Social Influence

Social influence is the process by which people change their thoughts, feelings or behavior in response to real or imagined pressure from others. Psychologists usually describe three main forms: conformity (going along with a group), compliance (agreeing to a direct request) and obedience (following orders from an authority). Social influence is a core topic in social psychology, and it is studied through some of the field's most famous experiments.

Influence is not inherently good or bad. The same processes that lead people to copy a harmful trend also help them learn customs, cooperate in teams, and behave safely when they don't know what to do. Understanding how influence works helps explain everything from fashion and voting to emergencies, workplace decisions and online behavior.

Types of Social Influence

Conformity

Conformity means changing behavior or beliefs to match those of a group. Morton Deutsch and Harold Gerard (1955) distinguished two motives behind it.

Informational social influence happens when we follow others because we believe they know better. It is strongest when a situation is ambiguous, new or urgent, such as following other people toward an exit when an alarm sounds in an unfamiliar building. It often produces genuine, private acceptance: we really come to believe the group is right.

Normative social influence happens when we go along with others to be liked, accepted or to avoid disapproval. Laughing at a joke you didn't find funny, or agreeing with a group opinion you privately doubt, are everyday examples. It tends to produce public compliance without private agreement.

Herbert Kelman (1958) added a further distinction between three levels of conformity: compliance (public agreement only), identification (adopting a group's views because you value membership of that group) and internalization (accepting the views as your own, so they persist even when the group is absent).

Compliance

In social psychology, compliance also refers to agreeing to an explicit request, often from someone without formal authority, such as a salesperson, a charity worker or a friend. Compliance does not require any change in private attitude. Researchers have studied many techniques that increase it:

• Foot-in-the-door: agreeing to a small request makes people more likely to agree to a larger one later (Freedman and Fraser, 1966).
• Door-in-the-face: after refusing a large request, people are more likely to agree to a smaller one, partly because it feels like a concession that should be reciprocated (Cialdini and colleagues, 1975).
• Low-ball: people commit to a deal, then the terms worsen, yet many stick with their commitment.

Robert Cialdini's book Influence (1984) grouped such tactics under broad principles including reciprocity, commitment and consistency, social proof, liking, authority and scarcity. Our guide to persuasion and compliance covers them in detail.

Obedience

Obedience is following a direct order from someone perceived as a legitimate authority. Unlike conformity, the pressure is explicit and comes from someone of higher status, and unlike ordinary compliance, it involves commands rather than requests. Obedience is necessary for functioning institutions, from hospitals to air traffic control, but it becomes troubling when it leads people to act against their own conscience.

Classic Experiments in Social Influence

Sherif's autokinetic studies (1935)

Muzafer Sherif used a visual illusion called the autokinetic effect: a stationary point of light in a dark room appears to move. When people estimated the movement alone, their estimates varied widely. When they made estimates in small groups, their answers converged on a shared group norm, and individuals continued to use that norm when later tested alone. Because there was no correct answer, the study is a classic demonstration of informational influence and of how norms form. Later work showed such arbitrary norms could be passed down across several "generations" of group members as people were gradually replaced.

Asch conformity experiments (1951–1956)

Solomon Asch asked people to judge which of three lines matched a target line, a task so easy that people working alone made errors less than 1% of the time. Each real participant sat with a group of confederates who, on certain trials, unanimously gave the same wrong answer. Across these critical trials, participants conformed roughly a third of the time; about 75% conformed at least once, and about 25% never conformed.

Asch's variations showed what drives conformity. It rose as the majority grew to about three people and then leveled off. It fell sharply when just one other person gave the correct answer, and when participants could write their answers privately. Interviews suggested most conformers knew the majority was wrong but went along to avoid standing out, which points to normative influence. A 1996 meta-analysis by Rod Bond and Peter Smith found that conformity in Asch-type studies had declined in the United States since the 1950s and tended to be higher in more collectivist cultures. See our full guide to the Asch conformity experiment.

