Obedience AS Psychology: Social Influence


Possible Questions
Home  AS      A2  Links


This area of the course on social influence, covers one of the most interesting and controversial areas in Psychology.  Hence the critical issue is the ethics of using human participants in Psychological research.


First a distinction:

Differences between conformity and obedience:





What is it?

Going along with the crowd/yielding to group pressure

Behaving as instructed!

Who ‘asks’?

Nobody, we act to please peers, friends, social group

Authority figures: parents, teachers (no don’t laugh!), police, government etc.

Why do we do it?

To be accepted, liked or just to fit in or to avoid feeling silly

To avoid punishment or unpleasant consequences




1. Conformity and Minority influence

It’s important to remember at the outset that although psychological research generally gets conformity a bad name without it Society would not function.  In the majority of real life situations conformity is seen as good!

Informational Social Influence

This happens when there is no obvious right answer so we look to others for information in order to be right.

Real life examples:

Looking at the people around you in a posh restaurant to see what knife and fork to use

Putting on car lights in the evening when others start to do the same.

Examples of Psychological research:


Jenness (1932): ‘Beans.’

You all did the experiment so you should be okay with the procedure. 

Sherif (1935): ‘Autokinetic effect’

Participants sit in a darkened room and stare at a pinpoint of light that appears to move, (try it sometime).  They are asked to estimate the distance it moves.  Since the movement is only apparent the correct answer is it doesn’t, but Sherif’s participants were obviously not aware of this.  Again, when put in rooms with others their guesses converge towards a group norm. 

In a follow up experiment Sherif started the participants in groups were they agree on an approximate answer.  When individuals are taken from this group and do the experiment on their own they stick to the answer agreed earlier.


Findings of this sort of research

Clearly there is conformity when people are unsure of the answer since group norms emerge.


Both studies are very artificial so lack ecological validity.  Can we generalise from this to real life situations?

The Sherif study does involve deception since participants are told the light is moving when it isn’t, so there are ethical concerns.

Note: The everyday examples are for clarification.  Avoid using them as examples in your answers.


Normative Social Influence

This happens when we go along with the crowd because we want to be accepted or liked or because we want to avoid embarrassment or being ridiculed.

Real life examples:

Smoking because others in your peer group smoke

Dressing like your friends in order to fit in or avoid bullying

Having a conservatory built because the neighbours have


Examples of Psychological research


Asch (1951 etc.): ‘The lines’

Again, you are all aware of the procedure.  Briefly stated: participants are deceived into taking part in a study on visual perception.  They are seated at a desk with others that they believe to be fellow participants but who in reality are in league with the researchers (stooges or confederates).  Lines are presented on a screen and participants simply have to say which line (out of 3 possibilities, is the same length as the target line).  The stooges get the right answer on the first two trials but then start to make deliberate mistakes. 

Conformity is measured by counting the number of times the real participant conforms when stooges give the wrong answer.

Mind Changers: Solomon Asch

Possible questions:

Describe the procedure.’  Easy peasie, describe the experiment as above.  You could mention some of the variations.  You could also mention the pilot study that Asch carried out first in which errors were only made on 3 trials out of 720.

Describe the findings.’   This one is more likely and also more troublesome.  What you must avoid doing is wasting time by describing the procedure. 

To answer this one, first of all mention Asch’s initial findings:

Overall conformity rate was 32% (unless your surname is Eysenck or Flanagan, in which case it’s 37%).  This means that participants conformed on 32% of all trials.

However, within this there were substantial individual differences:

Nobody conformed on 100% of trials

13 out of the original 50 never conformed at all

Highest rate of conformity was a participant who conformed on 11 out of 12 trials (must have felt a right plonker when he was debriefed!).

75% conformed at least once.

Also mention what Asch found in his variations:


Description and conformity

Size of group

One stooge (3%), two stooges (14%), three stooges (32%).  Further increases in group size do not increase conformity.  With very large groups conformity actually begins to fall!


