SHM Social identity theory

Inquiry questions
* Why do people discriminate against others?
* Why do people feel a need to belong to their group?
* How does group membership affect our behaviour?

Key learning:

® Social identity theory (SIT) is better understood when placed in the context
of previously existing theories of intergroup conflict: the aggressive
personality theory and the realistic group conflict theory.
® Social identity theory is based on six key claims (examined in this section)
and supported by studies using the minimal group paradigm. The main
idea of the theory is that the mere perception of belonging to a group is
sufficient to trigger out-group discrimination and in-group favouritism.
* Tajfel et al.’s classic experiments used the minimal group paradigm to
demonstrate that people tend to maximize their in-group’s profit at the
expense of the out-group. They even tend to sacrifice their in-group’s profit
for the sake of positive group distinctiveness.
* Tajfel's theory has a large explanatory power. For example, it explains why
conflict and discrimination can occur even in the absence of competition
over resources between the two groups. It has also contributed to our
understanding of stereotypes. It can be useful for designing strategies for
overcoming inter-group conflict.
® Onthe other hand, Tajfel’s theory has limitations: the self-esteem
hypothesis is debated, the positive-negative asymmetry continuum has
been discovered, and it has been argued that the theory’s explanatory
power is better than its predictive power.

Key terms: realistic group conflict theory, social identity theory, social
categorization, out-group discrimination, in-group favouritizm, self-esteem,
distinctiveness, social comparison, minimal group paradigm

In a wider context

Social identity theory (SIT) was developed as a theory of intergroup conflict: its
aim was to explain why conflict and discrimination occur.

Social identity theory (SIT) is better understood when placed in the context of
previously existing explanations of intergroup conflict, such as:

* Theories that stressed intrapersonal reasons (i.e., factors within the

personality). According to these theories, group conflict emerges because
groups are composed of people with “aggressive”, “violent”, or “evil”
personalities. Such theories ignored social variables.

335
 5 Human relationships

* Realistic group conflict theory (originated in the studies of Sherif, 1956).

This theory suggested that intergroup conflict is caused by real conflict
between group interests (competition over limited resources). In other
words, groups conflict with each other because they want to possess
resources (food, territory, energy, romantic partners) that are not abundant
enough for everyone to possess them.

What is social identity theory?

&y Social identity theory

Social identity theory (Tajfel and Turner, 1979) is based on six key claims:
1. Competition over resources is not always necessary
for the development
of conflict between groups.
2. The mere perception of belonging to two groups (social
categorization) is sufficient
to trigger out-group discrimination
and in-group favouritism.
3. Social categorization provides individuals with a means of building
their social identity. People define themselves in social terms, in terms
of being similar to or different from, “better” or “worse” than members
of the other groups.
4. Individuals strive to achieve a positive social identity because it increases
their self-esteem.
5. Positive social identity is based on distinctiveness: the in-group must be
perceived as positively different from (“better than”) certain out-groups.
This involves the process of social comparison.
6. When social identity is not positive, individuals will try to either leave the
group or make the existing group more positively distinct.
Some of these claims have been supported by research studies using
the so-called minimal group paradigm. In this experimental paradigm,
participants are randomly classified as members of two groups on the basis of
avery trivial criterion (like the letter their name starts with or even based on a
coin flip). The groups are purely cognitive as there is no objective reason for
the group members to compete.

ltis a good idea to thoroughly analyse one of Tajfel et al.’s (1971) studies
presented in this section to understand what the theory is based on, as well
as its inherent limitations. This will help you apply your knowledge of SIT to
any unseen situation that can be presented in Paper 1B.
For Paper 1A purposes, however, you will be limited in time, soitis a good
idea to know about a simplified real-life example to illustrate the six claims.
Alternatively, practise explaining one of Tajfel’s studies in justa few sentences.

Henry Tajfel (1919-82) was a social psychologist from Poland who was best
known for his pioneering work on social identity, prejudice, and discrimination.
He played the key role in creating social identity theory.

336
 5.1 Social identity theory

There s a lot to explore in this theory because it is a construct that has been
shaped over decades in multiple research studies. However, let us start with
the classic experiments in which the theory was born. These experiments do
not support all six claims outlined above, but they certainly shed light on three
claims: 1, 2, and 5.

Tajfel et al. (1971): minimal group

experiment 1
The main theoretical claim that was tested in this experiment was claim 2. This
is that the mere perception of belonging to two different groups is sufficient to
trigger in-group favouritism and out-group discrimination, no matter how trivial
the basis of group membership is.

