Psychology Core Studies.


Background, Milgram (1963).

Obedience involves: being instructed or ordered to do something; being influenced by an authority figure of a superior status; or the maintenance of social power and status of the authority figure in a hierarchical society.

When given extreme commands by legitimate authority figures, subordinates adopt an agentic state where they become the instrument for carrying out another person’s wishes.

From 1933-45, millions of innocent people were systematically slaughtered on command. Such inhumane actions may have originated in the mind of one person, but they could only have been carried out on such a massive scale because large numbers of people obeyed. There was the hypothesis that this behaviour was unique to the German race, so Milgram decided to test it on Americans.

1 of 24

Research Method, Milgram (1963).

Although Milgram refers to this study as an experiment, it is generally considered a controlled observation as there was, in fact, no independent variable.

The aim of this study was to investigate the process of obedience by testing how far an individual will go in obeying an authority figure, even when the command breaches the moral code that an individual should not hurt another person against his will.

Prior to the study, 14 Yale Seniors, all psychology majors, estimated the percentage of participants who would administer the highest level of shock. Estimates ranged from 1-3%.

2 of 24

Sample, Milgram (1963).

40 male participants aged between 20 and 50 years, from the New Haven area were obtained by a newspaper advertisement and direct mail solicitation which asked for volunteers to participate in a study of memory and learning at Yale University.

There was a wide range of occupations in the sample.

Participants were paid $4.50 for simply presenting themselves at the laboratory.

The experimenter was played by a biology teacher and the victim was an accountant called Mr Wallace.

3 of 24

Procedure, Milgram (1963).

The study took place in a laboratory at Yale University where participants were told that the study was about how punishment effected learning.

The ‘teacher’ then sat in front of an electric shock generator in an adjacent room. He had to conduct a paired word test on the learner and give him an electric shock of increasing intensity for every wrong answer. The machine had 30 switches ranging from 15-450 volts, in 15 volt increments. The ‘learner’ produced (via a tape recording) a set of predetermined responses, giving approximately 3 wrong answers to every correct one. At 300 volts he pounded on the wall and thereafter made no further replies. At 315 volts the learner pounded on the wall again and then fell silent.

If the ‘teacher’ turned to the experimenter for advice on whether to proceed, the experimenter responded with a series of standardised prods eg’ “Please continue / Please go on.”. If the participant asked if the learner was in pain he would respond ‘Although the shocks may be painful, there is no permanent tissue damage’. The study finished when either the ‘teacher’ refused to continue (was disobedient) or reached 450 volts (was obedient). The participant was then fully debriefed. Data was gathered through observations made by both the experimenter who was in the same room as the participant and others who observed the process through one-way mirrors.

4 of 24

Findings, Milgram (1963).

Quantitative data: All participants (40/40) / 100% continued to 300 volts.

26/40 / 65% of participants continued to the full 450 volts. 14 were disobedient and walked out.

Qualitative data: Many participants showed signs of extreme stress whilst administering the shocks e.g. sweating, trembling, stuttering, laughing nervously.

3 had full blown uncontrollable seizures.

Milgram offered 13 possible explanations for the high levels of obedience shown by participants e.g. The fact that the study was carried out in the prestigious university of Yale influenced participants as to the worthiness of the study and the competence of the researcher; the participants were told the shocks were not harmful; the situation was completely new for the participant so he had no past experience to guide his behaviour.

In a follow up study 84% of the participants felt glad they had participated and only 1 said that he he felt sorry he had participated.

5 of 24

Conclusions, Milgram (1963).

Inhumane acts can be done by ordinary people due to being in agentic state

People will obey others whom they consider legitimate authority figures even if what they are asked to do goes against their moral beliefs.

People obey because certain situational features lead them to suspend their sense of autonomy and become an agent of an authority figure.

Individual differences, such as personality, influence the extent to which people will be obedient.

6 of 24

Evaluation Strengths, Milgram (1963).

A strength of Milgram’s study was the use of both quantitative and qualitative data. This is a strength because it means we can make more valid conclusions about obedience. Solely looking at voltage we may infer that people are uncaring. But by also looking at qualitative behaviour we can conclude that not only do people obey, but they found the experience highly stressful.

The measure was highly reliable/consistent. This is known as internal reliability i.e. consistency of the procedures or materials.

Has external reliability i.e. when the results are found to be the same when repeated. Milgram’s study has been replicated many times and found very similar obedience levels.

Understanding why people obey destructive orders can be used as a preventative measure and therefore has useful applications.

The features of the study are highly comparable to the Nazi officers e.g. respectable environment with uniform to signify authority, participants were told that the study would enhance science, the electric shocks increased in small increments.

7 of 24

Evaluation Weaknesses, Milgram (1963).

40 male participants from the USA, the findings on obedience levels could be difficult to generalise to the whole population.

The use of self-selected and snowball sampling increases the likelihood of an unrepresentative sample. It will only attract a certain type of helpful person.