If one of stooges also disagrees with others conformity drops sharply

Difficulty of task

As task becomes more difficult conformity increases

Familiarity of task

We are less likely to conform when we are confident in our ability, e.g. men are less likely to conform to incorrectly named tools than they are to incorrectly named kitchen utensils.  Clearly research of the 1950s!


Note: write your answers in continuous prose, not in tables like this.

Evaluation of Asch’s Paradigm (as it is often called)

The method

The procedure is very artificial (it lacks ecological validity) in that participants are being asked to conform when there is clearly a different and obviously correct answer.  In everyday life disagreements occur over politics, religion, tastes etc., when correct answers are not obvious, except we all agree that Kylie is lush!

Results do not appear to be consistent over time.  Later studies such as Perrin and Spencer’s in Britain in the 1980s found much lower levels of conformity.  It has been suggested that Asch’s original was post war when America was very wary of Communist take over when US citizens were worried about being seen to be different for fear of incrimination.  Levels of conformity did fall in the late 60s when it was popular for students in particular to protest against the Vietnam War, showing low levels of conformity.

The study is androcentric.  Only male participants took part.


The ethics

Participants were deceived so were unable to give their informed consent.  Note: whenever stooges are used there is always deception.

Participants were clearly stressed and some must have been embarrassed by the procedure and suffered some loss of self esteem once they had been informed that it had all been a big con.  This all constitutes ‘psychological harm.’


Crutchfield (1955): ‘The question booth.’

Crutchfield thought Asch’s experiment was far too expensive, time consuming and inefficient.  Lots of stooges were required to test each participant.  So he devised a method of testing lots of participants quickly and cheaply.  They were sat in cubicles and questions projected onto a screen.  In one corner were the answers given by other participants.  In fact these were made up and often wrong.  Conformity was measured by the number of times participants would go along with these incorrect answers. 

Example of question used: ‘The life expectancy of the average US male is 25.’

Participants answer true or false.  Since the screen indicates that the majority have answered ‘true’ many of the real participants do the same.  In fact Crutchfield found about the same level of conformity as Asch; 30%.

Also worth mentioning in an appropriate part c. question is the information Crutchfield found out about the personalities of conformist individuals by administering a personality test after the procedure.  According to this, conformist people tend to be: ‘intellectually less effective’, submissive, inhibited, have feelings of inferiority and have less mature social relationships.


Note: one of the questions asked by Crutchfield was; ‘true’ or ‘false’, ‘U.S males sleep 4 to 5 hours a night and eat 6 meals a day.’  Now that one I could believe!


Identification (or conforming to social roles and expectations)

This happens because we learn expectations of how we should behave in certain situations and then conform to these expectations when that situation arises.

Real life example:

Feeling very full of ourselves when dressed in evening wear such as a Tuxedo.  Sorry this is the best I can do!

Example of psychological research

Zimbardo’s Stanford prison simulation (1973)

Won't insult your intelligence here!

Again, in the unlikely event that the question asks for a description of the study, assume its party time but try to stick to the key details such as the way the guards were empowered by their dress (khaki uniform, dark glasses etc.), and the way the prisoners were humiliated by being strip searched.

In the more likely case of the question asking for findings:

Mention the effects on the prisoners who showed signs of ‘Pathological prisoner syndrome’ in which disbelief was followed by an attempt at rebellion and then by very negative emotions and behaviours such as apathy and excessive obedience.  Many showed signs of depression such as crying and some had fits of rage.  Zimbardo put these effects down to depersonalisation or deindividuation due to loss of personal identity and lack of control.

Mention also the effects on the guards who conversely showed the ‘Pathology of power.’  They clearly enjoyed their role; some even worked unpaid overtime and were disappointed when the experiment was stopped.  Many abused their power refusing prisoner’s food and toilet visits, removing their bedding etc.  Punishment was handed out with little justification.  Most notable was the way in which the ‘good guards’ never questioned the actions of the ‘bad guards.’ 



The experiment was a role play so it lacks realism with participants behaving as they think they should behave.  However, there is evidence for the guards not just simply role playing, for example their brutal behaviour wasn’t there at the start but developed over the first few days and they did not play up to the cameras as might be expected.  In fact their behaviour was worse when they knew they weren’t being observed.


