AS Social approach detailed ppt

Download Report

Transcript AS Social approach detailed ppt

SOCIAL PSYCHOLOGY UNIT
1
SOCIAL PSYCHOLOGY:
Definition & key social psychological terms
 How other people, groups, culture & society can influence our behaviour.
Individuals with perceived authority or charisma can influence us, e.g.,
obedience, self-fulfilling prophecy; groups can influence behaviour, e.g.,
compliance & conformity – Asch, Latane & Darley (Smoke filled room), crowd
behaviour & diffusion of responsibility.
 Culture & society can influence us: our culture can affect our response to
individuals & groups, who we believe has legitimate authority over us and may
also affect our general tendency for obedience (we may have an ethnocentric
bias).
Culture & society may influence our beliefs about perceived
attractiveness – rates of anorexia may reflect social & cultural influences.
 The power of the social situation can be very influential: even ‘good’ people
will do evil things if they are in an evil situation
SOCIAL PSYCHOLOGY:
Key Terms
 Agentic state: we surrender our free will & conscience to serve the interests of
the wider group; we see ourselves as primarily the agents of those with power and
in authority and only secondarily as individuals – responsibility shifts to those in
charge and we become de-individuated, denying personal responsibility
 Autonomous state:
essentially the opposite – we feel free to act as we wish,
including how our conscience dictates.
 Moral strain:
this is the result of having to do something we believe to be
immoral in order to function as an agent of those with power & authority, and so
benefit society. Denial (a defence mechanism) is often used to avoid the distress
associated with moral strain and having to do things we might normally find
abhorrent.
 In-groups/out-groups: in-group loyalty refers to our tendency to identify
ourselves as part of a particular group group & to classify others as either within or
outside that group; thinking of ourselves as belonging to one or more groups is
regarded as a fundamental feature of human nature. We tend to judge people not
in our particular ‘in-group’ more harshly than those we identify with, I.e., people
are either ‘us or them’: in-group loyalty V out-group hostility.
Key Terms (continued)
 Social categorisation:
categorising ourselves as members of a
particular group; these categories are ones we learn to be important, e.g.,
Goths, Emos.
 Social identification: Adopting the identity of the group to which we
have categorised ourselves, adopting consistent behaviours with this
categorisation, adopting the attitudes & behaviours of the in-group to
distinguish oneself from non-members.
 Social comparison:
comparing your group favourably to others; we
need to compare ourselves & our in-group favourably to others to maintain our
self-esteem. In & out-groups are measured against each other, the out-group is
devalued and the self-esteem of the in-group members is thus raised.
Research Methods/How
Science Works & Practical
 For the social approach you will need to know a
range of scientific terminology & be able to
describe & evaluate a number of different social
psychological research methods
 Many of these terms you will need to apply to a
short practical based on principles from social
psychology.
 When carrying out your own social psychology
practical you will need to keep a record of:
Research Methods/How
Science Works & Practical
 How you planned it;
 How you carried it out/methodology & procedure;
 How you analysed it;
 Your conclusions;
 Your evaluation of your practical.
Research Methods/How
Science Works & Practical
 Qualitative data: descriptive, nonnumerical
information,
such
as
generated by open-ended questions,
unstructured interviews etc..
 Difficult to analyse statistically
therefore to generalise from.
&
 Greater validity as provides more detail
& means answers can be explored in
more depth making conclusions more
meaningful.
 Quantitative data: information is
numerical in nature, such as generated
by closed questions, likert scales etc..
 Numerical data only tells you how often
behaviour occurs, not the underlying
motivation for behaviour, reduces
thoughts & attitudes to numbers which
undermines validity.
 Is reliable, as easy to repeat, easy to
statistically analyse & therefore to
generalise from.
Sampling methods (a representative sample is
drawn from the target population)
Method
Random
Procedure
Strength
Weakness
Every member of the target
population has an equal
chance of taking part in
research: like names being
drawn out of a hat
Most representative,
unbiased sampling
technique of a ‘normal’
population, researcher has
no influence over who is
chosen.
Very hard to do unless you
have a small population
group.
Opportunity
Participants selected from
whoever is available at the
time.
Ethical, no pressure to take
part , easy & quick
Not very representative, I.e.,
sample drawn from who is
available at time.
Stratified
A proportional
representation of the target
group, e.g., if target
group=60%male & 40%
female our sample will have
same proportion of males &
females
Likely to be very
representative of target
population
Time-consuming & difficult
to establish correct
proportions from target &
sample populations.
Volunteer/
self-selected
Participants select
themselves, e.g., respond to
advert
Well-motivated & people
might not normally have
access to
As volunteers their high
motivation may make them
as differently to other types
of participants.
Sampling methods (a representative sample is
drawn from the target population) continued.
•[For many psychological studies the sample is made up of students because
researchers will often use opportunity sampling]
Method
Procedure
Strength
Weakness
Systematic
sampling
This involves selecting
every nth person from a
list or group, e.g., from
a list of 1000 selecting
every 10th person
Likely to be more
representative than
simply choosing first 10
on a list or from a group
Still not entirely
representative as
method is not entirely
random & need big
sample for it to be
effective, which may
not always be practical.
Surveys: questionnaires & interviews
Method
Description Strengths
Weaknesses
Unstructured
interview
Questions are open-ended,
not systematic, each
question depends on
answers given previously:
qualitative data
Flexible, more detail can
expand on answers, get
information about attitudes,
beliefs underpinning
behaviour, high validity.
Cannot be replicated,
unsystematic, time
consuming, subjective
interpretation
Semi-structured
interview
Is a schedule of questions
but some flexibility to
expand on responses
See above
See above
Structured
interview
Systematic, pre=set
questions asked of every
interviewee
Systematic, very reliable as
easy to repeat & compare
answers.
Less detail elicited as little
opportunity to expand on
answers: lower validity
Closed Questions
1 word, yes/no answers,
Likert scales, limited
responses, yields
quantitative data
Reliable, easy to replicate,
easy to interpret &
statistically analyse
answers, objective, can be
large-scale
Low validity, little detail,
superficial, only tells you
how often, not why.
Open questions
Can be answered how
participant chooses, words
not numerical response,
yields qualitative data
Detailed, more valid
(opposite to closed
questions)
Low reliability, subjective
interpretation (opposite to
closed questions)
Research Methods/How
Science Works
 Surveys – questionnaires & interviews generate self-
report data, i.e., information elicited from questions
which relies on the participants reporting their own
behaviour, feelings, attitudes etc.
