Why do people agree with a group when the group is clearly wrong? Why might someone follow an authority figure’s instructions despite feeling uncomfortable? And why can a room full of witnesses sometimes hesitate to help?
Social psychology experiments investigate how other people—and our beliefs about them—shape thoughts, feelings, and behavior. By changing specific features of a social situation, researchers can test explanations for conformity, obedience, prejudice, persuasion, and cooperation.
The best-known studies offer memorable examples, but their findings need careful interpretation. Some used narrow samples, some raised serious ethical concerns, and some became famous for conclusions that their designs cannot firmly support.
This guide explains 13 influential social psychology experiments: what researchers did, what they found, how the proposed psychological mechanisms work, and where the evidence stops.
What are social psychology experiments?
Social psychology experiments are studies in which researchers deliberately change a social condition and measure its effect on an outcome.
The changed condition is the independent variable. The measured outcome is the dependent variable. For example, a researcher might vary whether a participant answers a question alone or after hearing a group’s answers, then measure the participant’s accuracy.
A well-designed comparison helps separate the effect of the social condition from other explanations. Random assignment—allocating participants to conditions by chance—helps prevent preexisting differences between groups from systematically determining the results.
Three distinctions make these experiments easier to understand:
- Conformity: adjusting a judgment or behavior to align with a group.
- Compliance: agreeing to a request.
- Obedience: following an instruction from someone perceived as an authority.
These behaviors can look similar from the outside, but they involve different relationships and pressures.
Famous social psychology experiments at a glance
| Experiment | Main topic | Central question |
|---|---|---|
| Asch’s conformity experiments | Majority influence | Will people agree with an obviously incorrect group? |
| Sherif’s autokinetic experiment | Social norms | How do shared standards develop under uncertainty? |
| Milgram’s obedience experiment | Authority | How far will participants follow troubling instructions? |
| Stanford Prison Experiment | Roles and research validity | What can a simulated prison actually establish? |
| Robbers Cave experiment | Intergroup conflict | Can competition create hostility and cooperation reduce it? |
| Bobo doll experiment | Observational learning | Can children learn aggressive actions by watching adults? |
| Festinger and Carlsmith’s experiment | Cognitive dissonance | Can weak external justification encourage attitude change? |
| Bystander seizure experiment | Diffusion of responsibility | Does believing others can help delay intervention? |
| Smoke-filled room experiment | Interpretation of emergencies | Can other people’s calm behavior suppress reporting? |
| Good Samaritan experiment | Time pressure | Can being hurried affect whether someone helps? |
| Minimal group experiments | Ingroup favoritism | Can a minor group distinction influence resource allocation? |
| Foot-in-the-door experiments | Sequential requests | Does a small initial commitment increase later compliance? |
| Halo effect experiment | Social judgment | Can an overall impression alter ratings of specific traits? |
1. Asch’s conformity experiments: when the majority is wrong
Researcher: Solomon Asch
Period: Early 1950s; a widely cited account appeared in 1955
Main concept: Conformity
What happened?
Participants judged which of three comparison lines matched a target line. The task was straightforward, but a genuine participant answered after several confederates—people working with the researcher—who deliberately chose the same incorrect answer on selected trials.
What did the experiment find?
In the standard condition described by Asch, participants followed the incorrect majority on roughly one-third of the critical judgments. A substantial minority remained independent throughout. Introducing a supporter sharply reduced conformity.
How does it work?
One explanation is normative social influence: someone may know the answer but avoid publicly disagreeing. Another is informational social influence: unanimous disagreement can make someone question their own perception.
Public agreement therefore does not automatically demonstrate private belief.
Why does it matter?
Imagine a meeting where everyone supports an inaccurate forecast. An employee may stay silent to avoid standing out. Asch’s work provides a framework for understanding that pressure.
The limitation is equally useful: these were controlled perceptual judgments with a narrow participant population. The findings do not establish that everyone conforms, or that the same rate applies to consequential workplace decisions.
2. Sherif’s autokinetic experiment: how social norms form
Researcher: Muzafer Sherif
Year: 1935
Main concept: Norm formation under uncertainty
What happened?