Participants in this study were 64 male students from a school in Bristol (England),
aged 14-15. They were tested in eight groups of eight participants.

Each group of eight was further split in half according to the principles of the
minimal group paradigm. This means that the difference between the groups was
not significant and only important in participants’ minds but not in reality.

Here is how it was achieved in practice. At stage 1 (categorization), participants

were rapidly shown 40 slides with clusters of dots on a screen and requested to
estimate the number of dots. Participants thought that estimating the number
of dots on the screen was the main point in that experiment. However, after the
slideshow was finished, the experimenter told participants that there was another
study going on (as in, “since you are already here, why don't you participate in
the other study too?”). Participants were told that for convenience, they would be
divided into groups based on the previous task—a group of four “overestimators”
and a group offour “underestimators”.

At stage 2 (distribution of rewards), participants were taken to another room one by

one where they worked on their own in separate cubicles. They had to distribute
rewards and penalties in real money to others. For this, they were given a booklet
with 18 matrices and required to choose one of the options for the distribution of
rewards and penalties. Table 5.1 shows an example of such a matrix.

You are in the group of: overestimators

Rewards and penalties for:

Member # 2 of your group
Member # 4 of the other group

A Table5.1 Example matrix. The numbers on top (1-14) represent the choice that the
participant is required to make. Based on: Tajfel et al. (1971)

In this matrix, as you move from left to right, the reward of your own group
member increases, but the reward of the member of the other group decreases.
Some matrices in the booklet involved two members of the different groups (as in
the example given in Table 5.1). In some matrices, the choice was between two
anonymous members of the in-group, and in some it was two members of the out-
group. In the example above, the choice that gives maximum profit to an in-group
member is 14, and the point of maximum fairness is 7.5 (that is, between 7 and 8).
 5 Human relationships

( Jaeiiey S
Participants always allotted money to others and never to themselves. All
participants were anonymous because names were replaced by codes. They
were told that at the end of the task they would be given whatever amount of
Complete the following steps
money others had allocated to them.
individually, then get together in
a small group and compare your Results of this study revealed that the mean choice in the different-group matrices
answers. was 9.2 (on the scale from 1 to 14 as shown in the example in Table 5.1). The
® How were the constructs of choices in the same-group matrices (two members of the in-group or two members
“in-group discrimination” of the out-group) were more closely clustered around the point of fairness (7.5).
and “out-group favouritism” These results indicate that participants demonstrate in-group favouritism and out-
operationalized in this study? group discrimination when categorized into groups based on trivial criteria.

* How do we know that this study

employed the minimal group
paradigm?
Chat with Al
* Why was the matrix designed Social identity theory may be complex to explain, especially if you are writing
this way? an answer to a question in Paper 1A. Therefore, it is important to explain things
concisely: briefly but without losing depth and clarity.
ATL . .
Research, Communication Practise writing a response to the following question: “Explain social identity
theory using one example”. Follow your comfortable pace and style of
writing, but time yourself. Count the number of words in your response.

Your next task is to condense your answer to a maximum of 300 words (or
however much you can write comfortably in 10 minutes). This is where Al can
help. Try making your answer shorter yourself and at the same time ask your
favourite generative Al to make it shorter for you. Compare your approach to
“shortening” your response to that of Al.

What can you learn from this comparison?

Tajfel et al. (1971): minimal group

experiment 2
To deepen the findings and to further test the claims of social identity theory,
Tajfel et al. (1971) conducted another experiment published in the same paper.
This time the key claim that was being investigated was claim 5 (suggesting that
positive social identity is based on distinctiveness).

A Figure 5.1 Kandinsky's painting

338
 5.1 Social identity theory

Participants in this experiment were 48 boys of the same age from the same
school. The procedure was largely the same as in the first study, but the way
boys were categorized into groups and the matrices used in the booklets were
different.

At stage 1 (categorization), participants were shown slides with pairs of abstract

paintings—one by Kandinsky and one by Klee—and asked to choaose the one they
liked better. They were told that they would be split into groups based on whether
they preferred Klee or Kandinsky, but in reality, the allocation was random.

The matrices at stage 2 (distribution of rewards) were manipulated in such a way

that three different contradictory strategies could be used:

1. maximum joint payoff (MJP): choosing the option that provides the highest
total number of points no matter the group membership

2. maximum in-group payoff (MIP): the best outcome to your own group,
regardless of what the other group’s outcome would be

3. maximum difference in favour of the in-group (MD): choosing the option that
maximizes the difference between the payoffs for the members of the two
groups.