Research Methods/How
Science Works
 Before starting a piece of research a hypothesis has to
be made: a hypothesis is a testable prediction.
Prediction of human behaviour is made and then
tested to ascertain if this hypothesis, or prediction, is
actually true for most people
 There are 3 types of hypothesis.
Research Methods/How
Science Works: Hypotheses
 Experimental hypothesis: This is a prediction testable by means of
either laboratory, field or natural experiment.
 Alternative hypothesis: This is a prediction testable by means of
research methodology other than experiments, e.g., questionnaires,
interviews, observations, correlation studies, longitudinal, Crosssectional & cross-cultural studies.
 Null hypothesis:
This is NOT the opposite of the alternative or
experimental hypothesis – it is a rejection of it. The null hypothesis
states that the prediction is wrong, there is no such effect (as had been
predicted) other than effects produced by chance.
 E.g., Exp Hypothesis: coffee makes you more alert: Null hypothesis:
coffee has no effect on alertness, any increase in alertness due to
chance factors.
Ethics & research on human participants
 Introduction: researchers must ensure that public, after taking part in research, have
confidence in the psychology profession & have a positive perception of psychologists.
All participants should be treated with respect & their dignity & well-being should be
safe-guarded at all times.
 Informed Consent: participants must be made aware of the aims & procedure of the
research to enable them to make a fully informed decision about whether to take part or
not. Sometimes, to avoid demand characteristics, participants may be deceived about
the nature of the research, or they may be in a field experiment or observation where
informed consent cannot be obtained prior to research. In these cases participants must
be fully debriefed after the research. Where informed consent cannot be obtained
presumptive consent can be obtained (would other people, if the scenario was explained
to them agree to take part in the study themselves).
 Debriefing: participants must be fully debriefed at the end of the research, I.e.,
everything about the nature of the research must be revealed to them, & they must be
reminded of their right to withdraw their results from the study & given the chance to ask
any questions about the study. They must leave in same emotional state as they arrived.
 Withdrawal: participants must be made aware that they can withdraw at any time, even
if they have been paid, and they can also withdraw their data.
Ethics & research on human participants
 Competence: Researchers should do research and make judgements only in areas
appropriate to their area of expertise; must check with colleagues if there is any doubt – or
not carry it out research.
 Deception: wherever possible participants should not be deceived unless vital to preserve
experimental validity & should be fully debriefed at the end.
 Confidentiality: unless agreed with participants in advance, confidentiality should be
maintained, no personal information should be disclosed & pseudonyms used.
 Protection from harm: Participants should be protected from physical & psychological
harm & should be exposed to no more risk than they would normally encounter in their
usual lives. Participants should leave the research feeling positive about themselves & the
experimental experience.
 Observation: participants should only be observed in places where public behaviour is
expected.
Research Methods/How
Science Works
 Reliability: this refers to the consistency of data – if the research is reliable we
would expect that if it were repeated, with similar types of participants in
similar circumstances, the same results would be obtained.
 Validity: Does the research actually measure what it is supposed to measure.
In psychology testing abstract concepts can be difficult, we rely on observing
measurable changes in behaviour & attitudes/beliefs; however, we cannot
always be sure that what we think we are testing is actually being reflected in
the participants’ responses – these responses may be due to factors other than
the ones we are thinking we are testing.
 Subjectivity: This refers to the interpretation of data, could participants’ data
be interpreted differently, is the interpretation of data completely free from
bias?
 Objectivity: Essentially the opposite, is data able to be interpreted in such a
way that it is deemed unbiased, untainted by attitudes, beliefs & values of the
researcher.
Research Methods/How Science
Works: The Practical
 Develop a hypothesis & null hypothesis.
 Consider the ethics of your practical: questions should not cause distress,
embarrassment.
 Consider how you will generate quantitative & qualitative data: closed
questions & open-ended questions, questionnaire & small-scale unstructured
interview/semi or structured interview.
 Sampling: who is your target group what type of sampling method are you
going to use to ensure a representative sample, what issues are there with your
sampling method, how big will the sample be?
 Operationalising your research: how will you operationalise your questions &
variables, e.g., if you are trying to measure attitudes, levels of prejudice, out/in-
group loyalty or obedience how will your define & measure these terms?
What/who will you be comparing?
 Conduct a pilot study: ask a very small number of participants the questions
you wish to first to ensure that they fulfil the criteria you want, if they don’t
you have the opportunity to change them at an early stage. These responses
can be included in your final results.
Research Methods/How Science
Works: The Practical
 Analysing the results: Quantitative data: look for numerical trends by establishing mean,
median & mode, range & standard deviation.
 Analysing the results: Qualitative data: look for trends/themes in the answers given to
open-ended questions.
 Reliability: is your study reliable? What have you done to standardise instructions &
procedure to avoid experimenter bias & ensure the research is well-controlled?
 Validity: What have you done to ensure high validity & avoid demand characteristics and
confounding/extraneous variables affecting your results?
 Do your results support the alternative/experimental hypothesis or the null hypothesis?
Why? Have your participants been fully debriefed?
In depth area of study:
Obedience & Prejudice
Define the terms
prejudice &
discrimination
Social Identity
Theory of
prejudice
Agency theory
Milgram’s study
& 1 variation of it
Ethical issues &
the study of
obedience
1 cross-cultural
study of
obedience
OBEDIENCE
Obedience: to allow ourselves to be directed by an individual[s] who we perceive
to have power/authority over us. It differs from compliance, which means simply
going along with suggestions or instructions without being directly ordered to;
and conformity which is where we adopt the attitudes & behaviour of those
around us without being directly ordered to by an authority figure.
 Studies of obedience, why people carry out orders which they seemingly find
abhorrent and go against their moral believes and values, gained impetus after
the destructive obedience demonstrated by some Nazis during World War II,
such as the Holocaust.
 Genocide, as illustrated by the Holocaust, is the most obvious example of
destructive obedience; other examples of destructive obedience include
genocide in Rwanda, Iraq and the former Yugoslavia, the Mi Lai massacre and
human rights abuses in the US run military prison in Iraq – Abu Ghraib.
Milgram’s classic study of obedience (1963)
Description - APRC