  • Consent was obtained in advance and participants were told the nature of the research!
  • But, participants were not told that they would be arrested by real police officers and strip searched. 
  • Right to withdraw at best appears dubious. 
  • Although Zimbardo claims they were free to leave, and indeed some did, word got round the prisoners that this was not he case.
  • Participants were clearly subjected to physical and psychological harm. 
  • There is still a debate as to whether the experiment should have been stopped sooner, which brings into question Zimbardo’s dual role as researcher and self appointed ‘prison governor.’

However, in defence of Zimbardo you can mention the therapeutic debrief given to all those who took part.


Minority Influence

So far in all of the studies considered such as Asch etc., a majority have had influence over a minority, such as six stooges influencing one participant.  However, in real life if this were always the case, and the minority always went along with the majority, there would be no change in Society.  For change in ideas, religions, politics etc. there are times when a minority of people with different views have to exert their influence on the rest of us.  This so called minority influence tends to be a slow process, but it does bring about a change both in public and privately held opinions.

Real life examples:

The suffragette movement changing attitudes towards women’s rights, Galileo’s ideas on planetary movements, the Nazi’s reign in Germany etc…


Psychological experiments:

Moscovici et al (1969): ‘calling a blue slide green’

I can’t emphasise enough how important it is to remember this study, ‘cos ‘minority influence’ is a likely question and this is the only study to use!


Groups of six participants are presented with blue slides varying in intensity.  Participants have to say what colour the slides are.  Two of the participants are stooges and these answer in one of two ways:

  1. They always say the slides are green
  2. They say the slides are green on two thirds of occasions.


  1. When the stooges say ‘green’ every time: 8% of the majority agree
  2. When the stooges are less consistent this falls to 1%


Consistency is vital for minority influence to occur.

However, Nemeth et al (1974) agree that consistency is important but is not always enough in itself.  They carried out a variation on the procedure but allowed the participants to answer with a combination of colours. This time there were three conditions:

  1. The stooges randomly answer ‘green’ on half of the trials and ‘blue-green’ on the other half.
  2. The stooges answer ‘green’ to the brighter slides and ‘green-blue’ to the darker slides
  3. The stooges answer ‘green’ on every trial.

Assuming Moscovici et al to be correct, we would expect the third condition, in which the stooges are consistent to have the greatest influence on the minority.  However this was not the case.


Findings and conclusions:

The majority were most influenced by condition 2 since it is seen as flexible.  21% of participants were influenced by the minority in this condition.

In the other two conditions few participants were influenced.  In the first there is lack of consistency, (supporting Moscovici’s findings), and in the third there is a total lack of flexibility and no attempt for the stooges to use the more complex descriptions allowed.

Moscovici concluded that minorities are more likely to be influential if they are consistent but not to the point of being dogmatic.

Hogg & Vaughan (1995) claim that the following are important for minorities to be influential:

  • Principle: if the minority seem to be acting on principle rather than out of self interest
  • Sacrifice: if the minority have had to make sacrifices to maintain their position
  • Share characteristics with the majority:  if the minority are similar in age, race, social class etc.
  • Social trends: if the views of the minority are in keeping with social trends.  For example current trends in Western Society are tolerance and liberalisation.  Therefore calls by a minority for equal rights for a minority group are more likely to meet with acceptance.

Evaluation of Moscovici experiment:


The experiment uses stooges so deception is employed.  Whenever there is deception consent cannot be informed.              


It lacks ecological validity since it is a very trivial exercise, i.e. a silly disagreement over a slide that is very obviously blue.  This is not the sort of thing we normally disagree over, so does it tell us anything about minority influence in real life when very weighty matters of principle tend to be involved.

Explaining conformity and minority influence

These two phenomena do seem to contradict one another.  Indeed when Moscovici met Solomon Asch at a conference he reportedly felt embarrassed and concerned that Asch would be critical of his findings.  In fact Asch was very enthusiastic about the research on minority influence since he thought it helped to redress the balance that his research had tipped towards conformity.  Social impact theory proposed by Latane & Wolf (1981), is seen as one way of being able to explain both majority influence (conformity) and minority influence.