The use of a lab for the obedience task reduces ecological validity, however, to many of the participants the experience appeared very real (realism) due to the fact that they were deceived.

Extremely unethical, participants were deceived about the true aims of the study and were at risk of psychological harm due to believing they had possibly killed someone.

8 of 24

Background, Bocchiaro (2012).

Social power refers to the influence an individual has to change another's thoughts, feelings or behaviours. Individuals in authority, be it legitimate or illegitimate, have social power to influence those with lower social status within their social hierarchy.

A whistle blower is a person who exposes/informs on a person or organisation regarded as engaging in unlawful or immoral activity.

In most situations, when someone disobeys, it is assumed that there would be a lower level of whistle-blowing than disobedience because it involves a potential direct confrontation of the person and the authority.

Milgram found that people have strong inclinations to obey legitimate authority, irrespective of their beliefs, feelings or intentions.

9 of 24

Research Method, Bocchiaro (2012).


(a) Who are the people that disobey or blow the whistle?

(b) Why do they choose the challenging moral path?

(c) Do they have personal characteristics that differentiate them from those who obey?

Like Milgram, there was in fact no independent variable (IV) so the study may be best viewed as a laboratory study, or as Bocchiaro et al say a ‘scenario study’.

The study took place in a laboratory (standardised and controlled procedure) at the VU University in Amsterdam.

10 of 24

Sample, Bocchiaro (2012).

149 undergraduate students took part in the research in exchange for either €7 or course credit. NB.

A total of 11 participants were removed from the initial sample of 160 because of their suspiciousness about the nature of the study.

A comparison group of 138 students were used who predicted obedience, disobedience and whistle blowing, but did not participate in the actual study.

11 of 24

Procedure, Bocchiaro (2012).

8 pilot tests, involving 92 undergraduates from the VU University in Amsterdam, were conducted to ensure the procedure was credible and morally acceptable. These tests also served to standardise the experimenter-authority behaviour throughout the experimental period

Each participant was greeted in the laboratory by a male, Dutch experimenter who was formally dressed and stern. The experimenter proceeded with a (seemingly unjustified) request for each participant to provide a few names of fellow students and then presented the cover story about sensory deprivation. The experimenter then left the room for 3 minutes for thinking time.

Participants were then moved to a second room where there was a computer for them to use to write their statement. Participants were told to be enthusiastic when writing their statements and had to use two adjectives among “exciting”, “incredible”, “great” and “superb”. If a participant believed the proposed research on sensory deprivation violated ethical norms he/she could anonymously challenge it by putting a form in the mailbox. The experimenter told participants to begin and left the room for 7 minutes.

After the 7-minute interval the experimenter returned and invited the participant to follow him back to the first room where he/she was administered two personality tests, fully debriefed and asked to sign a second consent form, this time fully informed. Sessions were approx 40 mins.

12 of 24

Findings, Bocchiaro (2012).

Of the 138 comparison students e.g. prediction group: only around 4% indicated they would obey the experimenter. Most believed they would be either disobedient, around 32% or whistleblowers, around 64%

Of the 149 participants in the actual laboratory situation: – 114 obeyed the experimenter, 21 disobeyed and 14 blew the whistle.

Results for individual differences found that the only significant difference that was found was in relation to faith, with results suggesting that people with higher faith were more likely to be whistle-blowers.

13 of 24

Conclusions, Bocchiaro (2012).

People tend to obey authority figures, even if the authority is unjust.

Behaving in a moral manner appears to be challenging for people, even when the reaction appears to observers as the simplest path to follow.

Behavioural acts of both disobedience and whistleblowing are psychologically, socially and economically demanding for people, notably whistleblowers.

People are not very good at predicting what they or others will do.

14 of 24

Evaluation Strengths, Bocchiaro (2012).

The procedure takes place in a highly controlled environment and therefore it is possible to control for extraneous variables and be reasonably confident that the authority figure effected the participants obedience levels. It is also highly standardised and therefore easy to replicate.

The procedure was very lifelike because the situation investigated was a psychologist carrying out a study, and this is exactly what happened, there was nothing artificial about the procedure therefore increasing the ecological validity.

Large sample of 149 participants, this reduces the chance of participant extraneous variables.

The fact that religion was considered to play a role in disobedience meant that some cultural influences were considered, making the study less ethnocentric.

Practical applications are important: because of the publicity of whistle blowing, more and more people are coming forward and standing up to unethical treatments and procedures e.g, health care, Jimmy Savile case.

Only collecting quantitative data allowed direct comparisons of obedience, disobedience and whistle blowing.

15 of 24

Evaluation Weaknesses, Bocchiaro (2012).

Ethical issues can be raised, people were put into a situation facing a moral dilemma and this can cause potential stress. Also participants were deceived in order to avoid demand characteristics and achieve high validity.

Using students from a Dutch university meant that the sample was unrepresentative of the general population and other age and cultural groups.

16 of 24

Background, Piliavin (1969).