Name: Milgram’s study of obedience
Aim: to investigate how obedient people would be in a
situation where following orders would mean breaking
participants’ moral codes & harming another person:
to test the hypothesis that the ‘ Germans were
different’.
Method: procedure-Milgram advertised for 40
volunteers (males aged 20-50) to take part in a study
on human memory (really obedience).
Each
participant was introduced individually to Mr Wallace
(a confederate) and told that either they, or Mr
Wallace would be randomly allocated the roles of
either ‘teacher’ or ‘learner’. Mr Wallace was always
the ‘leaner’ who would receive an electric shock every
time he got a memory question wrong (Mr. Wallace
was in another room so could be heard but not seen).
The ‘teacher’ (the naïve/real participant) was given a
small shock at the beginning of the experiment to
illustrate its effects. Shocks went up in 15v increments
up to 450v marked ‘XXX’. The researcher & Mr
Wallace followed a carefully scripted set of responses
& prompts.


Results: Before the research Milgram asked
psychologists & psychiatrists to estimate what % of
naïve participants would inflict shocks – the estimate
was les than 1%. In reality 100% went to 300v & 65%
went to 450v. [NB. Obedience is operationalised as
going to 450v.]
Conclusion: Milgram concluded that the power of the
social situation is a powerful determinant of behaviour
– we are socialised from an early age to recognise
authority and obey those with perceived power.
Milgram’s study:
Evaluation GRAVE




Generalisability: can we generalise from Milgram’s
sample to the wider population, I.e., is there
population validity? Despite the fact that only US
males were used, they were from a wide range of
backgrounds & ages; furthermore, Milgram found very
similar results with female participants and crosscultural results tend to be similar (see table below).
Reliability: a standardised procedure was followed,
I.e., the same script & verbal prompts were used for
each participant.
Application to real life: there are numerous
examples in real life where we have to obey authority
figures.
Validity: did the study have experimental validity?
Did the participants really believe they were giving
electric shocks? Milgram did his best to convince the
naïve participants of the veracity of his research: he
staged a sample 45v shock for the naïve participant,
the equipment looked real and the cries of Mr Wallace
seemed genuine. Ecological validity: do the results
tell us anything about real-world behaviour? Would
the participants really behave as they did in the real
world and not in a laboratory situation?

Validity cont’d: Variations of Milgram’s experiment,
e.g., when done outside a lab and in a run down office
block, saw obedience levels fall, but where still very
high at 48%. Hofling’s (1966) field experiment found
similar results and later real life incidents, e.g., Mai Lai,
support the ecological validity of Milgram’s study.
Sheridan & King (1972) conducted a procedural similar
studies to Milgram’s on puppies, participants thought
they were shocking puppies: similar results were
obtained.

