Hook65% of ordinary people were willing to electrocute a stranger
In 1963, at Yale, Stanley Milgram sat 40 ordinary men in front of a machine and told them to give a stranger electric shocks that rose to 450 volts, past a switch marked 'Danger: Severe Shock'. The stranger — really an actor — screamed, begged, and then went silent. A man in a grey lab coat simply said, 'The experiment requires that you continue.' Sixty-five per cent of participants went all the way to the top of the dial. They were not sadists. They were teachers, clerks and salesmen, and many were shaking, sweating and pleading to stop — while their finger kept pressing the switch.
That gap between what people felt and what they did is the whole of social influence: how the mere presence of other people changes our behaviour. This section covers four versions of that force — conforming to a group we want to fit into, obeying an authority we see as legitimate, helping (or walking past) a stranger in trouble, and losing ourselves inside a crowd. In every case the exam is testing the same insight Milgram stumbled on: to explain what a person did, look less at who they are and more at the situation they were in.
ModelConformity — going along with the group
Conformity is changing your behaviour or opinion because of real or imagined group pressure. Solomon Asch showed just how little pressure it takes. In his 1951 line-judgement study, participants sat with a group of actors and stated aloud which of three lines matched a standard. The answer was obvious — yet when the actors deliberately gave the same wrong answer, about 75% of real participants conformed to the group at least once, and on average people gave the wrong answer roughly a third of the time, against the plain evidence of their own eyes.
Asch then varied the situation to find what drives it. Conformity rose with group size up to about three or four confederates and then levelled off; it collapsed when just one other person broke the unanimous majority (losing unanimity freed people to dissent); and it rose when the task was harder and the right answer less obvious. Psychologists explain why with two motives. Normative social influence is conforming to be liked and accepted — you go along publicly to avoid standing out. Informational social influence is conforming to be right — when you are unsure, you assume the group knows better and genuinely change your mind. Harder tasks tip people towards informational influence; the desire to fit in drives normative.
ModelObedience — following an order from authority
Obedience differs from conformity: it is a direct response to an order from someone of higher status, not quiet pressure from equals. Milgram's explanation was the agentic state — the idea that when we obey a legitimate authority we stop seeing ourselves as personally responsible and instead act as an 'agent' carrying out their will. The opposite is the autonomous state, where we feel responsible for our own actions. Handing responsibility upwards is what let ordinary men keep pressing the switch.
Milgram's follow-up variations mapped the factors that raise or lower obedience. Proximity mattered both ways: when the learner was in the same room, obedience fell; when the experimenter gave orders by telephone rather than in person, it fell further. Legitimacy of authority and location mattered: moving the study from prestigious Yale to a run-down office block cut obedience, because the setting lent less credibility to the authority. And there is a dispositional side too — Adorno's idea of the authoritarian personality, a rigid, obedient character type that respects hierarchy and is more willing to follow orders. The examiner's favourite tension is exactly this: Milgram's situational factors versus the dispositional claim that some people are simply built to obey.
Here is a 4-mark application question and a model answer, annotated so you can see where the marks land.
Question: 'Ravi is a new soldier. His sergeant orders the unit to do something Ravi believes is wrong, but the whole unit obeys and so does he. Using your knowledge of obedience, explain Ravi's behaviour.' (4 marks)
Model answer: 'Ravi is likely to have entered an agentic state, meaning he no longer feels personally responsible for his actions and instead sees himself as carrying out the sergeant's orders [this is the AO1 knowledge — the named concept, defined]. Because the sergeant is a legitimate authority within a clear military hierarchy, and the order is given in a setting that reinforces that authority, Ravi is more likely to obey than if a stranger had asked [this is the AO2 application — the concept is tied to Ravi's specific situation: the sergeant, the hierarchy, the setting]. The fact that the rest of the unit also obeys removes the dissent that might otherwise have freed him to refuse, so situational pressure, rather than Ravi's character, best explains why he acted against his own judgement [this closes with a short evaluative point weighing situation against disposition].'
Notice the shape: name and define the concept, then bolt every sentence back onto Ravi by name. A common way to lose two of these marks is to describe the agentic state correctly but never mention Ravi — that is AO1 without the AO2 the question actually asked for.
MechanismProsocial behaviour — helping, and the bystander effect
Prosocial behaviour is acting to benefit someone else — and the puzzle psychologists chased is why we so often fail to. The trigger was the 1964 murder of Kitty Genovese in New York, reportedly witnessed by dozens of neighbours who did not intervene. That produced the idea of the bystander effect: the more people present, the less likely any one of them is to help, because of diffusion of responsibility — everyone assumes someone else will act, so no one does.
Piliavin's 1969 New York subway study tested what tips people into helping. An actor collapsed in a train carriage, appearing either ill (carrying a cane) or drunk (carrying a bottle). The ill victim was helped far more often and faster than the drunk one, because bystanders weigh the cost of helping against the cost of not helping — a drunk stranger feels riskier and less deserving. The factors AQA wants you to marshal are the cost of helping, the similarity to the victim (we help people we see as like us more readily), the number of bystanders (diffusion), and whether we feel competent to help. Interestingly, in Piliavin's face-to-face carriage, help rates were high — showing that when responsibility cannot easily be passed off, the bystander effect weakens.
CaseCrowd and collective behaviour — losing yourself in the many
In a crowd, people do things they would never do alone — for better and for worse. The key idea is deindividuation: within a large, anonymous group we lose our sense of individual identity and personal responsibility, and the normal restraints on behaviour loosen. It can turn anti-social — the 2011 England riots saw ordinary people looting shops they walked past every day, emboldened by anonymity and the sense that no one could be singled out — or intensely pro-social, as when a charity crowd or a rescue effort acts with a generosity individuals rarely show alone.
Several factors shape which way a crowd tips. Deindividuation and anonymity reduce accountability. Social loafing — the tendency to put in less effort when responsibility is shared — can leave a crowd passive when action is needed. Culture shapes the norms a crowd falls back on, and collective faith in others can either spread panic (contagion) or spread calm and cooperation. The exam point to hold is that a crowd is not simply 'a mob': the same psychological mechanism — reduced individual identity — can produce a riot or a relief effort, and the direction depends on the situation and the norms it activates.
VocabularyKey terms the mark scheme pays for
TrapsMisconceptions that cost marks
ExamWhat examiners want
Social influence questions in AQA GCSE Psychology (8182) span all three assessment objectives, and knowing which one a question wants is half the battle. AO1 asks you to describe the classic studies and concepts accurately — Asch on conformity, Milgram and the agentic state, Piliavin on bystanders, deindividuation in crowds — so revise the studies with their key figures (65%, roughly 75%, the factors) attached. AO2 gives you a fresh social scenario (a soldier, an employee, a witness) and expects you to apply the right concept to that person by name; describing the theory without landing it on the stem caps your marks. AO3 wants evaluation, and this topic has a rich, reusable currency: generalisability (Milgram tested 40 American men, so the findings may not extend to women or other cultures), ethics (deception and protection from harm in Milgram), and ecological validity (Asch's line task is artificial, whereas Piliavin's field study is more lifelike).
On extended-writing questions worth 6 or 9 marks, examiners use a levels-based scheme that rewards a clear line of reasoning, not a scattergun of facts: make a point, back it with evidence from a named study, evaluate it, and reach a conclusion that answers the exact wording. Balance is the trap — strong answers spend real time on evaluation rather than pouring everything into description, and every evaluative point is tied back to the question rather than left floating.