Table 5.2 is an example ofa matrix that was used in the study.

Rewards and penalties for: 12|3|a|s|e|7]|8]|9]w]|n]|n

Member # XYZ of your own group (Klee) 7 8 9 (10N 1213 (14 (1516 17|18 |19
Member # XYZ of the other group (Kandinsky) [l 3 5 79| N |[13[15]17 19| 21| 23|25

A Table 5.2 Example matrix. The numbers on top (1-13) represent the choice that the
participant has to make. Based on: Tajfel et al. (1971)

In this matrix, as you move from left to right, MJP and MIP are maximized but
MD is minimized. If you care about the biggest total reward and do not care @
about group membership, you will choose option 13. The same is true for MIP.
However, if you mostly care about your group being better than the other group Why was it necessary for Tajfel
(more positively distinct from it) and you are even ready to sacrifice your group'’s etal. to conduct this additional
profit for the sake of this positive distinctiveness, then you will choose options experiment?
closer to the left side of the matrix. In option 1, your group performs seven times -
better than the other group. Communication, Thinking

Results of this experiment showed that participants demonstrated a clear

preference for the MD strategy—that they preferred their group to get less
money if this meant that their group would compare more favourably to the
other group. Therefore, the study supports the idea that subjects sacrifice
group and personal gain to achieve favourable intergroup differences—the
idea of positive distinctiveness (claim 5 in the series of claims that we used to
outline SIT).

Evaluation of Tajfel’s experiments

A major strength of procedure in Tajfel et al.’s (1971) studies was the high level
of control over confounding variables. The experiments were designed very
carefully to ensure that the mere perception of belonging to two different groups
is the only thing that is actually different. Also note the sophistication and the
339
5 Human relationships

precision with which the variables were operationalized: the matrices were
carefully designed and used in a very quantifiable way to measure in-group
favouritism and out-group discrimination.

It is commonly noted that Tajfel’s studies were conducted with only a very limited
sample of participants, so generalizability is questionable. However, it is worth
noting that Tajfel’s original studies triggered a very large number of follow-ups
and replications from other authors. The same hypotheses were tested in multiple
variations of the minimal group paradigm, with multiple populations.

Another argument is that conditions in the study were quite artificial and did
not closely resemble participants’ normal everyday activities—this may limit
ecological validity of the experiments. For example, teenage boys may be
naturally competitive, and the matrices used in the study involved forced
choice that was suggestive of competition. This may have triggered demand
characteristics (i.e., the boys may have felt that it was expected for them to
behave competitively in the experiment). However, this concern was addressed
in follow-up studies as well. Some of these studies demonstrated that social
categorization effects remain even when the groups are made to be explicitly
random and forced choice is removed from the matrices.

Evaluation of social identity theory

Tajfel’s theory has a large explanatory power. For example:

* [t explains why conflict and discrimination occur even in situations where there
is no competition over limited resources.
* |t provides an explanation for how and why people respond to status inequality.
For example, under what circumstances groups start engaging in protests.
* |t has contributed to explaining stereotyping. From the point of view of an
in-group, sterectypes may be reliable guides as to how to feel about and act
towards members of another group. Since positive group distinctiveness is,

R
according to the theory, important for our self-esteem, stereotypes too may
be said to be “important” to our self-esteem.
Social identity theory has played a role in developing evidence-based
strategies of resolving intergroup conflict. For example, researchers of SIT
Imagine you are a psychologist
have argued that if you want to reduce conflict between two groups, you
who has been invited to a
need to find a way to make their group membership less noticeable or
school that is experiencing a
important. You can introduce another categorization that cuts across the
problem with strong segregation
existing groups. Suppose you want to reduce conflict between boys and girls
among students (and teachers)
in a middle school class, so you introduce the idea of “houses” and divide
belonging to two different cultural
them into two mixed-gender teams. Another SIT-based strategy of conflict
communities. You are tasked
resolution that has been researched is redrawing the boundaries between
with designing a strategy to help
groups in such a way that expands the in-group to include what used to
overcome this segregation.
be the out-group. Such is the phenomenon of former enemies teaming up
What suggestions do you have against a third party.
based on the principles and
Researchers have also outlined limitations of SIT:
predictions of SIT?
* SIT attempts to explain social identity and behaviour through the concept of
= Thinking, self-esteem. Two predictions can be made based on the theory: (1) your self-
Self-management, esteem will increase if your group is judged to be better than other groups;
Communication and (2) people from low-status groups whose self-esteem is initially low
should seek more opportunities to present their group in a positive light in
 5.1 Social identity theory

