Psychology: Social psychology: Case Study 4: Experiments in intergroup discrimination

Psychology: Social psychology: Case Study 4: Experiments in intergroup discrimination

Author: Tajfel (1970)

Key term: Intergroup discrimination

Background: Tajfel believes that we as people categorize everything - including other people. As children, we decide who we like and don't like, putting things into an order that is easier and simpler to understand and deal with. We categorize people into 'us' and 'them', 'we' and 'they' - an 'in-group' and 'out-group'. The implication of this prejudice is we develop a 'generic norm of behavior' - a way of behaving in society and toward other people, favoring people in our in-group (anything can create an in-group; race, age, gender, likes, dislikes, etc) and discriminating against those in the out-group.

Prejudice - Forming an opinion before knowing the facts 

Aim/hypothesis: 

• Study 1: Test the theory that in-group favoritism and out-group discrimination will be shown even in the categorization into minimal groups (over and under estimating number of dots)- will just merely being in different groups cause discrimination?
• Study 2: Validate study 1 using a different criteria for minimal groups (artistic preference)

Method: Laboratory experiment

Variables:

• Independent variables:
  Study 1: Two conditions -
   Groups 'over-estimators of dots' and 'under-estimators of dots'  For half the boys, they were told some boys were more accurate at estimating than others.
  Study 2: Two conditions - groups 'preference for Klee' and 'preference for Kandinsky' (artists)
  However... The boys only thought they were sorted into these two groups - but each boy is in both in-group and out-group.
• Dependent variables: Choice of points made by the boys on each matrix.

Design: Repeated measures - Each boy is in both the in-group and out-group, although they were deceived into thinking that they were in different groups depending on their estimation of dots and preference of Klee and Kandinsky.

Participants/ Sampling technique:

• Study 1: 64 boys aged 14-15 from a comprehensive school in Bristol, UK. All the boys were in the same house and knew each other.
• Study 2: 3 new groups, 16 boys in each group from the same school as Study 1 - 68 boys in total, but only 48 boys did the experiment in the end.
• Probably opportunity sampling because boys were from the same school 

Experimenters: Unknown - probably Tajfel himself.

Apparatus:

• Study 1: Booklet with 18 matrices; 6 different matrices for 3 variables: ingroup choice (ingroup-ingroup), outgroup choice (outgroup-outgroup) and intergroup choice (ingroup-outgroup). In-group was always on the top row.

Example of a matrix used in study 1. 

• Study 2: 12 slides - 6 pictures by the artist Klee, 6 pictures by the artist Kandinsky. Booklet with 4 different matrices for the same 3 variables. One major difference was these matrices were employed to allow experimenters to investigate these three variables;
   Maximum joint profit (giving the largest reward to both groups)
   Largest possible reward to ingroup (largest reward to boy of their group regardless of the reward to outgroup)
   Maximum difference (largest difference in points in favour to the ingroup)
  The outgroup was also the top row in some of the matrices.

Eg. of matrix in booklet for study 2

Example of the variables that experimenters could investigate in the matrices in study 2

No box allows equal marks to be given - the marker has to favour one or the other boy.
Matrices are scored on a scale of 1-14.
Matrices consist of 13 boxes.

Each point was worth 1/10 of a 

Controls:

• All the boys received the same instructions.
• All boys completed the booklet separately in individual cubicles - no conformity or copying.
• Boys did not know who they were allocating rewards to, each boy was given a code name - prevent bias
• Boys could not allocate points to themselves - prevent bias
• In Study 2 signatures of Klee and Kandinsky were removed from the paintings shown to the boys so they would not know they were randomly allocated.
• Study 2 allows investigation of max. diff., max joint profit and max. profit - elimination of EVs
• Some matrices in Study 2 had the outgroup as the top row - elimination of EVs

Procedure:

• Study 1 - Estimating dots:

1. The 64 boys arrived in the laboratory in groups of 8.
2. They were told the experiment was on 'visual judgement' and shown 40 slides of varying numbers of dots and told to estimate the number of dots.
3. They were then divided into groups of  'over-estimators' and 'under-estimators' but the groups were actually randomly allocated
4. Half of the boys (4 groups of 8) were told some people in the group were 'better as estimating' than others.
5. Each boy was given a booklet of the matrices to complete alone. The boys would choose a column in the matrix with two different points they wished to allocate to two boys.
6. Boys were told each point was worth 1/10 of a penny 

• Study 2 - Aesthetic preference:

1. 68 boys arrived in the laboratory in groups of 16. 
2. They were told the experiment was about 'aesthetic preference for two foreign painters' and shown 12 slides of paintings - 6 by Klee, 6 by Kandinky. Autographs of the painters were removed.
3. The boys completed a booklet of the 12 matrices for Study 2.

Data: Quantitative data - the points boys chose in each matrix were totalled; they were not asked why they allocated those points.

Results:

Study 1: 

• The results were scored on a scale of 1-14, 1 being minimum amount of points possible and 14 being the max. amount of points possible to a fellow ingroup member.
   For the ingroup-ingroup and outgroup-outgroup matrices, the average result was 7.5 out of 14.
• This shows that for same group matrices, the boys gave both boys in the matrix points that were as close as possible to being fair. 
• However, when it came to intergroup choices, the boys allocated more points to their ingroup; the average score was 9 out of 14. This difference is statistically significant.

Study 2: 

• Study 2 was to further investigate the findings in Study 1, the relative weightings given to three variables - max. diff., max. joint profit, and max. in-group profit. 
• The boys did not choose the max. joint profit rewards to both in-group and out-group or maximum in-group profit, despite them all knowing each other. 
• Instead, they chose maximum difference, meaning that although the in-group got less overall, the out-group got less than the in-group. 
• Maximumising the difference between the two groups, even if it meant less profit for the group, was more important than maximum profit.

Findings:

• Study 1: The boys showed in-group favoritism and out-group discrimination even on the basis of 'flimsy and unimportant criteria' (in this study, being an over/under-estimator of dots).
• Study 2: The boys discriminated the out-group, choosing to maximise the difference between the two groups instead of maximum profit on the basis of 'flimsy and unimportant criteria' (in this study, artistic preference)

Conclusion:

• When people are placed in a group based on a similar characteristic they share, no matter what kind of characteristic, they will favor those in their group and there will be discrimination towards people who don't have that characteristic.
• If school boys showed this trait even over a small difference, this proves that bigger differences (race, religion) can cause more inter-group conflict.
• Competition vastly increases inter-group conflict as shown in Sherif's Robbers cave experiment.

Strengths

• Laboratory experiment - High levels of control - Confidence that IV directly affects DV
  Autographs removed from paintings so boys did not know they were randomly assigned into groups
   -Each boy did his experiment individually so there was no copying and bias
   -Boys knew they were not allocating points to themselves
   -Names were not used in the booklets - boys only saw code names, not the real names of the boys they were allocating points to
• Laboratory experiment - High levels of control - Reliable, can be repeated and results should be the same 
• Quantitative data - Easy to analyze and objective - Results show clearly that boys favored the in-group from the points allocated

Weaknesses

• Laboratory experiment - Lacks ecological validity  - A laboratory where boys have to sit in a cubicle and mark matrices is not normal in real life 
• Quantitative data - We don't know why those boys allocated the points the way they did, there could be another reason besides grouping
• Reductionism - This study generalizes the world from a bunch of boys in a school in UK, maybe only teenagers in UK act like this.
  - All the boys knew each other - Strangers may behave differently in the same experiment 

Ethics

• Use of money is considered unethical
• Deception
   -Boys did not know the experiment was on inter-group discrimination - They were told the experiment was on "visual judgement" and "aesthetic preference", and the booklets were on "judgement". This is unethical because the boys were not told the real intention of the experiment.