Social Impact Theory

This theory proposes that a person’s behaviour can be predicted in terms of three factors:  I’ll use the Iraqi debate as a contemporary example:

  • Strength

A message is stronger if it is repeated by a lot of people who are all in agreement.  This equates to Moscovici’s ‘consistency.’  You are more likely to be convinced that War on Iraq is right if all of your friends are in agreement.

  • Status and knowledge

The message will be strengthened if the person doing the convincing is an expert in the field.  A person who has lived under the Saddam Regime is likely to be more convincing than a politician who has never visited the area.

  • Immediacy

The message will have more impact if it comes from friends rather than a strangers.  Your friend trying to convince you of the need for war is going to have more impact than a bloke you’ve just met in the pub.


Other factors that may affect conformity

Changes over time

The original study by Asch was carried out in 1950s USA.  America was a very paranoid society, fearful of Communist take over and under the grip of McCartyism in which the government and other institutions in positions of influence were being purged of possible ‘Commies.’  People were afraid of appearing different or stepping out of line, so it is not surprising that Asch found such levels of conformity.

Later studies by Perrin & Spencer have found much lower levels of conformity.  However, some of these studies were on engineering students at a British University.  Since they were experts on accurate measurement of length it isn’t surprising that they failed to conform.  Out of several hundred trials Perrin & Spencer found only one incidence of conformity, despite the students being ‘very puzzled’ by the stooges’ bizarre answers!

When the study was carried out on young men on probation the rate of conformity was similar to those reported by Asch.

Cultural differences

If we consider culture in broader terms rather than narrow nationalistic ways, we can break societies into two broad kinds:

  1. Individualistic: for example Western Societies were the need to be independent and self sufficient is taught as the ideal. 
  2. Collectivistic: for example Asian and some African cultures were the needs of the family and larger social group are seen as more important.


Smith & Bond (1993) carried out a meta-analysis (see your notes on ‘validity’ in Research Methods), and found that collectivist societies tend to be more conformist.  Not surprising since they rely on each other to a much greater extent than selfish individuals in the West.

See later notes on the Temmi and Eskimos for similar example.


This is loss of self identity and was evident in Zimbardo’s Prison Simulation when the guards wore dark, reflective glasses.

Zimbardo thought it was responsible for the behaviour he observed in a study he carried out in 1969.  He found that female participants were more likely to administer electric shocks to other women if they were wearing lab coats and hoods that partly covered their faces. 

Role play

Johnson & Downing (1979) disagreed with Zimbardo.  They felt that Zimbardo’s participants were dressed like the Klu Klux Klan and were behaving accordingly, i.e. conforming to expectations.  They got participants to dress as nurses and found that despite the deindividuation that resulted that participants were less likely to deliver shocks.  They were conforming to the caring image of nurses.

Personality, intelligence and gender

Students are found to be less conformist.  This could be due to higher intelligence or to education that teaches independent thought and inquiry.

People who are measured high in ‘desire for personal control’ are less conformist than those measured lower.

Eagly & Carli (1981) found that women tend to be more conformist than men.


Explaining Conformity



Normative Social Influence


Need to be liked or accepted



Others are able to reward

or punish us


Conflict can arise between our own and other’s opinions





Informational Social Influence


Need to be certain



Seek information to reduce our uncertainty



Look to others for guidance




According to Kelman (1958) there are three types of conformity:

Compliance: you go along with the crowd and publicly agree with them.  However, internally you maintain your original views.  So using the S club analogy, you might outwardly conform and pretend to love the Life and works of this revered group in order to impress a fan, however, deep down inside you still consider their music to be vacuous, plagiaristic and infantile. Normative SI leads to compliance.

