Social influence: key studies
10 studies. Tap a card to flip it, then say whether you knew it. Cards you miss come back every round; cards you know come back less often. Progress is saved in this browser. Terms deck · Match-up game · Back to the topic
Asch (1951)
Aim To see whether people conform to a majority on a task with an obvious right answer.
Method 123 American male students judged which of three comparison lines matched a standard line, in groups where the others were confederates who gave the same wrong answer on 12 of 18 trials.
Findings Participants conformed to the wrong answer on about 37% of critical trials; 75% conformed at least once; 25% never conformed. Many said afterwards they conformed to avoid disapproval.
Conclusion People conform to a clear majority even when they know it is wrong, mostly through normative social influence.
Evaluation Artificial task and setting (low ecological validity); all male American students in the 1950s, so the findings may be time- and culture-bound (Perrin and Spencer found far less conformity in 1980 with engineering students); deception and stress raise ethical issues.
Examiners say Reports warn against muddled statistics: 37% is the conformity rate on critical trials, 75% is the proportion who conformed at least once. Questions on Asch's variables want the direction of the effect, not just the name of the variable.
Asch's variables (1951–56)
Aim To find which situational factors change the level of conformity.
Method Variations of the line task altering group size, unanimity of the majority and task difficulty.
Findings Conformity rose with group size up to about three confederates then levelled off; a single dissenter (even one giving a different wrong answer) cut conformity to around 5%; making the lines more similar increased conformity.
Conclusion Conformity depends on the size and unanimity of the majority and on how uncertain the task makes people feel (informational social influence when the task is hard).
Evaluation Same limitations as the original study; the effect of group size may depend on the task; the findings allow practical applications such as social support in resisting pressure.
Examiners say 'Outline how two variables affected conformity' needs both the variable and what happened to conformity. Naming the variable alone scores nothing.
Zimbardo (1973), Stanford prison study
Aim To see whether brutality in prisons comes from the personalities of guards and prisoners or from the social roles they take on.
Method 24 emotionally stable male volunteers were randomly assigned to be guards or prisoners in a mock prison in the Stanford psychology department; prisoners were arrested at home, de-individuated with numbers and smocks, and the study was planned for two weeks.
Findings Guards became increasingly abusive; prisoners became passive and some showed severe distress. The study was stopped after six days.
Conclusion Situational factors and social roles have a powerful effect on behaviour; ordinary people conform to roles.
Evaluation Demand characteristics and play-acting (Banuazizi and Movahedi); only about a third of guards behaved brutally, so disposition mattered too; serious ethical problems, though Zimbardo did debrief; Reicher and Haslam's BBC replication found different outcomes.
Examiners say Credit for Zimbardo needs the roles and the process, not just 'it got out of hand'. Reports note students who evaluate only on ethics when the question says 'apart from ethical issues'.
Milgram (1963)
Aim To find out how far ordinary people would obey an authority figure who told them to harm another person.
Method 40 American men, recruited as a memory study, acted as 'teacher' and gave what they believed were increasing electric shocks (15–450 V) to a confederate 'learner' whenever he made a mistake, prompted by the experimenter's prods.
Findings All participants went to 300 V and 65% went to the full 450 V; many showed signs of extreme stress.
Conclusion Obedience to a legitimate authority is high in ordinary people; situational pressures override personal conscience.
Evaluation Low internal validity if participants did not believe the shocks were real (Orne and Holland; Perry's analysis of the tapes), though Sheridan and King's puppy study found real obedience; ethical issues of deception, protection from harm and right to withdraw; supported by Hofling's nurses in a field setting.
Examiners say Questions on Milgram's situational variables (proximity, location, uniform) want the figures: touch proximity 30%, run-down office 47.5%, member of the public in plain clothes giving orders 20%.
Milgram's variations (1974)
Aim To test how situational variables affect obedience.
Method Repeats of the shock procedure changing proximity (same room, touch proximity, experimenter by phone), location (run-down office rather than Yale) and uniform (experimenter replaced by a 'member of the public').
Findings Same room 40%; touch proximity 30%; experimenter absent 20.5%; run-down office 47.5%; ordinary clothes 20%.