Ethics: Participants were deceived but that was
necessary to ensure experimental validity, but this
meant they could not give informed consent.
However, participants were fully debriefed at the end
& 84% said they were glad or very glad they had taken
part, only 2% said they were sorry to have taken part &
74% claimed to have learned something of personal
importance. Withdrawal was made difficult, but in
real life situations it is often not easy to disobey &
follow your conscience, so it was important to make
withdrawal hard to simulate real life.
Milgram’s study (continued)
 Ethics cont’d: Participants were put under a great deal of stress and caused much
distress, but they did not have to administer the shocks and could, in reality,
withdraw whenever they wanted. Moreover, Milgram consulted experts before the
research & no one predicted the level of obedience obtained. Finally, the wider
benefits to society might be considered to outweigh the costs to the individual
participants: a cost-benefit analysis.
Variations of Milgram’s original study
WE DID THIS ONE!!! Milgram Variation ~ The liberating effect of group pressure (1965)
Aim: To see if obedience was affected by the presence of others who disobey the authority of the
experimenter.
Procedure: the same
Results: 1 ppt dropped out 105volts, 3 dropped at 150volts, 12 dropped at 210 volts when the 2 nd
confederate refused to continue. 4/40 went all the way to 450volts
Others:

Prestige: the experiment was moved from Yale university
to a rundown office block – obedience levels dropped to
47.5%.

Responsibility: when the participant was not directly
responsible for the shocks, I.e., they simply had to read the
paired words & someone pressed the electric shock button,
obedience rose to 92.5%. When the ‘teacher’ (naïve
participant) had to hold the ‘learner’s’ hand on the
electric shock plate obedience dropped to 30%.

Buffer: if the ‘learner’ & ‘teacher’ were together in the
same room so that the ‘teacher’ could be seen & heard,
obedience dropped to 40%. When the ‘learner’ could not
be seen or heard, all the participants went to 450v.




Personal control: when participants were allowed to
choose their own shock level they always choose the
lowest.
Witnessing disobedience: where a participant saw
another ‘participant’ refuse to beyond administering 150v,
only 10% then obeyed the experimental instructions &
carried on to 450v.
NB., these variations in the social situation, and their
impact on the levels of obedience, illustrate how
powerful the social situation is in determining human
behaviour.
NB., the evaluative points applied to Milgram ’ s
original study basically apply to the variations.
Ethical issues & the study of obedience
 Studies of obedience often involve deception, preventing participants giving informed
consent, but this is often necessary to ensure experimental validity.
 Participants may experience significant distress, not least because they may find out
quite negative things about themselves, I.e., they are prepared to obey & cause harm to
others.
 Withdrawal is sometimes made difficult to simulate the effects of obedience in real life
situations – but participants can be fully debriefed afterwards.
 The benefits of the research to wider society may outweigh the psychological &
emotional costs to the participants, i.e., finding out why even good people do horrible
things in order to understand why atrocities are committed & prevent them from
happening in the future.
Milgram’s agency theory
 Milgram argued that general tendency to obey those we perceive to have authority is a
mechanism to ensure a stable society. To run smoothly, complex societies require us to
obey a vast range of social rules; keeping to these rules means that we have to give up a
certain amount of our free will.
 To enable us to give up a degree of our free will we have evolved 2 states: autonomous &
agentic (see the definitions of these terms above).
 We are socialised from an early age into developing the capacity for the agentic state.
This process starts in the home, continues in school and into the workplace: to maintain
order in different social situations we give up our free will and obey parents, teachers &
employers (people often put the needs of their employers above their own, e.g., they work
longer than their contracted hours, take work home, sacrifice family & quality of life for
their jobs).
 We use this agentic state to avoid moral strain; when we do things against our conscience
we believe we have little choice as we are acting as agents of authority figures.
Evaluation of agency theory
 It has face validity (on the surface it seems to explain human behaviour in many
situations, school, the workplace, armed forces etc.)
 There is a lot of experimental support for it, e.g., the studies of Milgram & Hofling.
 A study by Blass (1996), where participants saw extracts from the original Milgram study,
showed that these participants blamed the researcher for what the naïve participants did,
i.e., they accepted that these naïve participants were the agents of the authority figure, in
this case, the researcher.
 A study by Bushman (1988) varied the authoritativeness of the authority figure, when the
authority figure had more authority/status (e.g., a uniform) obedience was more likely.
We are more likely to become agents when perceived level of authority increases.
 When participants are reminded of their potential for autonomy, e.g., they see someone
else disobey, obedience levels drop.
 Not everyone in the Milgram study gave up their autonomy and entered into the agentic
state: this theory cannot explain individual differences in levels of obedience (see
authoritarian personality to explain personality differences in obedience).
Evaluation of agency theory
(continued)
 The obedience alibi: David Mandel argues that agency theory ‘lets people off the hook’
for their heinous actions, I.e., it gives concentration camp guards an excuse for their
deplorable behaviour; ‘I cannot be held personally responsible for my actions, I was only
obeying orders’. This might have some validity in some cases but should be treated with
caution.
 Circular argument: it cannot be defined independently from obedience – people obey
because they are in an agentic state, but are in an agentic because they obey. Circular
arguments have limited explanatory value – because you simply go around in circles!
 Personality (charisma): it doesn’t take into account personality variables & obedience,
some people might be naturally more predisposed to obey, whilst some people can get
others to obey them even when they have little or no authority over them, it is simply the
force of their personality (charisma) which elicits obedience.
 There are other explanations of obedience, e.g., French & Raven (1959) Social Power
theory. This theory states that different people in different social situations have
different types of social power: Reward, punishment (coercive), legitimate, expert &
charismatic power. Milgram’s researcher had some of these types of power.
Cross-cultural studies of obedience