Internalisation: occurs when people take on the views of others both publicly and privately.  For example a person confused by the meaning of life may look to others for guidance.  A passing cult may spot the person and tell them that their leader was visited by Aliens many years ago and they have been told to go out and clone humans to be like their Alien masters.  Since this helps to reduce the person’s uncertainty about life they are taken in by these teachings, hook, line and sinker, and adopt their views both publicly, but more importantly inwardly too.  (Again not a good one to use just in case the marker belongs to the Raelian Movement).

Identification: occurs when a person conforms to the role that society expects them to play.  As with compliance there does not have to be change in private opinion.  The classic example here obviously is Zimbardo.




2. Obedience


Historical perspective

The work on obedience stemmed from Nazi atrocities during WW 2.  It was widely believed that Hitler himself was an evil genius, but he relied on the co-operation of millions of people to carry out his plans, including ‘the final solution.’ 

Hannah Arendt (1963) published her account of the trial of Adolf Eichmann titled ‘A Report on the banality of evil.’  Eichmann was the mastermind of the ‘final solution’ that involved using gas chambers in the death camps.  In her account Arendt  described Eichmann as ‘a dull, uninspired, unaggressive bureaucrat who saw himself as a cog in the machine.’ 

She concludes that the Nazis were mostly just ordinary people following orders.  Most controversially she believed that the rest of us would behave in a similar way, given a similar set of circumstances.

Milgram (1974) wrote ‘Gas chambers were built, death camps were guarded, daily quotas of corpses were produced with the same efficiency as the manufacture of appliances.’


In the 1960s it was still comforting to see the Germans as somehow a race apart, Milgram set out to show this.


Milgram's shocker (1963)

You are all aware of the study.  The danger is that in any question on Milgram you will regurgitate the procedure.  Read the question and tailor your answer to suit.  If it asks for findings concentrate on the percentages and the variations such as how close the ‘teacher’ stands to the ‘learner’ etc. 


·      An advertisement is placed in a local paper. Participants are paid $4.50 for taking part.  (Issue of payment is important).

  • Experiment is supposed to be on learning (deception).
  • Participant introduced to ‘Mr. Wallace’ (a harmless looking accountant in his 50’s, with a dickey ticker).  Mr Wallace was in fact a stooge or confederate.  (More deception).
  • Mr. Wallace and the participant draw lots to see who will be teacher and learner.  The real participant always becomes the ‘teacher.’
  • Mr. Wallace goes next door.
  • Participant is shown the equipment, and procedure is explained.
  • Mr. Wallace will be asked a series of questions.
  • An incorrect answer will result in an electric shock, delivered by the teacher.
  • The teacher is given a 45V shock to show that the equipment is real. (This is the only shock used in the experiment!!!!!).

·         The teacher sits in an adjoining room with the experimenter.

  • Control panel has switches, 15V to 450V, (labelled slight shock to Danger severe shock and XXX).
  • Each incorrect answer gets a shock 15V higher than the last.
  • The experimenter encourages the teacher with various instructions.
  • As the experiment proceeds Mr Wallace is heard to make various noises:


    • 75V, 90V and 105V a little grunt
    • 120V  complains about the pain
    • 150V ‘Experimenter get me out of here/’
    • 180V ‘I can’t stand the pain.’
    • 270V  an agonised scream
    • 300V  he shouts that he will answer no more questions.
    • 315V  violent scream
    • 330V  silence
    • 345V  and onwards there is silence




How it was done

% Obedience




Standard procedure

Teacher and learner in adjacent rooms


Closer proximity

Teacher 1 metre from learner


Touch proximity

Teacher has to push learner’s hand onto electrodes


Less prestigious setting

Experiment repeated in a run down office


Telephoned orders

Experimenter has to leave and phones instructions in.


An ally

A stooge disagrees with the experimenter


Less responsibility

A stooge gives the shocks when the ‘teacher’ says so.



Remember:  No shocks were ever received!!!

Milgram’s findings have been confirmed by others and there appear to be few sex differences.  Although Milgram’s original study was only carried out on men others have shown the same effect with women participants.  The experiment has also been replicated around the World.  Below are some of the findings.   