Milgram obedience studies (1961–1962)

Stanley Milgram at Yale recruited men from the community for what they were told was a study of learning. Each participant, as the "teacher," was instructed to give increasingly strong electric shocks to a "learner" for wrong answers. The learner was a confederate and received no real shocks. In the baseline condition reported in 1963, all 40 participants continued to at least 300 volts, and 26 (65%) went to the final 450-volt level, many while showing obvious distress.

Milgram then ran around 20 variations, and obedience changed dramatically with the situation. It dropped when the learner was in the same room, when the teacher had to physically place the learner's hand on a plate (about 30%), when the experimenter gave orders by telephone (about 20%), and when two peer "teachers" refused to continue (10%). Moving the study to a run-down office building reduced obedience somewhat. Jerry Burger's partial replication in 2009, stopping at 150 volts for ethical reasons, found rates similar to Milgram's. Archival work by Gina Perry and others has since raised questions about how consistently the scripted procedure was followed and how many participants believed the shocks were real. For the full story, see our guide to the Milgram experiment.

Stanford prison experiment (1971)

Philip Zimbardo's study randomly assigned student volunteers to be "guards" or "prisoners" in a mock prison at Stanford. It was stopped after six days instead of the planned two weeks, after some guards behaved abusively and several prisoners showed acute distress. It was long presented as proof that situations alone can turn ordinary people cruel. Later analysis, including Thibault Le Texier's archival research, showed that guards were given guidance on how to behave, that experimenters were actively involved, and that only some guards were harsh. The study had no control group and was never published in a major peer-reviewed journal as a standard experiment, so it is now treated as a vivid demonstration rather than strong evidence. Our Stanford prison experiment page covers the debate.

Group Dynamics and Processes

Influence does not only flow from a majority or an authority to an individual. Groups also change how their members think and perform. Our guide to group dynamics explores these processes further.

Group polarization

In the early 1960s, James Stoner found that groups often made riskier decisions after discussion than their members had made individually, a pattern called the "risky shift." Later research showed that discussion can also produce a "cautious shift," and that the general rule is group polarization: discussion tends to push a group toward a more extreme version of the view its members already leaned toward. Two explanations are well supported: people hear more persuasive arguments for the dominant position, and they shift to present themselves favorably relative to the group norm.

Groupthink

Irving Janis (1972) coined groupthink to describe faulty decision-making in highly cohesive groups that value agreement over critical evaluation. He analyzed policy failures such as the 1961 Bay of Pigs invasion. Symptoms include an illusion of unanimity, self-censorship, pressure on dissenters and "mindguards" who shield the group from contrary information. Recommended safeguards include assigning a devil's advocate, inviting outside experts and having leaders withhold their own view at first. Research has supported some elements of the model more than others.

Social facilitation

Norman Triplett (1898) observed that cyclists rode faster when racing others than when riding alone against the clock. Later studies found that an audience sometimes helps and sometimes hurts performance. Robert Zajonc (1965) proposed a resolution: the presence of others increases arousal, which strengthens the dominant (most practiced) response. On simple or well-learned tasks this improves performance; on complex or new tasks it impairs it. Concern about being evaluated also plays a role.

Social loafing

Social loafing is the tendency to put in less effort on a group task when individual contributions can't be identified. It was first noticed by Max Ringelmann in rope-pulling studies and was systematically studied by Bibb Latané and colleagues (1979) using clapping and shouting. A 1993 meta-analysis by Steven Karau and Kipling Williams found loafing is reduced when contributions are identifiable, the task is meaningful, the group is valued and the group is small. Some studies suggest it is weaker in collectivist cultures.

Deindividuation

Deindividuation refers to a reduced sense of personal identity in groups, especially under anonymity and arousal. Classic accounts linked it to impulsive or antisocial behavior, such as crowd violence. Later research, notably the social identity model of deindividuation effects (SIDE), suggests that anonymity doesn't simply release people from norms. Instead, it often makes them follow the norms of the group they identify with more closely, which can be prosocial or antisocial depending on the group.