Diffusion of responsibility is where the responsibility for the situation is spread (diffused) among the people present. This implies that the more people present, the more the bystander believes the responsibility is spread out so they feel less personally responsibility and are therefore less likely to help.

Another explanation for not helping a victim in need is that a bystander may believe that someone else will do what’s necessary so there is no need for them to offer assistance. This is known as ‘bystander apathy’.

Since the murder of Kitty Genovese in 1964 (a woman stabbed to death over a period of 30 minutes in front of a reported 38 unresponsive witnesses), many social psychologists believed that this was an example of diffusion of responsibility.

Research by Darley and Latané (1968) found that bystanders hearing an epileptic fit over earphones, led to those who believed other witnesses to be present being less likely to help the victim than bystanders who believed they were alone.

17 of 24

Research Method, Piliavin (1969).

The study was a field experiment on a New York subway. The journeys lasted about 7 1⁄2 mins.

The experiment had four independent variables (IVs):

  • Type of victim (drunk or carrying a cane).
  • Race of victim (black or white).
  • Effect of a model (after early – 70 or late – 150 seconds, from the critical or adjacent area), or no model at all.
  • Size of the witnessing group (a naturally occurring independent variable).

The dependent variables (DVs) recorded by two female observers in the adjacent area were:

  • Number of helpers.
  • Speed of help.
  • Race of helper.
  • Sex of helper.
  • Movement out of critical area.
  • Verbal comments by bystanders.
  • Verbal comments by helpers.
18 of 24

Sample, Piliavin (1969).

Participants were about 4,500 men and women who used the New York subway on weekdays between 11.00 am and 3.00 pm.

19 of 24

Procedure, Piliavin (1969).

There were 4 teams of 4 researchers: 2 female observers, 2 males – one acting as victim, one the model

The victims (3 white, 1 black) were all male, General Studies students, aged 26-35 years, and dressed alike. They either smelled of liquor and carried a liquor bottle wrapped tightly in a brown bag or appeared sober and carried a black cane. In all aspects they acted identically in both conditions.

The victim stood near a pole in the critical area. After about 70 seconds he staggered forward and collapsed. Until receiving help he remained on the floor looking at the ceiling. If no help was offered, the role model would step in and help after either 70 seconds or 150 seconds. The point of this was to see if this affected helping behaviour. If no one helped up until the train stopped the model would help the victim up.

After each trial, they then changed platforms to repeat the process in the opposite direction.

The observers recorded the dependent variables as stated above. Both observers recorded comments spontaneously made by nearby passengers and helpers.

20 of 24

Findings, Piliavin (1969).

The cane victim received spontaneous help 95% of the time compared to the drunk victim 50% of the time

Help was offered more quickly to the cane victim (a median of 5 seconds compared to 109 seconds delay for the drunk victim).

90% of the first helpers were males.

There was a slight tendency for same race helping especially in the drunk condition.

No diffusion of responsibility was found, in fact response times were faster with larger groups than smaller, and the more passengers that were near the victim the more likely that help was given.

More comments were made by passengers in the drunk than the cane condition and most comments were made when no help was given within the first 70 seconds.

21 of 24

Conclusions, Piliavin (1969).

An individual who appears ill is more likely to receive help than someone who seems drunk. With mixed groups of men and women, men are more likely than women to help a male victim. People are more likely to help people of the same race especially if they are drunk.

People may or may not help because: An emergency situation creates a sense of empathy in a bystander. This empathy is increased if someone feels a sense of identity to the victim, or if they are physically close to them. The emotions felt can be reduced by directly helping or rationalising reasons to not help.

Helping is also determined by a cost-reward model. If the possible cost of helping e.g. risk to self are greater than the reward e.g. praise, feeling good about yourself, then someone is less likely to help.

22 of 24

Evaluation Strengths, Piliavin (1969).

A main strength of the study has to be its high level of ecological validity. The study was done in a true to life environment and consisted of an incident, which could and does happen. However, some of the participants were very close to the victim and were in a situation where they could not escape. This is often unlike many other situations where we come across emergency situations and this may be one of the reasons why diffusion of responsibility did not occur.

The sample size was also very large and we would assume a fairly representative sample of New Yorkers. The researchers should therefore be able to generalise their findings with much more certainty than if they had carried out a study on say 40 students. 

23 of 24

Evaluation Weaknesses, Piliavin (1969).

The study can be criticised on ethical grounds. A problem with the field experiment is that the participants cannot give their consent, because they do not know that they are participants in an experiment. Similarly the participants are being deceived because they are unaware that it is not a genuine emergency. Participants were also not debriefed as this would have been almost impossible. Following from this it is possible that participants had feelings of guilt, distress, and anxiety.

A further problem with field experiments is that they are more difficult to control than laboratory experiments. For example we could question whether travellers on the trains saw more than one trial. Field experiments are also more difficult to replicate and more time consuming and expensive.

24 of 24


No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Core studies resources »