Cross cultural variations




% Obedience






Milgram (1963)

Male, general population




Female, general population



Mantell (1971)

Male, general population



Barley & McGuinness

Male students



Shanab & Yahya (1978)




Kilham & Mann (1974)

Female students



Ancona & Pareyson (1968)




Evaluation of the evidence

The research does tend to confirm Milgram’s original findings.  Most of the studies do suggest very high levels of obedience.  However, it is difficult to make comparisons between studies since there are differences in their methodologies.

  1. Different studies have used different populations, i.e. some have used students, others the general population.

  2. Milgram used a mild mannered Mr. Wallace with a dickey ticker.  In the Australian study a female student replaced him. 

  3. In most scenarios the ‘learner‘ was male, in the Australian she wasn’t.

  4. In the Italian study the maximum shock was 330 Volts.

The study that does stand out is the Australian study but this was women giving shocks to other women!


Evaluation of Milgram’s work

It is traditional to split this into two main sections:

1.       Methodology or validity

a.       Experimental validity

b.       Ecological validity

2.       Ethics.


1a. Experimental (or internal) Validity

By now you should know what validity means!  Did the participants taking part in the study actually believe that they were administering electric shocks to Mr Wallace?  Orne & Holland (1968) make a number of claims, each of which is refuted by Milgram:


Orne & Holland's claim

Milgram's defence

The participants realised that the set up was a sham.

70% of participants in later studies report afterwards that they thought it was genuine.

The participants obeyed because of the lab conditions, simply doing as was expected of them.

This criticism seems to be missing the point.  Milgram was trying to show that the situations we find ourselves in could cause obedience.

Obedience was due to payment in advance and the idea that a contract had been entered into.

This does happen in everyday life.  Presumably the SS were paid for their services in WW II.


1b. Ecological (or external) Validity

Can the results of the experiment be generalised to situations outside of the laboratory setting?  Since the person in the white lab coat was an authority figure, then Milgram believes that it does.  After all he was trying to show that we do obey such figures in real life.

Other studies that appear to support Milgram:

Experimental validity

Sheridan & King (1972) carried out a similar procedure but used a puppy as the ‘learner.’  The puppy carried out a learning exercise and each time it made a mistake it would receive an electric shock.  Participants, acting as the teacher, were led to believe that the shocks were becoming increasingly severe, as in Milgram’s original procedure.  In fact the puppy was getting a small shock each time, just enough to make it jump and show obvious signs of receiving a shock.  Eventually the puppy receives an anaesthetic to put it to sleep, and the participants think they’ve killed it.  Most participants continue to give it electric shocks.  The participants can be in n doubt that the puppy is receiving the shocks, so answering Orne & Holland’s first criticism.

Ecological validity

Hofling (1966) set up an experiment (natural, field or quasi?), in which a nurse receives instructions over the phone, from a Dr Smith, to administer 20mg of a drug Astroten to a patient Mr. Jones.  This instruction breaches three rules:

a.       The nurse did not know Dr Smith

b.       The nurse did not receive written authority

c.       20mg was twice the maximum dose suggested on the bottle.

Despite this, 21 out of 22 nurses were prepared to administer the drug.  Since this is a natural setting, it does have ecological validity, and as such is telling us something about obedience in real life.


For future reference, there are clearly ethical problems with the study:

a.   Nurses were deceived

b.   There was no consent

c.       No right to withdraw.

Bickman (1974).  People in the street are asked to pick up a piece of litter or stand on the other side of a bus stop etc.  The person doing the asking is dressed either as a milkman, a civilian or a guard.  People were more likely to obey the guard, showing, presumably, the power of uniform or of perceived legitimate authority.

2.      Ethics of Milgram (aaaaaaggghhh overload, overload!!!)



By who

Milgram’s defence

Measures were not taken to protect participants from physical or psychological harm

Baumrind (1964)

The results were unexpected.  Before starting Milgram asked professionals for their opinions.  Most thought the teacher would stop when the learner protested.

The right to withdraw from the experiment was not made clear to participants. 

Use of phrases such as ‘You have no choice, you must go on,’ would suggest participants did not have a choice.