Participants viewed a stationary point of light in a dark room and estimated how far it moved. Without visible reference points, the light appeared to move—a perceptual illusion called the autokinetic effect.
Participants made estimates individually and in groups.
What did the experiment find?
Individual estimates developed their own ranges. When participants made judgments together, their estimates converged toward a shared standard. Group-influenced standards could persist when participants later answered alone.
How does it work?
When objective evidence is ambiguous, other people’s judgments provide a reference point. Over repeated exchanges, a group can establish what feels like a reasonable answer.
Persistence outside the group suggests more than temporary public agreement: the shared standard can become part of an individual’s judgment.
Why does it matter?
The study offers an analogy for how newcomers learn informal expectations, such as what counts as an acceptable workload.
Its scope is specific, however. It examined judgments about an ambiguous visual stimulus. Extending its findings to organizational culture or political beliefs requires additional evidence.
Compared with Asch’s clear line task, Sherif’s experiment highlights influence when people have good reason to feel uncertain.
3. Milgram’s obedience experiment: following an authority figure
Researcher: Stanley Milgram
Publication year: 1963
Main concept: Obedience to authority
What happened?
Forty men took part in what they believed was a learning study. Assigned the teacher role, they were instructed to deliver increasingly strong electric shocks when a learner made mistakes.
The learner was a confederate and received no shocks. Participants faced a generator labeled up to 450 volts and encouragement from the experimenter to continue.
What did the experiment find?
In the reported condition, 26 of 40 participants—65%—continued to the highest setting. Many displayed considerable distress.
How does it work?
The procedure combined an apparently legitimate scientific authority, incremental escalation, and pressure to continue an established task. These features help explain why distress did not always produce refusal.
However, the original condition did not independently isolate every proposed mechanism.
Why does it matter?
The study raises questions about responsibility within hierarchies: when does following a procedure discourage independent judgment?
The 65% figure belongs to this particular sample and procedure. It is not an estimate of how many people would commit violence on command. Deception, distress, and uncertainty about participants’ understanding also complicate interpretation and ethical evaluation.
4. The Stanford Prison Experiment: a famous study with serious flaws
Researchers: Philip Zimbardo and colleagues
Year: 1971
Main concepts: Social roles, institutional power, and methodological validity
What happened?
Male student volunteers were assigned prisoner or guard roles in a simulated prison. The study ended after six days amid mistreatment and participant distress.
It became widely associated with the claim that assigning ordinary people powerful roles can rapidly make them abusive.
What can the evidence support?
That familiar conclusion is much stronger than the design warrants. Archival research by Thibault Le Texier identified problems including instructions given to guards, incomplete data collection, and evidence that participants were responding to perceived expectations.
Why do those problems matter?
Demand characteristics are cues that suggest how researchers expect participants to behave. If guards receive guidance that encourages harsh treatment, their conduct cannot cleanly demonstrate the effect of the guard role alone.
Researchers’ involvement in the simulation further complicates interpretation.
Why include it?
It is historically influential and useful for teaching critical evaluation. It should not be treated as decisive proof that power inevitably produces cruelty.
The study illustrates why a dramatic event, even one presented as an experiment, may leave competing causal explanations unresolved.
5. The Robbers Cave experiment: competition and cooperation
Researchers: Muzafer Sherif and colleagues
Year: 1954; full report published in 1961
Main concept: Intergroup conflict
What happened?
Researchers organized boys at a summer camp into two groups. After separate activities encouraged group identities, competitions placed the groups in opposition.
Later, researchers introduced superordinate goals: desirable outcomes that required both groups to cooperate, such as resolving a shared water-supply problem.
What did the experiment find?
Competition was accompanied by hostility and unfavorable views of the other group. Simply bringing the groups together did not readily resolve the conflict. A series of shared tasks requiring cooperation was followed by improved relations.
How does it work?
Competition can make another group an obstacle to something valuable. Shared goals change that relationship: the other group becomes necessary for success.
The proposed mechanism concerns the structure of interdependence, rather than a general instruction to be friendly.