Studies of obedience carried out across cultures find similarly high levels of obedience to those found
by Milgram. This shows that we are, by nature, social beings, heavily influenced by our social
environment & setting: the power of the social situation.
However, cross-cultural studies of obedience often use different methodologies, so like is not always
being compared with like.
In an Australian study, Kilham & Mann (1974), the ‘learners’ had long hair & may have been
perceived as more or less deserving of electric shocks as a result. Also, in this study female students
were asked to shock another female (in Milgram, the ‘learner’ being shocked was always male).
Hamilton & Sanders (1995) presented participants from US, Japan & Russia with scenarios where a
crime was either an individual’s idea or the order of a superior. Little responsibility was attributed to
the person who acted criminally under orders, but that reversed when they acted on their own
volition. However, cultural differences emerged: US participants attributed more personal
responsibility to individuals acting criminally under orders than did the Japanese & Russian ones;
thus obedience might be deduced to be more important in Japanese & Russian culture than US
culture, I.e., obeying even criminally wrong orders might be seen as appropriate more in Japan &
Russia than in US.
Finally, most obedience research involve studies that were conducted in western industrialised
nations, so universal conclusions about human nature cannot necessarily be drawn (see table below).
Cross-cultural replications of Milgram’s
obedience research
Study
Country
Participants % obedient
Ancona & Pareyson
(1968)
Italy
Students
85
Kilham & Mann
(1974)
Australia
Male students
Female students
40
16
Burley &
McGuinness (1977)
United Kingdom
Male students
50
Shanab & Yahya
(1978)
Jordan
Students
62
Miranda et al. (1981)
Spain
Students
90+
Schurz (1985)
Austria
General population
80
Meeus &
Raajimakers (1986)
Holland
General population
92
Meeus & Raaijmakers (1985) Administrative
Obedience: carrying out orders to use psychologicaladministrative violence



Name: Meeus & Raaijmakers (1985)
Aim: To test the concepts of obedience
illustrated by Milgram by in a more up-to-date
way & in a culture more liberal than 1960s US
– 1980s Holland. Would obedience still be
high if psychological, as opposed to physical,
harm was to be applied?
Method: 24 naïve participants took part in
what they thought was a job interview that
required the applicant being able to tolerate
stress. In 1 condition an experimenter sat in
with the naïve participant who was to
interview the applicant (a confederate or
stooge, much like Mr. Wallace in the Milgram
study). The naïve participant was told to
cause the applicant stress being making a
series of graduated cutting comments to the
applicant, with 1 being the most innocuous &
least offensive to 15 being the most offensive.
 1 = your answer to question 9 was wrong:
15=according to the test it would be better
for you to apply for lower functions. The
applicant would show increasing levels of
distress as the offensive comments
progressed up the scale to 15.
 Results: In the experimental condition,
where the experimenter sat in on the
interview, 22/24 (92%) of participants
made all 15 stressful comments. In the
control condition, where the naïve
participant was alone, none did. Rates of
obedience were higher than in the Milgram
study.
 Conclusion: Even in a liberal culture like
the Netherlands, people obeyed an
authority figure & were prepared to abuse a
stranger psychologically. People are even
more willing to abuse people emotionally
under direct orders.
Meeus & Raaijmakers: Evaluation GRAVE



Generalisability: adults from the general
population, not just students, so is therefore
more representative and so generalisable, and
the results are consistent with other studies
done in Europe, so arguably there is good
population validity.
Reliability: well-controlled, standardised
statements, I.e., from 1-15 generating
quantitative data which can be objectively &
easily analysed & interpreted. The study
supports the findings of Milgram & Hofling;
the high levels of obedience found in this
study can be explained by the non-physical
nature of the abuse required to be obedient.
Application to real life: psychological abuse
is far more common in the real world than
physical abuse, especially in the workplace.


Validity: researchers maintained that ecological
validity was high, arguing that the type of abuse
depicted in the study was more common in society
than the physical abuse depicted in Milgram’s
study and therefore more realistic. However,
experimental validity was still quite low because the
scenario is still quite extreme and bizarre.
Ethics: although the level of distress experienced
by the naïve participants might be less than in
Mlgram’s study, not giving electric shocks so no
physical harm was thought to being perpetrated, it
might still be distressing for the participants
because they are seemingly causing mental stress to
another person. The experiment also required a
high level of deception to work. There was no ‘lie’
about the reality of the ‘shocks’ as in the Milgram
study – so consent was more informed; however,
deception was still used, the interviewee was an
actor, but deception necessary for experimental
validity.
In depth area of study:
prejudice & discrimination
PREJUDICE: To pre-judge, when we allow our
stereotypes to affect our beliefs & attitudes about a group
of people (often based on little or no knowledge of them).
DISCRIMINATION: Actions or treatment based on
prejudice. Unequal treatment of individuals or groups –
often based on characteristics such as race or sex.
STEREOTYPE: A set of fairly fixed, simplistic
generalisations (or characteristics) about a group or class
of people.
Prejudiced Attitudes
Attitude: how positive/negative you feel about
something
A=AFFECT – how you feel about something, e.g., mood,
emotional state: anger, fear, suspicion, hostility.
B=BEHAVIOUR – how you behave, e.g., insulting,
avoidance, physical attack
C=COGNITION – what you think about something:
knowledge & attitudes based on stereotyping.
PREJUDICE: Social Identity Theory (minimal
groups)
 Social identity theory states that simply being in a group, or





