Asch's Line Study (1951)
Method: Lab experiment with independent groups design. Groups of 8 participants shown a set of lines and had to say which lines were the same. Each group contained one real participant, other seven were confederates. Real participant was always last or second last to answer, hearing peers' answers first. Each participant did 18 trials and on 12 of these (the critical trials) the confederates all gave the same wrong answer. Also a control group, where participants made guesses in isolation.
Results: Wrong answer given 0.7% of the time in control group. This number rose drastically in the critical trials, with participants conforming to give the wrong answer 37% of the time. 75% of participants conformed at least once. Afterwards, some participants said they knew their answer was wrong, but didn't want to look different.
Conclusion: Normative social influence can affect results in an unambiguous task.
Evaluation: The study was a lab experiment, meaning there was good control of variables and the effect of extraneous variables was minimised. Strict control of the variables means the study could be repeated and the same results are produced.
The study lacks ecological validity as it did not occur in a natural situation. Participants were deceived and may have been embarrassed when they found out the true nature of the study.
Sherif (1935) - conformity and autokinetic effect
Method: Lab experiment with repeated measures design. Visual illusion called the autokinetic effect made a stationary spot of light appear to move. Participants had to estimate how far they thought it had moved. In the first phase participants made repeated estimates. Then they were put into groups of 3 and made estimates with the others present. They were then retested individually.
Results: When alone, participants made their own stable estimates (personal norms), which varied between participants. Once the participants were in a group, their estimates tended to converge. When participants were retested individually, their estimates were closer to the group estimates.
Conclusion: Participants were influenced by the others' estimates, and a group norm developed. Estimates converged because participants used info from others to help them- they were affected by informational social influence.
Evaluation: Lab experiment- controlled results should be possible to recreate and establish cause and effect. Repeated measures design meant participant variables that may have affected results were kept constant.
Experiment created an artificial situation, so the study lacks ecological validity. Limited sample - male participants, so results can't be generalised to everyone. Ethical issue of deception- participants were told the light was moving when it wasn't.
Zimbardo et al (1973) - Stanford Prison
Method: Male students recruited as either guards of prisoners in mock prison. Behaviours observed. Participants were 'arrested' at home, taken to 'prison' and given uniforms and numbers. Guards wore uniforms and mirrored sunglasses.
Results: Initially, guards tried to assert authority and prisoners tried to resist by sticking together. Prisoners became more passive and obedient, while the guards invented nastier punishments. The experiment was abandoned early because some prisoners became very distressed.
Conclusion: Guards and prisoners adapted their social roles quickly. Zimbardo claims this shows that our social role can influence our behaviour - seemingly well-balanced men became unpleasant and aggressive in the role of guard.
Evaluation: Controlled observation, so good control of variables.
As the study took place in an artificial environment, the results can't really be generalised to real-life situations.
In terms of ethics, some participants found the experience very distressing.
There's an issue with observer bias, as Zimbardo ran the prison himself and later admitted he became too personally involved in the situation.
The experiment doesn't take individual differences into account.
Moscovici et al (1969)
Method: Lab experiment into minority influence using 192 women. Groups of 6 at a time, participants judged the colour of 36 slides. All slides were blue, but the brightness varied. Two of the six participants in each group were confederates. In one condition the confederates called all 36 slides 'green' (consistent) and in another condition they called 24 of the slides green and 12 of the sides 'blue (inconsistent). A control group was also used which contained no confederates.
Results: In the control group the participants called the slides 'green' 0.25% of the time. In the consistent condition 8.4% of the time participants adopted the minority position and called the slides 'green'. 32% of participants called the slides 'green' at least once. In the inconsistent condition the participants called the slides green only 1.25% of the time.
Conclusion: The confederates were in the minority but their views appear to have influenced the real participants. The use of the two conditions illustrated that the minority had more influence when they were consistent in calling the slides 'green'.
Evaluation: The study was a lab experiment, so it lacks ecological validity because the task was artificial. The study only included females and so can't be generalised to men. The use of the control group makes it certain that the participants were actually influenced by the minority. In a similar experiment, participants were asked to write down the colour rather than saying it out loud. In this condition, even more people agreed with the minority, which provides more support for minority influence.