Why does it matter?
A possible organizational application is to examine whether departments are rewarded for competing or for solving shared problems.
However, the camp involved a small, homogeneous sample and substantial researcher orchestration. Its results cannot establish that shared goals alone will resolve conflicts involving unequal power, historical grievances, or incompatible interests.
6. The Bobo doll experiment: learning by watching others
Researchers: Albert Bandura, Dorothea Ross, and Sheila Ross
Year: 1961
Main concept: Observational learning
What happened?
Children observed an adult behaving aggressively toward an inflatable Bobo doll, behaving nonaggressively, or received no model exposure. Researchers later observed the children’s behavior in a setting containing the doll and other toys.
What did the experiment find?
Children exposed to the aggressive model reproduced more of the modeled aggressive actions and speech than children in the comparison conditions. They did so when the adult model was absent.
How does it work?
Observational learning means acquiring information about behavior by watching someone else. The observer need not personally receive a reward for every action before learning how to perform it.
A model can also communicate which actions appear available or acceptable in a particular setting.
Why does it matter?
The experiment provides a clear example of how demonstrations can shape behavior, making it relevant to discussions of adult role models.
Its outcome was short-term behavior involving a toy designed to be struck. That is different from demonstrating lasting aggression toward people. Broader claims about violent media or long-term development require studies that measure those outcomes directly.
7. Festinger and Carlsmith’s experiment: cognitive dissonance
Researchers: Leon Festinger and James Carlsmith
Year: 1959
Main concept: Insufficient justification
What happened?
Participants completed repetitive tasks. Some then received either $1 or $20 to tell another person that the tasks were enjoyable. Researchers subsequently measured participants’ own evaluations of the experience.
What did the experiment find?
Participants paid $1 rated the tasks more favorably than those paid $20 and the control group on the enjoyment measure.
How does it work?
Cognitive dissonance refers to psychological tension associated with inconsistent cognitions, such as believing a task was boring while knowing one described it as enjoyable.
A large payment offers a clear external explanation: the person said it for the money. A small payment provides less justification, potentially encouraging an attitude shift that makes the statement feel more consistent with personal belief.
Why does it matter?
The study helps explain why behavior can shape attitudes, reversing the familiar assumption that attitudes always come first.
It does not establish that low pay improves motivation or makes unpleasant work enjoyable. The result concerned a particular counterattitudinal statement under controlled conditions, not the general relationship between compensation and job satisfaction.
8. The bystander seizure experiment: diffusion of responsibility
Researchers: John Darley and Bibb Latané
Year: 1968
Main concept: Diffusion of responsibility
What happened?
Participants believed they were discussing personal problems through an intercom. During the discussion, another apparent participant seemed to experience a seizure.
Researchers varied whether the genuine participant believed they were the only other listener or that additional listeners were present.
What did the experiment find?
Believing that other people could also hear the emergency reduced the speed of reporting it. The presence of other potential helpers could inhibit an individual’s response.
How does it work?
With diffusion of responsibility, a person feels less individually obligated to act because responsibility appears shared.
Someone may recognize a serious problem yet assume that another listener will respond. This differs from failing to care about the person in distress.
Why does it matter?
A practical inference is to make responsibility explicit when coordinating assistance: address a particular person and specify the needed task.
The experiment did not demonstrate that crowds always prevent help. It tested perceived listeners in an unusual communication arrangement. Whether others can coordinate, communicate, or provide useful support can change the situation substantially.
9. The smoke-filled room experiment: reading other people’s reactions
Researchers: Bibb Latané and John Darley
Year: 1968
Main concept: Social interpretation of ambiguous events
What happened?
Male students completed questionnaires while smoke entered the room. Participants were alone, with two confederates who remained passive, or with other genuine participants.
What did the experiment find?
Seventy-five percent of participants tested alone reported the smoke, compared with 10% of participants seated with passive confederates. Smoke was reported in 38% of the three-person groups of genuine participants. The last figure measures groups, so it should not be treated as an individual reporting rate.
How does it work?