Conclusion Obedience falls as the authority's legitimacy and the victim's distance are reduced; the agentic shift is harder to maintain when the consequences are close.
Evaluation Cross-cultural replications support the pattern (Meeus and Raaijmakers in the Netherlands); the variations may have made the deception more obvious; the research has been used to explain real atrocities, though Mandel argues this gives an 'obedience alibi'.
Examiners say Reports want the variation named, the percentage, and the explanation (e.g. a uniform signals legitimate authority).
Adorno et al (1950), authoritarian personality
Aim To find out whether some personalities are more obedient to authority.
Method Over 2,000 middle-class white Americans completed the F-scale measuring fascist tendencies, with interviews about childhood.
Findings High scorers identified with strong people, were contemptuous of the weak, held rigid conventional views and had often experienced harsh, disciplinarian parenting.
Conclusion A dispositional explanation: obedience is partly a personality type formed by a strict upbringing, with hostility displaced onto those seen as inferior.
Evaluation Correlational, so cause cannot be inferred; the F-scale suffers from acquiescence bias (agreeing with every item scores high); cannot explain whole populations obeying; Elms and Milgram found fully obedient participants scored higher on the F-scale, offering some support.
Examiners say A 2-mark limitation needs an explanation, not a label: 'the F-scale has response bias because every item is worded in the same direction, so agreeing with everything looks authoritarian'.
Moscovici et al (1969), minority influence
Aim To see whether a consistent minority can influence a majority.
Method Groups of six (four participants, two confederates) judged the colour of 36 blue slides. In one condition the confederates consistently said green; in another they said green on two-thirds of trials.
Findings The consistent minority led participants to say green on 8.4% of trials, with 32% agreeing at least once; the inconsistent minority had little effect (1.25%).
Conclusion Consistency is what gives a minority its influence, by making the majority reconsider.
Evaluation Artificial task; only female American participants; later work (Martin et al) shows minority views are more deeply processed and resistant to change, which supports the idea of conversion rather than compliance.
Examiners say Reports note confusion between consistency and commitment: commitment is demonstrated by sacrifice (the augmentation principle), consistency by repeating the same view over time and across the group.
Hofling et al (1966), nurses
Aim To test obedience to authority in a real hospital.
Method An unknown 'doctor' telephoned 22 nurses and asked them to give a patient twice the maximum dose of an unfamiliar drug, breaking several hospital rules.
Findings 21 of the 22 nurses obeyed.
Conclusion Obedience to a legitimate authority occurs in real settings, not just the laboratory.
Evaluation High ecological validity; Rank and Jacobson's replication with a familiar drug and time to consult found only 2 of 18 obeyed, so the finding may depend on the specific conditions.
Examiners say Useful as support for legitimacy of authority and for the external validity of Milgram's findings.
Rotter (1966), locus of control
Aim To measure whether people see events as controlled by themselves (internal) or by outside forces (external).
Method A forced-choice questionnaire placing people on a continuum from internal to external locus of control.
Findings Internals take responsibility for their actions and are more likely to resist pressure to conform or obey.
Conclusion Locus of control is a dispositional explanation of resistance to social influence.
Evaluation Holland's replication of Milgram found 37% of internals refused to continue against 23% of externals, supporting the link; Twenge et al found people have become more external but more resistant over time, which contradicts it; the effect may apply only to new situations.
Examiners say Application questions ask you to identify the internal from a stem (takes credit for success, blames self for failure) and then link that to resistance.
Gamson et al (1982)
Aim To see whether group support helps people resist an unjust authority.
Method Groups of participants were asked to produce video evidence for an oil company's smear campaign against a manager.
Findings 29 of 33 groups rebelled; discussion within groups made the request seem illegitimate.
Conclusion Social support makes resistance to authority much more likely.
Evaluation Supports social support as an explanation of resistance; the group setting differs from Milgram's isolated participants.
Examiners say Evidence for social support should name the mechanism: an ally breaks the unanimity and acts as a model for disobedience, even if the ally is later withdrawn (Asch: resistance did not persist once the dissenter left; Allen and Levine: a dissenter with poor eyesight still reduced conformity).
Work with an AQA psychology tutor
Know the studies but not scoring on the essays? That's what a tutor fixes.