Social Norms and Their Influence

Social norms are shared expectations about how people behave or should behave. Robert Cialdini and colleagues distinguish descriptive norms (what most people actually do) from injunctive norms (what people approve or disapprove of). Both shape behavior, and messages work best when they point in the same direction.

Field research shows how powerful this can be. In a study of hotel guests, Goldstein, Cialdini and Griskevicius (2008) found that a card saying most previous guests in the same room reused their towels increased reuse more than a standard environmental appeal. Schultz and colleagues (2007) found that telling households how their energy use compared with neighbors reduced use among high consumers, but could raise it among low consumers unless the message also included approval (a smiley face) for using less. This "boomerang effect" shows why norm-based campaigns need careful design.

Norms are learned through observation and instruction, a process linked to social learning theory, and enforced through approval, gossip and exclusion. Even being briefly ignored by strangers can be distressing, which is one reason the threat of rejection is such a strong motivator.

Minority Influence

If majorities always won, social change would be impossible. Serge Moscovici showed that a consistent minority can shift the majority. In a 1969 study with Lage and Naffrechoux, groups of six judged the color of blue slides. When two confederates consistently called the slides "green," real participants said "green" on about 8% of trials, compared with about 1% in a control group; when the confederates were inconsistent, the effect largely disappeared.

Minorities seem to be most effective when they are consistent over time, appear committed and confident, are flexible rather than dogmatic, and argue in line with broader social trends. Moscovici's conversion theory proposed that while majorities often produce public compliance, minorities encourage people to think more carefully about the issue, sometimes producing private or delayed attitude change. Charlan Nemeth's research suggests dissenting minorities can also stimulate more creative, divergent thinking in groups.

Bystander Effect and Helping Behavior

The bystander effect is the finding that individuals are less likely to help in an emergency when other people are present. Research on it was prompted by the 1964 murder of Kitty Genovese in New York, which a newspaper reported as having been watched by 38 people who did nothing. Later investigations showed that report was seriously inaccurate: far fewer people saw any part of the attack, most heard only fragments, and at least one person called the police and another went to help. The research it inspired, however, stands on its own.

Latané and Darley proposed that to help, a bystander must notice the event, interpret it as an emergency, take responsibility, know how to help and decide to act. Others' presence interferes at several steps: through diffusion of responsibility (assuming someone else will act), pluralistic ignorance (seeing others stay calm and concluding nothing is wrong) and audience inhibition (fear of looking foolish).

The picture is more nuanced than "crowds don't help." A 2011 meta-analysis by Fischer and colleagues found that the bystander effect is weaker in clearly dangerous emergencies, and a 2020 analysis of CCTV footage of real public conflicts in three countries found that someone intervened in about nine out of ten incidents. Helping becomes more likely when the need is clear, when someone is directly asked for help, when bystanders know each other or the victim, and when others model helping. See also prosocial behavior.

Power and Status Dynamics

Who can influence us depends heavily on power and status. John French and Bertram Raven (1959) described five bases of social power, and Raven later added a sixth:

• Legitimate power comes from a recognized position or role, such as a manager or police officer.
• Reward power comes from the ability to provide benefits.
• Coercive power comes from the ability to punish.
• Expert power comes from knowledge or skill.
• Referent power comes from being liked or admired.
• Informational power comes from controlling useful information.

Research also suggests that having power changes people. The approach/inhibition theory of power (Keltner and colleagues, 2003) proposes that power makes people more action-oriented and less attentive to others' perspectives. Status characteristics such as age, gender or ethnicity can also shape whose views are taken seriously in groups, even when those traits have nothing to do with the task, a topic closely linked to prejudice and discrimination.

Cultural Variations in Social Influence

Social influence works everywhere, but its forms vary across cultures. Studies comparing more individualistic and more collectivist societies tend to find somewhat higher conformity in collectivist settings, where fitting in with the group may be seen as tact rather than weakness. In such settings, the classic social loafing pattern is sometimes reduced or reversed when working with in-group members.