Coolican (1990)

Milgram believes that they did have the right to withdraw, in fact, some did.


The experiment should have been stopped.


Milgram did not believe the distress caused was sufficient to warrant stopping!

Although participants gave their consent to take part, this was not informed since they did not know the purpose of the study or what it would entail.  Deception was used.

Baumrind (1964)

Milgram refers to deception as ‘technical illusions.’  Without them the experiment would have been meaningless.


Other points worth making in an essay on ethics of Milgram.

Milgram's main defence centres on the debrief that all participants received afterwards.  During this participants were reassured about their behaviour:

1.       They were reunited with an intact Mr Wallace

2.       They were assured that no shocks had been given

3.       They were assured that their behaviour was normal.  (Picture the scene, 'its okay Mr Smith, we all have maniacal, homicidal tendencies and feel the need to electrocute to death mild mannered accountants with dickey tickers!').

4.       They all received a full report of the procedure and findings

5.       They were all sent a questionnaire.


The questionnaire:  a staggeringly high 92% returned the questionnaire.  Of these:

·         84% were glad or very glad that they'd taken part.

·         74% claimed that they'd learned something of 'personal importance.'

·         Only 2% were sorry or very sorry that they'd taken part.

One year later, 40 of the participants were interviewed by a psychiatrist who concluded that none of them had suffered long term harm.

Many psychologists are still uneasy about the procedure.  Wrightman & Deux (1979) say that Milgram reports with awe and relish the extreme degrees of tension that his subjects experienced.  For example: they would 'sweat, stutter, tremble, groan, bite their lips and dig their fingernails into their flesh.  Full blown, uncontrollable seizures were experienced by three subjects.'

It is also worth mentioning that Milgram did not breach ethical guidelines, since they did not exist at the time!  In fact it was Milgram's study that was largely responsible for the introduction of such codes of conduct.

Each year Aronson (1988) says he asks his University students how many of them would behave like Milgram's participants.  Typically 1% believes they would!  This figure is the same as 1963, when, before conducting his experiment, Milgram asked students and psychologists to predict how many would deliver 450 Volts.

In 1965 Milgram was awarded the prize for 'Contribution to Psychological Research' by the American Association for the Advancement of Science.

Social Psychology’s hall of fame (or hall of horrors!):



Psychological explanations of obedience


Bit of examination advice 

This question:

“Outline two psychological processes that may be involved in obedience to authority.

An innocuous enough question that has appeared a whopping three times in the past few years.  This suggests to me that it’s generally badly answered and they will keep asking it to sort out the thoroughbreds from the also-ran candidates.  I think the problem with the question is that students don’t get what its asking for.  Faced with this question do not fear; simply  use two of the following explanations!


Legitimate authority

Society gives power or authority to certain people that they are able to exercise over others.  Obvious examples include the police.  Many examples are situation specific, for examples teachers (supposedly) have authority in schools, traffic wardens in parking areas, doctors over their patients etc.  Hofling (1966) and Bickman (1974) are examples of this.  Respect for authority, like this, clearly has its advantages in allowing for the smooth running of a society, and its rules are hammered home in all of us from a very early age.  The danger comes when we blindly obey such figures and as a result behave in an immoral way as a result.  This would help to explain some of the differences found in levels of obedience between different countries.  Some countries such as Australia have a history of questioning authority whereas countries like Germany teach their children from an early age to respect authority.

Agency theory

Milgram believed that we operate on two levels:

1.       As autonomous individuals, conscientious and aware of the consequences of our behaviour.

2.       As agentic individuals seeing ourselves as the puppets of others and no longer responsible for our actions

Normally we behave as autonomous, but under certain circumstances we undergo agentic shift and move to the agentic level.  They are then responsible only to the person giving the orders and their responsibility to others disappears.  He believed this explained the behaviour of participants in his own studies, with the experimenter being in charge during the agentic state.  It would also explain the behaviour of people like Eichmann who could switch from ordinary, dull uninspired etc., to mastermind of the final solution.