perceiving that you are in a group, is enough to create ingroup loyalty & out-group hostility.
Social identity theory & the process of generating in-group
loyalty & out-group hostility is made of 3 features:
SOCIAL CATEGORISATION
SOCIAL IDENTIFICATION
SOCIAL COMPARISON
(SEE DEFINITIONS OF THESE TERMS ABOVE)
Social Identity Theory (continued)
 Social identity theory is illustrated by the research carried out by Tajfel (1970).
 Participants were placed in groups according to minimal criteria – whether or





not they liked the same paintings, or, when estimating the number of dots on a
screen, they were under or over-estimators. In reality the participants were
allocated to groups entirely randomly.
A member of each group was then given the same task to perform.
Members of each group then had to allocate rewards to the people performing
the simple task.
Despite the fact that the task being performed was the same for the individuals
from each group, both groups decided to reward the member of their own
group more highly than the member of the other group.
There was no direct competition between the 2 groups and what members of
each group thought they had in common with each other was minimal, i.e.,
liking the same painting, or being an under/over-estimator.
Nevertheless, members of both groups were prepared to discriminate in favour
of the member of their particular group; presumably because this increased
their own social standing/self-esteem by the process of social comparison.
Evaluation of Social Identity Theory
 The theory has a certain amount of face validity as it can successfully
explain many aspects of real-world behaviour & be applied to a wide
range of social situations, e.g., football teams, racism, Emos/Goths.
 There is great deal of empirical, scientific research which supports the
theory, e.g., the research carried out by Tajfel, to an extent Sherif (1961),
as the boys immediately developed strong in-group loyalty & out-group
hostility when they knew there was another group of boys in the
woods.
 The theory has useful applications: because it can explain how
prejudice & discriminate originate, it can also be used to reduce
prejudice & discrimination, i.e., by preventing in-groups from forming,
mixing up social groups, trying to prevent social categorisation &
identification.
Evaluation of Social Identity Theory
(continued)
 Some contemporary research into minimal groups suggests discrimination &
prejudice is more complex. E.g., Dobbs & Crano (2001) showed that where
individuals perceived that their in-group was in the majority there was much
less in-group favouritism & out-group hostility than when they perceived their
in-group was in the minority, (then the situation reversed); I.e., more likely to
be anti-English if you are Welsh or Scottish than the other way around, because
Welsh & Scottish people are the minority in the UK.
 Social Identity Theory cannot explain individual differences in levels of ingroup loyalty & out-group hostility; not everyone in a particular in-group will
have the same level of loyalty towards the in-group & hostility towards the out
group (authoritarian personality theory may explain individual differences in
prejudice better).
 Finally, people may have all sorts of complex reasons for identifying with each
other, not just minimal reasons based on social standing/self-esteem, e.g.,
shared cultural history, shared histories of conflict & battles for resources.
Key studies in detail from Social
psychology
Milgram
(1963)
Milgram
variation
(The
liberating
effect of
group
pressure
(1965)
Meeus &
Raaijmaker
(1965)
Hofling et
al. (1966)
nurses
study.
OR Tajfel
(1970)
minimal
groups
Hofling et al. (1966): Experimental
study in nurse-physician relationships




Name: Hofling et al. obedience in a natural setting
Aim: To investigate nurse-physician relationships, I.e.,
investigate effects of authority on obedience in a
natural environment (a hospital)
Method: field experiment; 12 graduate nurses & 21
student nurses asked to fill in a questionnaire about
how they would act in the experimental situation. 22
nurses from 2 separate hospitals took part in
experimental condition. While alone on a ward they
were asked by an unknown doctor over the phone to
break 4 hospital rules:.
Method cont ’ d: 1 Give an over dose of a drug
{Astroten 5mg}to a patient (it was really a placebo). 2.
Instructions were given over the ‘phone, not in person.
3. The particular drug was unauthorised for use on
that specific ward. 4. The instruction was given by an
unfamiliar voice. The ‘doctor’ used a written script to
standardise the procedure & all conversations were
recorded. Results were operationalised thus: nurse
complies & goes to give drug; consistently refuses to
give drug; goes to get advice; becomes upset; call lasts
longer than 10 minutes.
 Results: results of questionnaire=10/12
graduate nurses & 21/21 student nurses
believed no one would give medication.
Results of experimental condition=21/22
started to give the medication, calls were
brief, only 11 nurses were aware of dosage
limits for drug Astroten, none were
hostile towards the caller & all admitted
knowing what they were doing was
against hospital rules but said it was a
fairly common practice.
 Conclusion: the perception of authority
(in this case a doctor) is enough to
generate obedience, even when this
could possibly endanger a patient’s life.
Hofling et al.
Evaluation GRAVE


Generalisability: although the study only
involved nurses, and nurses might conceivably
be more inclined to obey doctors, the nurses
in the study were simply those on duty at time,
not specifically chosen. Also, nurses are not a
particularly unique set of people, therefore,
the study might be considered fairly good in
terms of population validity.
Reliability: the study was run 22 times with
similar results, and the procedure was
standardised. However, a field experiment has
limited control over extraneous variables, so
reducing reliability, I.e., the nurses might have
been very tired, or overworked that particular
day, so reacted unusually & without thinking,
I.e., not how they would normally react, but
Hofling did have an observer on the ward to
ensure that conditions were right for the
experiment to proceed.