The findings illustrate a process associated with pluralistic ignorance: people may privately feel concerned but interpret others’ outward calm as evidence that concern is unnecessary.
If everyone watches everyone else, uncertainty can sustain collective inaction.
Why does it matter?
This differs from diffusion of responsibility. The central issue is deciding whether an emergency exists, rather than deciding who should respond.
The study helps explain why visible reactions matter under uncertainty. It does not imply that people will ignore clear danger in every group setting; the ambiguity of the event is central to the explanation.
10. The Good Samaritan experiment: the effect of being in a hurry
Researchers: John Darley and C. Daniel Batson
Year: 1973
Main concept: Situational influences on helping
What happened?
Seminary students prepared to give a talk and walked to another building. Some were told they were late; others experienced less time pressure.
On the way, they encountered someone slumped in a doorway who appeared to need assistance. Some participants were preparing a talk about the Good Samaritan; others had a different topic.
What did the experiment find?
Participants under greater time pressure were less likely to stop. Preparing to discuss the Good Samaritan did not produce a statistically significant difference in whether they helped.
How does it work?
Hurry can alter what receives attention and make stopping feel costly. That is a plausible interpretation; the experiment did not directly establish every mental step behind the decision.
Why does it matter?
The study encourages a broader explanation of helpfulness that includes immediate circumstances alongside personal values.
Its small, specialized sample limits generalization. A nonsignificant topic effect also does not prove that moral reflection never matters. It means this study did not establish such an effect on the measured decision.
11. Tajfel’s minimal group experiments: how easily “us” and “them” emerge
Researchers: Henri Tajfel and colleagues
Publication year: 1971
Main concept: Ingroup favoritism
What happened?
Participants were categorized using relatively trivial distinctions, such as supposed visual judgments or artistic preferences. They then allocated rewards to anonymous individuals identified by group membership.
The procedures minimized prior relationships and direct personal gain from particular allocations.
What did the experiments find?
Participants often favored members of their own group. In some allocation choices, creating a favorable difference between groups mattered more than maximizing the ingroup’s absolute reward.
How does it work?
An ingroup is a group someone identifies with; an outgroup is a comparison group outside that membership.
The findings showed that even minimal categorization can influence decisions. They helped motivate social identity approaches, which examine how group membership contributes to self-definition and evaluation.
Why does it matter?
The studies offer a starting point for examining favoritism in teams or institutions.
However, allocating points in a laboratory is different from enduring prejudice, discrimination, or violence. These experiments identify a possible foundation of group bias; they do not provide a complete explanation of conflicts shaped by history, social norms, and unequal resources.
12. The foot-in-the-door experiments: from small requests to larger commitments
Researchers: Jonathan Freedman and Scott Fraser
Year: 1966
Main concept: Sequential-request compliance
What happened?
Researchers examined whether accepting a small request made people more willing to accept a larger request later.
Their studies included household requests and a well-known field procedure involving small initial actions followed by a request to display a large road-safety sign.
What did the experiments find?
Under some conditions, people who had accepted a small initial request were more likely to agree to a later, larger one. The effect could occur even when different people made the two requests.
How does it work?
One interpretation is that an initial action changes self-perception. Someone may begin to see themselves as helpful or supportive of a cause, making a related request feel consistent with that identity.
However, observing increased compliance does not uniquely establish that explanation.
Why does it matter?
The experiment provides a useful framework for understanding how commitments can develop gradually.
It does not mean that a small agreement guarantees a larger one. The specific requests and circumstances matter. An ethical application should preserve an easy, informed opportunity to decline each new request rather than treating earlier agreement as continuing consent.
13. The halo effect experiment: when an impression changes the details
Researchers: Richard Nisbett and Timothy Wilson
Year: 1977
Main concept: The halo effect
What happened?
Students watched one of two videotaped interviews featuring the same instructor. In one, he behaved warmly; in the other, he behaved coldly.
Participants evaluated the instructor and characteristics including his appearance, mannerisms, and accent.
What did the experiment find?
Students who saw the warm interview evaluated those characteristics more favorably. Participants did not adequately recognize how the overall impression had influenced their judgments.
