Tajfel et al (1970): The Minimal Groups Experiments

Focus on the findings if this study when evaluation the theory, and use the procedural details to evaluate the study, if the study only provides weak scientific evidence, this also weakens the theory. 

Aim: To investigate the minimal circumstances under which two groups, not in competition, with each other, and created purely for the purposes of the study, would become prejudiced towards one another? Tajfel wanted to answer the question “Is group membership alone enough to create in-group favouritism?”

Research Method: Lab experiment

Participants: 64 14 and 15 yr old British school boys

The story: A study to investigate vision

Procedure: Asked to estimate the number of dots on a screen; told that they were either an over estimator or an under estimator. In reality, they were randomly assigned to groups ?(hence minimal groups – members had nothing in common whatsover except that they had been told they were either over or under estimators). They were then told they had to assign points to pairs of boys (but not to themselves) and that the points would later be converted into money. They did not know who each boy was, the other supposed group members were unseen, and referred to by a code number, the only information available was whether the boy was an in or outgroup member.

The boys were shown matrices, or grids of numbrs like the one here: minimal-groups-matrices they had to choose a number from the top row, and this number of points would go to boy A in the pair, and the number of points below would then automatically be assigned to Boy B. Tajfel varied whether the boys were both in-group, both out-group or one of each to see how this affected their choices.

IV: Whether the boys were in or out-group members

Condition 1: Two boys both from in-group

Condition 2: Two boys both from out-group

Condition 3: One boy from in-group and one boy from out-group

DV: Number of points awarded

Hypotheses:

14-15 yr old boys will award a significantly greater number of points to boys they ?believe are in-group members than out-group members.

Results: Boys overwhelmingly allocated points on the basis of group membership regardless of the accuracy of the estimate. When allocating pairs of points to two boys where one was in-group and one was out-group, the Pps would choose the pair of scores which had the greatest difference between them , sacrificing the greatest gain for the in-group in order to simply have the greatest difference between in and out-group.

Conclusion: The mere existence of a group, which the boys did not see themselves as belong to, was enough to cause them to treat other supposed group members differently, dependent upon which group they perceived them as belong to (in or out-group).

Further information and research studies which can be used to support/refute SIT:

The minimal groups experiments have been replicated many times and therefore the evidence they provide is deemed reliable; robust!

If social comparison leads to increased self  esteem then we should expect that those taking part in minimal group study should demonstrate temporarily higher self esteem when allowed to show bias than non-categorised control Pps;

This has been supported by Lemyre and Smith (1985)
IV: could make intergroup decisions or only within group decisions;
DV: self esteem the intergroup decision makers showed higher self esteem

This said, if social comparison increases self esteem, then those with lowest self esteem should be most motivated to show bias; this is not well supported by the literature; Rubin and Hewstone (1998) did a meta analysis and found that only 13/129 studies supported this!

Replication – Dobbs and Crano (2001) majority in-group Pps had to give reasons for points allocations; in group favouritism decreased; but when minority in-group members asked to justify discrimination of majority out group members, in group favouritism increased! Much more complex than Tajfel and Turner would have it!

Individual differences in allocation of points; Platow et al (1990); minimal groups replication with wide ind. diffs; more competitive individuals showed great in group favouritism but more cooperative individuals showed much less in group favouritism

Cultural differences; think cooperative (collectivist) cultures less likely to show in-group favouritism and out-group hostility in minimal group studies; Wetherall (1982) white and Polynesian children in New Zealand; Polynesians less biased, more cooperative; cultural norms reflected

Groups in society rarely equal: numbers/status/power; dominant group has more power to oppress the out-group/minority

minority out groups can however act together and demonstrate collective action against dominant group

Collective action more likely when groups seen as impermeable (not possible to move between the groups) and when the dominant status of one group over another seen as unfair, (see Reicher and Haslam, 2006)

You should also use Sherif’s Robbers Cave (1954) study in your SIT essay; there is a whole page on this study as it is you classic study for social.

Next up, you can also use Hovland and Sears (1940) in your SIT essay; there is more detail on this study in the notes on Realistic Conflict Theory, (an alternative theory of prejudice, which you should use as comparison and shown why RCT is better than SIT for explaining certain phenomena).