Milgram (1963) - Obedience to authority
Method: Lab experiment. 40 male participants volunteered by responding to newspaper advert. Placed in room with experimenter wearing grey technician's coat. Participants introduced to a confederate. Study fixed so participant was always 'teacher'. They saw confederate strapped into a chair and connected to shock generator in next room. It didn't actually give shocks, participants deceived to believe it did. Switches ranged from 15 volts to 450 volts. When learner answered incorrectly, participant gave increasing level of shock. As shocks increased, learner began to scream and ask to be let out. After 330 volt shock, they made no further noise. If participants hesitated, the experimenter told them to continue. Debriefing included interview, questionnaires and 'reunion' with the learner.
Results: 26 participants (65%) administered 450 volts and none stopped before 300 volts. Most showed signs of stress during the experiment, like sweating, groaning and trembling.
Conclusion: Ordinary people will obey orders to hurt someone else, even if it means acting against their consciences.
- Internal validity: Possible that participants didn't really believe they were inflicting electric shocks - they might have just been going the along with the experimenter's expectations (showing demand characteristics)
- Ecological valditiy: Participants did a task they were unlikely to encounter in real life.
- Ethics: Participants were deceived of the true nature of the study and weren't informed of their right to withdraw.
Milgram (1963) Continued
Factors that affected obedience
- Presence of allies: When there were 3 teachers (1 participant and 2 confederates), the real participant was less likely to obey if the other two refused to obey. Having allies can make it easier to resist orders than when you're on your own.
- Proximity of the victim: Milgram's results suggest an important factor was the proximity (closeness) of the learner. In the 'remote learner' condition, 65% gave the maximum shock. This dropped to 40% with the learner in the same room, and 30% when the participant had to put the learner's hand on the shock plate. Proximity made the learner's suffering harder to ignore.
- Proximity of the authority: When the authority figure gave prompts by phone from another room, obedience rates dropped to 23%. When the authority figure wasn't close by, orders were easier to resist.
Milgram's Agency Theory (1973)
1) When people behave on behalf of an external authority (do as they're told), they're said to be in an agentic state.
2) This means they act as someone's agent, rather than taking personal responsibility for their actions.
3) The opposite of this is behaving autonomously - not following orders.
4) Milgram claimed that there were some binding factors that might have kept participants in the agentic state:
- Reluctance to disrupt the experiment - participants had already been paid, so may have felt obliged to continue.
- The pressure of the surroundings - took place in prestigious university. Made experimenter seem like a legitimate authority.
- The insistence of the authority figure - participants were told they had to continue when hesitant.
- Experimental evidence to support agency theory - Milgram's participants often claimed they wouldn't have gone as far by themselves, but were just following orders.
- Sometimes people resist the pressure to obey authority. This can be because of the situation, or because of individual differences. Agency theory doesn't explain why some people are more likely to exhibit independent behaviour than others.
Zimbardo (1970) - Deindividuation
Method: Replicated Milgram's experiment and examined the effect of different conditions. Compared participants who wore their own clothes and were treated as individuals, to ones who wore hoods covering their faces and were spoken to as a group.
Results: Found that the average level of electrick shock doubled when participants were wearing a hood.
Conclusion: When participants were deindividuated, they became more obedient and antisocial. This was replicated in Zimbardo's prison study in 1973, when the prison guards wore uniforms and sunglasses. It seems they stopped taking personal responsibility for their actions, and changed their behaviour to fit into their social role.
Mann (1981) Deindividuation in Large Crowds
Method: Looking at newspaper coverage of suicide attempts. Focused on crowds that gathered below when someone was threatening to jump off a tall building or bridge.
Results: The newspaper reports showed that people in large crowds were likely to start jeering and telling the person to jump. This was even more common when it was dark.
Conclusion: The anonymity you get in a big group can lead to more extreme behaviour, because the sense of personal responsibility is shifted onto the group.