Michele Gelfand and colleagues (2011) compared 33 nations on cultural "tightness" (strong norms and little tolerance for deviance) versus "looseness" (weaker norms and more tolerance). They found tighter cultures were more likely to have faced historical threats such as natural disasters, disease and conflict. Societies also differ in "power distance," the extent to which hierarchy and obedience to authority are accepted. Cross-cultural replications of Milgram-type studies have produced varying results, though differences in procedure make direct comparison difficult. Our guide to cross-cultural psychology covers these comparisons.

Modern Contexts and Digital Influence

Online environments create new channels for old processes. Visible likes, shares and follower counts act as signals of social proof, while recommendation algorithms shape which norms people see. Anonymity online can reduce restraint, a pattern sometimes called the online disinhibition effect, but as the SIDE model predicts, it can also strengthen identification with online groups and their norms.

Researchers debate how strong "echo chambers" really are. Many people encounter a wider range of views online than they choose to engage with, but selective attention, group polarization and rapid sharing can still amplify misinformation. Influencers draw on referent power and parasocial relationships (one-sided feelings of closeness to a media figure). For more on the wellbeing side of this, see our guide to social media and mental health.

Resistance to Social Influence

People are not passive targets. Several factors help individuals resist pressure:

• Social support: as Asch and Milgram both found, a single ally or a model of disobedience greatly reduces conformity and obedience.
• Privacy and time: responding anonymously or after reflection reduces normative pressure.
• Expertise and confidence: people who know a topic well conform less on it.
• Prior commitment: having publicly taken a position makes people less likely to abandon it.
• Personal responsibility: reminding people that they are accountable for their own actions reduces blind obedience.

Jack Brehm's reactance theory (1966) describes the motivation to restore freedom when it feels threatened: heavy-handed pressure can backfire and push people the opposite way. Some people also have a strong need for uniqueness, which leads them to avoid seeming like everyone else. Research on locus of control suggests people who believe they control their own outcomes may be somewhat more resistant to social pressure.

Applications and Interventions

Organizations. Understanding conformity and groupthink helps teams make better decisions, for example by encouraging dissent, using anonymous input and creating psychological safety so people feel able to speak up.

Education. Peer norms shape classroom behavior and bullying. Interventions that correct misperceived norms (for example, showing that most students disapprove of bullying) and cooperative learning structures draw directly on social influence research.

Public health. Social norms campaigns have been used to address issues such as student drinking and energy use, with mixed but sometimes useful results. Their success depends on accurate, credible information and avoiding boomerang effects.

Legal settings. Social influence research informs understanding of jury deliberation, eyewitness memory (which can be contaminated by co-witness discussion) and police interrogation, where pressure can produce false confessions. See forensic psychology.

Ethical Considerations

Many classic social influence studies relied on deception and exposed participants to stress. Milgram's participants were misled about the shocks, and some left the lab distressed, which fuelled debate that helped shape modern research ethics: informed consent, the right to withdraw, minimizing harm and thorough debriefing. Studies like Milgram's original would not be approved in that form today, which is why later researchers used modified designs such as Burger's 150-volt limit or virtual-reality versions.

There are applied ethical questions too. The line between legitimate persuasion and manipulation is often thin. Techniques that work in marketing or politics can bypass careful thought, so transparency and respect for people's autonomy matter. Our page on the psychology of influence discusses these issues.

Future Directions

Current research is moving in several directions. Neuroscientists are studying how the brain responds to disagreement with a group. Large, multi-site and cross-cultural studies are testing which classic effects replicate. Computational models and social network analysis track how behaviors and beliefs spread through populations. Researchers are also examining how algorithms, AI systems and virtual reality create new forms of social presence and influence, and how collective behavior can be mobilized for challenges such as public health crises and climate change.

Summary

Social influence describes how other people, real or imagined, shape what we think and do. Conformity, compliance and obedience each have distinct motives, and classic studies by Sherif, Asch and Milgram showed how powerful they can be. Later research has added nuance: influence depends on group size, unanimity, authority, culture and identity, and people can and often do resist. Knowing how these forces work makes it easier to recognize them, to use them responsibly, and to stand apart when it matters.