Application to real life: this was done a in a
real world environment and the negative
effects of nurses obeying inappropriate
instructions from doctors is very real &
important.
Validity: the study has very good ecological
validity because it was done a real hospital
with nurses who were unaware they were
taking part in a study – it demonstrated real
behaviour – nb., the difference between the
questionnaire results 7 actual behavioural
results. It also has experimental validity as the
nurses were not aware they were in an
experiment so behaved naturally, there were
no demand characteristics so experimental
validity was high, participant behaviour was
not for the benefit of the study or the
researcher.
Hofling et al. (continued)

Ethics: ethically the study was very dubious.
No informed consent was obtained and the
nurses were deceived. This was necessary for
the experimental validity of the study & to
avoid demand characteristics but would be
upsetting for the nurses. Although the nurses
were fully debriefed, to offset the deception,
they admitted to feeling shame, guilt &
embarrassment at their professional behaviour
– research should ideally leave participants
feeling positive about themselves, or they
should leave in the same emotional state as
they entered it. However, the nurses were
reassured that they had acted normally,
patient care had not been affected & they were
not criticised for their conduct.
Tajfel et al. (1971) Social Categorisation
& Intergroup Behaviour




Name: Tajfel et al. (1971)
Aim: To test whether the act of placing people
into 2 clearly identifiable groups, based on
minimal intra-group similarities & not in
competition, would be enough to produce
prejudice between groups of very similar people.
(NB., inter=between/intra=within)
Method: See explanation of Social Identity theory
above. 2 versions of experiment, 1 involving
paintings by Klee or Kandinsky, and 1 involving
estimating numbers on a screen (being an under
or over-estimator). The participants were initially
placed into groups according to whether they were
under/over-estimators,
or
their
painting
preference (in reality the allocation to groups was
entirely random).
Method cont’d: Participants were then given
the opportunity to allocate points, which
could be converted into prizes, to members of the
two groups. The participants did not know who
they were allocating points to, but did know which
group they belonged to. In another variation,
Tajfel further manipulated the experiment by
ensuring that when participants favoured
members of their in-group, the out-group would
automatically get more points.
 Results: The participants overwhelmingly
chose to favour their own group by allocating
more points to members of their own group,
even when this meant the out-group would
then get more overall points, & therefore
prizes.
 Conclusion: Even when categorised into
meaningless/minimal groups, participants
still chose to favour members of their own
group over members of the other group.
This shows we have a natural tendency in
social situations to favour people we have
identified & defined as being part of our
‘ group ’ & discriminate against those
perceived to be members of a different group.
One explanation of this is that by favouring
members of our own perceived in-group, we
boost our own self-esteem, because we are
part of that group.
Tajfel et al. (1971) Social Categorisation
& Intergroup Behaviour (continued)
 Generalisability: The research has been
replicated on many different social
groups, not just schoolboys, as in the
original studies, e.g. adults in Cardiff,
female adults in California, soldiers in
Germany: all showed similar minimal
group effects. Therefore, this research
does have population validity.
 Reliability: The study was easy to
replicate because the procedure was
strictly controlled & very similar results
have been obtained across different
cultures & groups.
 Application to real life: In the real
world we are very often allocated to
groups based on minimal criteria, e.g.,
school, workplace. Can be used to
reduce prejudice, e.g., merging in-groups
& out-groups.


Validity: The study lacked ecological validity
because it was quite removed from the real life
experiences of the participants, I.e., being asked to
estimate dots on a screen, or be placed in a group
according to painting preference. Furthermore,
the study was carried out in a university setting
which would be unfamiliar to many of the
participants.
Experimental validity may be
questioned because arguably there was implied
competition created by the forced nature of the
choice participants had to make between
members of their in-group or those of the outgroup.
Ethics: There are no real ethical issues and the
participants were not caused distress. As some of
the participants were schoolboys, informed
consent & withdrawal may have been issue the
boys had felt intimated by the adult researchers &
university setting. The research does provide very
useful insight into the mechanics of prejudice &
discrimination and, therefore, ways of reducing
prejudice & discrimination.
This was an additional study – you do not
need to have APRC and GRAVE if you are
using Tajfel
Sherif et al. (1961): Intergroup conflict & co-operation:
The Robber’s Cave experiment




Name: Sherif et al. (1961).
Aim: To see if prejudice can be created
between two very similar groups by putting
them in competition with each other.
Method: A field experiment: 22 12 year old
white, lower middle-class protestant boys were
taken to a summer camp in Robber’s Cave
national park, Oaklahoma. They were all very
similar & psychologically well-adjusted. They
were put into 2 separate groups & for first 5
days each group given tasks to perform to help
them bond as a group & given names (Rattlers
& Eagles). Over next 4 days tension was
generated between the 2 groups by staging a
series of competitions between the 2 groups.


