Aim: would people conform to a unambiguous task?
Method: lab, independent groups, male undergrads. Say which line matched the standard line. 1 PP per group of 8.12/18 trials - confederates gave same wrong answer.
Findings: 6 control trials: gave the wrong answer 0.7% of the time. 12 trials - 37%. 75% of PPs gave the wrong answer at least once. Task became more difficult - more people conformed. Non unanimous majority reduced conformity. Being dissented - 28% conformity. Majority of 2 - not enough for people to conform.
Conclusion: control trials showed the right answer was clear. People will conform to things which they know to be wrong (NSI).
Evaluation: high int. validity (control, easy to replicate), low ext. validity (sample bias, zeitgeist - more conformity/stricter schools in the 50s), time consuming/expensive. Ethics (deception, ends/means?). Lacks mundane realism (no significance, conformity depends on the subject).
Aim: would people conform in an ambiguous task?
Method: lab, repeated measures. Auto kinetic effect. PPs had to guess how much the light moved (first individually, then in front of 2 others, then individually again).
Findings: individual estimates varied widely. Estimates converged in the group. On their own again, estimates were more like the group estimates.
Conclusion: a group norm developed (influenced by others' estimates), affected by ISI.
Evaluation: high int. validity (no PP variables, easy to replicate, high control), low ext. validity (sample bias), lacks mundane realism, ethical issues (deception).
Aim: how people are affected by minority influence.
Method: lab, 132 women. Groups of 6 with 4 true PPs. Had to judge the colour of 36 slides (all shades of blue). Consistent condition: confederates said slides were all green. Inconsistent condition: said 24 were green and 12 were blue. Control: no confederates.
Findings: control group said green 0.25% of the time. Consistent: 8% said all green, 32% said green at least once. Inconsistent said green 1.25% of the time.
Conclusion: minority influence is stronger when it's consistent.
Evaluation: lacks mundane realism (trivial), control group shows hte task was unambiguous, low ext. validity (sample bias).
Aim: why people obey unjust orders.
Method: 40 male volunteers. Learner - stooge. Shock generator (15v increase). The teacher adminstered increasingly powerful shocks. Experimenter prods, e.g. "it is absolutely essential that you continue."
Findings: 65% went to 450v. All PPs went to 300v, 12.% stopped there. The others stopped when the leaner protested. Seedy office: 47.5% went to 450v. The teacher forced the learner's hand onto the plate - 30% went to 450v.
Conclusion: people are obedient to authority, even in inhumane ways. Crimes against nature possibly caused by situational factors.
Evaluation: ethics (deception, informed consent, right to withdraw, protection from harm, ends/means - personal importance), important research (Nazi Germany), lacks mundane realism, Orne and Holland suggested demand characteristics (PPs showed distress).
Schwarzwald, Bizman and Raz
Aim: investigate the foot-in-the-door effect.
Method: independent groups. Gradual commitment condition: asked to sign a petition + give a donation 2 weeks later. Control group: asked to give a donation, no petition.
Findings: gradual commitment condition - more people made donations and the donations were larger.
Conclusion: the foot-in-the-door effect is effective.
Evaluation: low int. validity (individual differences), mixed research (Bell found it ineffective).
Aim: investigate the role of authority figures.
Method: New York. Dressed up as a guard, milkman + normal citizen. Asked people to do things like pick up litter.
Findings: 90% obeyed guard, 50% obeyed citizen.
Conclusion: people are more likely to obey a perceived authority figure.
Evaluation: high mundane realism, ethics (deception, informed consent), low int. validity (sample bias).
Gamson et al (1982)
Aim: investigate whether people are more likely to obey than conform.
Method: PPs were asked to sign a document agreeing to the unfair dismissal of an employee, which would be used in court.
Findings: PPs disobeyed through minority influence (conformity). 1 or 2 PPs at first, then it spread throughout the group. 16/33 groups completely disobeyed, 9/13 partly disobeyed.
Aim: investigate individual differences in defying authority.
Method: American teenagers had to find a solution to a moral dilemma.
Findings: found 6 stages of moral reasoning. Those who reached the higher stages were more likely to disobey unjust demands.
Evaluation: Kohlberg interviewd PPs - those who obeted didn't consider the morals (those who resisted did), Milgram found education + religion affected obedience (less educated - more likely to obey, Catholics more likely to obey than Protestants).
Aim: investigate locus of control.
Method: Austria, similar to Milgram. Pressed a button releasing a painful stimulus to a learner everytime they got a question wrong. 20 switches, increasing in intensity. Measured their locus of control.
Findings: 80% pressed all buttons.
Conclusion: locus of control does not influence obedience.
Evaluation: Twenge et al: young Americans are becoming increasingly external (negative efects like depression). Caused by social changes since the 60s (e.g. increased divorce/crime rate) - people believe many aspects of their lives are outside their control.
Aim: more realistic study of obedience.
Method: naturalistic, field, 22 nurses. Stooge doctor phoned the hospital - asked them to give 20mg of a drug to a patient without authorisation (3 rules broken: over the phone, double max dosage of 10mg, unauthorised).
Findings: 95% obeyed. Another 22 nurses were asked what they'd do - said they wouldn't obey.
Conclusion: people are unwilling to question authority even when they should.
Evaluation: unethical (deception), high mundane realism, low ext. validity (sample bias).
Found deindividuation allowed people to be more obedient. Stanford Prison Experiment: guards became aggressive when they wore sunglasses/uniform - they stopped taking responsibility, changed their behaviour to fit the role.
Aim: investigate deindividuation in obedience.
Method: replicated Milgram. Some treated as individuals (wore own clothes), some treated as a group (wore hoodies).
Findings: average level of shock reached doubled when PPs wore hoods.
Conclusion: obedience is more likely when people aren't treated as individuals.
Aim: investigate deindividuation in crowds.
Method: newspaper coverage of suicide attempts, focusing on crowds below buildings.
Findings: people in large crowds start telling people to jump (esp. if dark).
Conclusion: anonymity leads to more extreme behaviour (responsibility shifted to the group).