Milgram believed this shift was possible because we are taught at an early age to obey without question.  Once in the agentic state, binding factors keep us there:

1.       Fear of being rude and for example spoiling someone’s experiment.

2.       Fear of increasing our levels of anxiety by disobeying

Other factors:

Graduated commitment

Used by sales people the World over and usually referred to as foot in the door.  Get people to make a small commitment, i.e. buy a small item or give a small electric shock, and then build up to bigger, expensive items or ‘fatal’ shocks.  Once we’ve agreed to a small concession, then in principle it becomes more difficult to refuse a larger one.




Clearly this plays a part.  Some people are naturally more dominant whereas others tend to be more passive.  Moriarty (1975), (when he wasn’t annoying Sherlock Holmes), described the passivity type who will endure all kinds of unpleasantness, such as noisy neighbours, rather than confront the problem.


Remaining Independent

In both conformity and obedience studies many participants remain independent.  In Milgram’s experiment 65% obeyed, meaning that 35% did not, and in Asch’s study on conformity participants remained independent on 68% of trials.  13 out of the 50 that took part in the original study failed to conform once!


Factors affecting independence

1 Situational

Seeing others behaving independently

If we see others behaving independently then we are more likely to do the same.   Asch’s conformity rate of 32% fell to 5% if one of the stooges gave the correct answer on all trials.

Milgram’s obedience rate of 65% fell to 10% when two stooges acted as additional teachers and were prepared to disobey.


When Milgram moved his study to a less prestigious setting, rates of obedience fell (i.e. rate of independence increased).


2. Individual differences

a. Social desirability scale (Crowne & Marlow 1964)

Stang (1972) used this to show that those who do not seek the approval of others are more likely to remain independent.  This is like normative social influence in reverse. 

b. Personality differences.

When Crutchfield carried out his conformity experiment he administered a questionnaire and found the following personality types were most likely to conform:

‘Intellectually less effective,’

Having less mature social relationships,

Lower self esteem

Less leadership ability. 

Others, however, have found little evidence for a consistent personality type.  Burger found that those who scored high on personal control were less likely to conform to what others thought was a funny cartoon. 


3. Sex differences 

Milgram found no sex differences in obedience, but Kilham & Mann in an Aussie replication of the Milgram procedure using female students found 90% independence (i.e. only 10% obeyed).

4. Cultural differences

Berry (1966) found that Eskimos are far more independent than other native groups, for example the Temmi of Africa.  (See later notes on socialisation for reasons).


Not Conforming

Reasons for those remaining independent in Asch’s study:

1.       Confidence based:

The more certain we are, or the more expert we are in a particular area the more likely we are to stick to our guns and not conform to group pressure.  There may be area were you feel yourself to be particularly expert e.g. the hits of S Club or Blue.   Further examples are men who feel comfortable naming tools being prepared not to conform when others mis-identify them, and women having the confidence to do the same with kitchen utensils!

2.       Individuality-based

Simply the desire to want to remain independent.  In Asch’s study those that did this clearly felt uncomfortable and avoided eye contact with the others.

3.       Task-based

Some just wanted to get on with the experiment.


Not obeying

35% of Milgram’s participants refused to obey.  Possible reasons for this:

·         Responsibility.  One of Milgram’s participants refused because she’d lived in Nazi Germany and had seen enough pain inflicted in her lifetime.  Milgram believed past memories had ‘woken’ her from her agentic state

·         Education.  Gamson et al (1982) were conducting a research study when one of the participants became suspicious of the procedure and persuaded others to withdraw.  The participant had read about Milgram's research and questioned the legitimacy of the experimenters.  The person made use of their education.

·         Morality.  Kohlberg (more of him next year) outlined a number of stages of moral reasoning that we progress through.  He found that people who have reached the higher levels are more likely to disobey unreasonable demands and question authority when it appears unjust.

·         Disobedient model.  Watching others disobey reminds us that we are able to do the same! For example the variation on Milgram were stooges refuse to obey.

·         Knowledge of authority.  For example the one nurse out of 22 in Hofling’s study that knew the rules and refused to obey instructions to administer the drug.





Home   A2   Links