.

Asch, Sharif, Milgram, and Zimbardo were all highly motivated to find remedies for social problems. Asch and Milgram were invested in understanding the Holocaust. Zimbardo was a prison reform advocate before his study started. Sharif was desirous to find a way to reduce group tensions because while he was living there, Turkey engaged in genocides against Armenians, Assyrians, and Greeks. The Greco-Turkish war in the early 1920s resulted in hundreds of thousands dead and over 1.5 million people forced from their homes so that Greece and Turkey could have an ethnically homogeneous population.

In hindsight, it was ridiculous to draw conclusions about intransigent social problems from these dramatic studies. Sharif took American boys with a common social background and created an arbitrary social division. Removing short-term animosity based on the outgroup’s perceived behavior probably never could teach the world much about reducing long-standing hatred between groups—some of which is based on real differences in biology (e.g., skin color), language, or cultural practices (e.g., religion). Likewise, the sample sizes for Asch’s, Zimbardo’s, and Sharif’s studies were small, and Asch and Zimbardo only studied college students. It is not clear what the behaviors of such a small, unrepresentative sample of humanity could teach psychologists.

In many ways, social psychology’s most dramatic, memorable studies make more sense if considered more as performance art than scientific experiments. Perhaps this is why Zimbardo stated in 2018 that, “It shouldn’t have been called the Stanford Prison Experiment. It should have been called the Stanford Prison Experience.”

Amid all the questions about replication in social psychology, it is ironic that Asch’s study was subjected to replications: in 1974, only about one-third of male college students conformed to the opinion of a group at least once; by 1981, only 3% did. But no one proposed at the time that the original study was flawed. Instead, the psychologists conducting these replications theorized post hoc that the 1950s were a much more conformist time than later eras were. (A recent replication in Eastern Europe, in which 40% of respondents conformed at least once, has muddied the waters, but it is still far below the percentage in Asch’s original study.) A replication of Zimbardo’s study also failed to produce the same results as the original Stanford Prison Experiment. In the replication—in which no one was coached— the “guards” were reluctant to impose their authority, and the “prisoners” banded together to extract extra privileges from the “guards” and later staged a successful rebellion. Yet, none of this stopped Asch’s conformity studies or the Stanford Prison Experiment from becoming a staple of psychology textbooks.

In response to these problems, many of social psychology’s leaders were dismissive or downright hostile. Susan Fiske (who was the editor that accepted the ridiculous hurricane name study) called those conducting replications and criticizing weak studies were “self-appointed data police” engaging in “methodological terrorism.” In response to the 2015 article reporting that only one-quarter of social psychology studies replicated, Richard Nisbett said the project involved, “a huge amount of incompetence.” He emphatically declared, “There is not a replication crisis in psychology and never has been.” After a replication failed to produce the same results as the Stanford Prison Experiment, Zimbardo stated that the replication effort “fraudulent” and “disreputable.”

While these reactions do not represent the opinion of every social psychologist, most of the field’s leaders have been remarkably unwilling to change. In contrast to researchers in other areas of psychology (such as educational or personality psychology), many social psychologists have generally not seen fraud and poor replication rates as a call to improve their science. Instead, there have been efforts to excuse prior work and dismiss tough questions about the validity of the field’s research. Meanwhile, the problems keep accruing. More of the field’s foundational ideas are questioned or completely discredited, and there are new accusations of fraud in high-profile studies using social psychology’s methods and theories.

As a result, social psychology’s stock has never been lower. Daniel Kahneman, one of the few psychologists to win a Nobel Prize, epitomized the change in attitude towards social psychology when he recently stated, “When I see a surprising finding, my default is not to believe it. . . . Twelve years ago, my default was to believe anything that was surprising.” Other psychologists regularly express their reservations about social psychology—opinions that would have been considered extreme or unrealistic before 2011.

The rise of social psychology took decades, but its fall happened over the course of just a few years. Whereas in the past, social psychology’s most memorable studies were cited for their insights into the dark side of human behavior, today they are held up as examples of the field’s shoddy approach to research. It is unclear whether social psychology can ever regain its past prestige. If it does, it will not be through memorable stories that arise from dramatic studies. Instead, it will be through careful research using high-quality methods that emphasize truth-finding, protect against confirmation bias, and guard against false positives and over-generalizations. There is already a name for this strategy: science. Hopefully, social psychology can embrace it.

Russell T. Warne has published over 60 scholarly articles in peer reviewed journals and is the author of Statistics for the Social Sciences: A General Linear Model Approach and In the Know: Debunking 35 Myths About Human Intelligence. He was an associate professor at Utah Valley University until 2022.

---