Method cont ’ d: Once hostility had been
created the researchers tried to reduce it by
bringing both groups together for joint
activities and problem-solving tasks.
Results: A strong in-group preference & outgroup hostility was shown by each group; this
was eventually reduced by the joint problemsolving tasks.
Conclusion: Competition increased prejudice
& discrimination, leading to clear inter-group
conflict; however, there was some hostility
between the groups as soon as they were aware
of each other. Working together on cooperative tasks successfully, but not entirely,
reduced prejudice & discrimination between
the 2 groups.
Sherif et al. (1961): Intergroup conflict & cooperation: The Robber’s Cave experiment



Generalisability: The sample was not very
representative, i.e., all white, protestant,
middle class young boys – lacked population
validity.
Reliability: although the boys were all tested
to ensure they were psychologically welladjusted and they were all similar
backgrounds, in a field experiment such as this
it is very hard to control confounding &
extraneous variables.
Application to real life: There are many
examples of tension & conflict over resources
leading to prejudice & discrimination, e.g.,
Northern Ireland, race riots in northern
England, Israel & Palestine. Also, how to
reduce prejudice, e.g., through 2 opposing
groups working together to solve a common
problem – called a superordinate goal.


Validity: Ecological validity was high because
they experiment was conducted in a natural
environment, therefore, eliciting natural,
uncontrived behaviour; also there was high
experimental validity as the boys did not realise
their behaviour was being observed & that
they were in a experiment, so there would be
no demand characteristics (trying to please
the researcher). However, even before
competition started, as soon as the groups
knew of the existence of another group there
was out-group hostility – simply being in a
group seemed to be enough to create this,
there was no need for competition. The
competition simply strengthened
that
hostility.
Ethics: the boys were not harmed or
distressed, physical hostility was prevented &
the researchers endeavoured to reduce the
discrimination & prejudice at the end.
Sherif et al. (1961): Intergroup conflict & cooperation: The Robber’s Cave experiment
 Ethics cont’d: Nevertheless, the did set out to deliberate create
something negative: prejudice & discrimination, but do the benefits
of the research outweigh the costs to the participants?
1 key issue in Social Psychology
Obedience during conflict:
destructive obedience –
Holocaust, Mi Lai massacre, Abu
Ghraib
Cult behaviour & obedience
Race riots
Football violence
Key issues in Social Psychology: destructive
obedience, cult behaviour, football/race related
violence
 All the above issues can be explained
using ideas, concepts & research from
social psychology
 But how?
 What ideas & research can be used?
 Agency theory
 Agentic V. autonomous states
 Moral strain
 Charismatic
leadership
&
reward/punishment (coercive power)
 Social Identity Theory – in-group
loyalty/out-group hostility
 Social categorisation, identification,
comparison
 Self-esteem
 Soldiers who commit war crimes,
football crowd violence & cult behaviour
can be explained by:
 Agency theory, they become the agents
of
those
with
perceived
authority/status/power; thus losing their
own autonomy. Moral strain is the
result, denying personal responsibility is
a coping mechanism.
 Some people may have charismatic, or
reward, or coercive, or expert power
which gives them an ability to get others
to do what they want them to. They may
have a combination of these types of
social power, thus eliciting obedience.
Key issues in Social Psychology: destructive
obedience, cult behaviour, football/race related
violence














All the above issues can be explained using ideas,
concepts & research from social psychology
But how?
What ideas & research can be used?
Agency theory
Agentic V. autonomous states
Moral strain
Charismatic leadership & reward/punishment
(coercive power) simplistic, emotional language
Social Identity Theory – in-group loyalty/outgroup hostility
Social categorisation, identification, comparison
Self-esteem
De-individuation (Zimbardo-guards)
De-humanisation(Zimbardo-prisoners, Milgram)
Emotional contagion
Conformity/obedience/power of social situation.
 Being in the army, or a cult, or member
of a football gang, or ethnic minority
group can generate a strong sense of ingroup loyalty & out-group hostility.
 A strong sense of in-group loyalty is
often fostered by the army, cults etc., and
the processes of social categorisation,
identification & comparison can increase
self-esteem.
 Any challenge to the in-group is also a
challenge to members’ self-esteem and
can be strongly resented.
 However, the level of in-group loyalty &
out-group hostility varies between
individual members of a group &
whether the out-group is perceived as
being in the majority or minority.
Key issues in Social Psychology: destructive
obedience, cult behaviour, football/race related
violence (continued)
 Not everyone becomes an agent of perceived authority & enters into an agentic
state – some people disobey despite the social, & sometimes physical, costs to
themselves. Agency theory struggles to explain this
 The concept of agency theory may provide an excuse for some people to
commit horrible acts & atrocities – they were psychologically powerless to act
any differently, but is this really case?
 Some conflict between groups is about competition for resources (Realistic
Conflict Theory); or is the result of negative propaganda about out-groups & so
is not a naturally occurring social phenomena.