One way to avoid this is to conduct an independent groups comparison, where one group or participants discusses the number of items (e.g. sweets in a jar) and then makes a private estimate, while a control group make an estimate without discussion.
|How many sweets? Image via https://pixabay.com|
The prediction would be that measures of dispersion such as the range would be lower in the discussion group, as everyone's estimate comes closer to a group norm.
This assignment topic is based on the theory of informational social influence - when uncertain of the correct response, people tend to change their views or behavior to be more like the group. Have a look at the following for more about this idea:
- Article on normative and informational social influence
While it's fine to mention normative influence too, be careful - if your study relates to informational social influence, that should be the focus of your introduction/background literature review.
The classic study of informational influence is the work of Jenness (1932), mentioned above, who looked for the emergence of a group norm in groups of three students:
- Summary of the Jenness study
A similar relevant study, also from the early days of social psychology, was conducted by Muzafer Sherif:
- Summary of the Sherif study
More recently, Abrams et al (1990) have suggested that the idea of informational social influence is over-simplistic, and that the change in group behavior is better explained on the basis of group membership. In their study, people were more likely to conform to the majority if they saw the others as belonging to the same social group as themselves:
- Abstract of Abrams et al research
It's important throughout to focus on studies that are relevant to the aim/hypothesis i.e. studies that link to informational influence. Other studies of conformity such as the work of Asch or Mori should only be included as part of an introductory paragraph, or (perhaps best) not at all, as they relate primarily to normative social influence.
Psychology students typically try this type of study by asking about something like the number of sweets in a jar or pieces of pasta in a packet, though there are simple ways to vary it which would give the experiment more real-world relevance. You could ask about any number-based information where the majority of people are unlikely to know the precise answer, for example, how many pupils are there in the school? What is the population of Scotland? How many pensioners are there in Glasgow? Again, you would predict that within a group the range of answers would be narrower.