How does it work?
The halo effect occurs when a general evaluation spills over into judgments of specific attributes.
Instead of evaluating each feature independently, an observer may interpret it through an existing impression. The same accent, for example, may seem appealing or irritating depending on feelings about the speaker.
Why does it matter?
A practical inference for assessment is to define separate criteria and seek evidence for each. A pleasant interview should not automatically count as evidence of technical competence.
The study concerned ratings of one instructor in a controlled setting. It illustrates judgment contamination without establishing that every favorable impression distorts every assessment.
How to interpret social psychology experiments critically
Famous social psychology experiments are most useful when their conclusions stay connected to their methods. Four questions help keep that connection clear.
Did the study measure the outcome being claimed?
A child hitting a doll, a participant selecting a line, and a person obeying simulated shock instructions are specific observed behaviors.
Claims about lasting aggression, political conformity, or real-world violence require further steps. Researchers must show that the experimental measure meaningfully relates to the broader outcome.
Could another feature explain the result?
A confound is a factor that varies with the condition of interest and offers an alternative explanation.
If a researcher changes both a person’s role and the instructions they receive, the resulting behavior may reflect either influence. A striking difference does not resolve that ambiguity by itself.
Who participated, and where?
External validity concerns whether findings generalize beyond the studied people and circumstances.
A study involving male students in one country may establish something valuable about that setting while leaving questions about other ages, cultures, and institutions unanswered. A narrow sample is a boundary on the claim, not automatic evidence that the finding is false.
Does the finding replicate?
Replication tests whether a result can be observed again with new data. In a major 2015 project, the Open Science Collaboration replicated 100 studies from three psychology journals and found that replication effects were, on average, substantially smaller than original effects.
That project did not test all 13 studies in this article and should not be used to label them collectively as disproven.
The practical lesson is to consider the broader evidence: repeated tests, methodological quality, effect sizes, and the conditions under which an effect appears.
Ethical questions raised by classic experiments
Several classic studies depended on deception, distress, or participants misunderstanding the situation. Those features create questions separate from whether the results are interesting:
- Meaningful consent: What did participants understand about the experience they were agreeing to?
- Freedom to withdraw: Could they stop without pressure or fear of consequences?
- Protection from harm: Were distress and other risks justified and adequately controlled?
- Debriefing: Were misleading explanations corrected afterward?
These questions also affect scientific interpretation. For example, pressure to remain in a study can influence the very behavior being measured.
Ethical scrutiny therefore belongs within the explanation of an experiment, alongside its design and findings.
Frequently asked questions about social psychology experiments
What do social psychology experiments study?
They study how social conditions influence thinking, feeling, and behavior. Common topics include group pressure, authority, helping, prejudice, persuasion, cooperation, and judgments about other people.
What is the difference between conformity and obedience?
Conformity involves aligning with a group’s behavior or judgments. Obedience involves following an authority figure’s instruction. Asch’s line judgments illustrate conformity, while Milgram’s procedure examined obedience.
Do these experiments prove that situations determine behavior?
No. They show that particular situations can influence measured behavior. Individuals still differ, and the strength of an effect depends on the procedure and context. Explaining an action usually requires attention to both the person and the situation.
Is the Stanford Prison Experiment reliable evidence about human nature?
Its methodological problems prevent it from supporting sweeping conclusions about inevitable cruelty or the automatic effects of social roles. It is best approached as a historically influential and heavily criticized study, rather than decisive evidence.
Why are these experiments still studied?
They introduce important research questions and experimental methods. They also teach readers to distinguish observations from explanations, examine ethical tradeoffs, and question conclusions that extend beyond the evidence.
What these experiments help us understand
Across these 13 social psychology experiments, researchers examined several ways social life can shape behavior: groups provide reference points, authorities create pressure, commitments influence later choices, and impressions affect judgments.
Their value lies in making those influences testable. To understand a finding well, connect it to a specific situation, ask what was actually measured, and examine which alternative explanations remain.
That approach makes these studies useful tools for thinking about human behavior while keeping their conclusions proportionate to the evidence.
Leave a Reply