order to restore self-esteem to a normal level (Brown, 2000). However, these

predictions have received little empirical support. Debate around the TOK
self-esteem hypothesis remains unresolved to this day.
Why is the ability to predict
* Another controversy that was uncovered is the so-called positive-negative
important for a theory? Does it
asymmetry phenomenon. It was observed that if you conduct a study that is
depend on the area of knowledge?
designed similarly to Tajfel et al.’s classic experiments but replace the matrices
For example, are there any
allocating rewards with matrices allocating punishment (such as exposure
theories in history that are able to
to aversive noise), then the effects of in-group favouritism and out-group
make predictions?
discrimination are not nearly as strong. Therefore, it seems that discrimination
and favouritism are evident in the positive domain, but not so much in the
negative domain. Why would this be the case?
* Finally, it has been argued that the explanatory power of SIT (i.e., its ability
to explain why a certain social behaviour occurs) is better than its predictive
power (i.e., the ability to predict how exactly groups will behave in given
social circumstances).

Conceptual analysis
Perspective
The starting point of SIT is of course a social factor—the fact of belonging to a
particular group. The end result is also social—it is our attitudes towards other
people, our behaviour towards other people. However, the theory is also
cognitive because it explains how we cognitively process this social fact. It is
not group membership per se that matters, but perceived group membership.
Perhapsit is this interplay between the social and the cognitive that gives SIT its
great explanatory power.

Bias
The most obvious form of bias described (and explained) by SIT is the so-called
“in-group bias”, also referred to as in-group favouritism and out-group
discrimination. It is a very general bias which accounts for other more specific
forms of bias in human behaviour and judgement. For example, evaluating the
performance of your team in a sports competition as more successful than it
actually is.

As always, there is a possibility that the theory itself could be biased. For
example, it seems like we might have overestimated originally how widely the
theory applies. As you have seen in this section, doubts have been raised that SIT
applies equally well to maximizing positive outcomes (distributing rewards) and
avoiding negative outcomes.

Measurement
Theoretical constructs of SIT such as “in-group favouritism” are quite subtle and
not easy to capture. A lot depends on how we operationalize them. From its
onset, research of social identity has been characterized by rigorous operational
definitions. For example, in Tajfel et al.’s (1971) studies “in-group favouritism” and
“out-group discrimination” were operationalized as the participant’s choice on a
matrix that outlined various possible distributions of rewards between members
of the two groups. Such an approach allows for precise measurement and easy
replication, which contributed immensely to the advancement of research in this
field over the years. The existence of a strictly defined “minimal group paradigm”
also contributed to better replicability.
341
5 Human relationships

Causality
From the start, research supporting SIT has been strictly experimental.
Conducting true experiments allows us to make cause-effect inferences with
more confidence. Carefully controlling the experimental situation by following
the principles of the minimal group paradigm ensures that we eliminate potential
confounding variables and alternative explanations. This contributes to the
theory's strong stance in science.

However, we have also considered a few challenges that SIT research is facing.
One of these challenges is the mismatch between the theory’s explanatory power
and its predictive power.

Change
Social identity theory can be used purposefully to change human behaviour
in desirable ways. Think about school education. We constantly promote the
inclusion of children in various social groups: the school choir, extracurricular
clubs, the house system within some schools, the “group” of scientists and
academic workers, and so on. Perhaps by manipulating perceived group
membership we can prevent such negative social phenomena as prejudice
and discrimination and promote positive ones such as helping and pro-social
behaviour? Racial segregation in classrooms is another problem that could be
addressed.

Responsibility
Manipulating social identity is quite easy. We have seen from the research studies
in this section that people actually cling to an opportunity to embrace another
group membership that sets them apart from others. Think about boys in Tajfel’s
study who readily became “overestimators” or “Kandinsky lovers” and acted in a
way that was discriminatory to the opposite group. If people so readily embrace
an identity that is clearly made up and has nothing to do with their personality,
how important would it be to them to cling to an identity that they may perceive
as an expression of who they are, such as supporting a football team or being
part of a musical subculture?

If introducing new identities is so easy and its effect on behaviour is so powerful,

we must tread lightly. Any policy should be carefully analysed based on
evidence, with potential costs weighed against potential benefits.