The relationship between group size and conformity

Group attitudes are more extreme than individuals - when asked to put odds on a risky course of action, a group will opt for lower odds even if previously as individuals all had estimated the risk as higher! (Kogan and Wallach, 1964)

Introduction

Conformity is defined by David Myers as a change in behaviour or belief as a result of real or imagined group pressure. Pressure is exerted by groups which are important to the individual e.g. peers (membership groups) whose values a person admires or aspires to but may not necessarily belong to (reference groups). Deutsch and Gerald state that people conform for two reasons; normative social influence (the desire to be liked) informational social influence (the desire to be right). Turner states that people conform in order to belong to a group a process called referential informational influence whereby the have the desire to be seen as part of a group and so learn the group’s norms and behave according to those norms

Sheriff (1935) performed an experiment using a visual illusion called the autokentic effect whereby a stationary spot of light seen in an otherwise dark room appears to move. Participants were asked to judge how far they thought the light had moved. They were first tested individually their given estimates fluctuated to begin with but the eventually became constant. They were then tested in a group situation, whereby they heard the estimates of two other participants in this situation the estimates of different participants became more similar developing a group norm. Different groups as well as different individuals produced different estimates even if they were in groups to begin with. Asch stated that because the task was ambiguous it made it difficult for any conclusion about conformity to be drawn about conformity in a group situation.

Asch proposed an experiment to investigate normative social influence testing people’s tendency to agree with other people who unanimously give the wrong answer on a task where the answers are obvious and unambiguous. Using the ‘line test’ where participants were asked to judge which of the three comparison lines matched the length of the standard line. Participants were asked to state aloud which line they thought was the longest, the answers were always obvious; apart from one naïve participant all the others were confederates. In the control group (where there were no confederates) there was almost 100% accuracy with only three mistakes in 720 trails (an error rate of 0.42%).  74% of naïve participants conformed at least once, Asch found a mean conformity rate of 32% where participant agreed with the incorrect majority this might not seem very high but note that the answers were obvious.

Apart from the ambiguity of the task Asch found that factors such as group size had an effect on conformity, with conformity rates being very low (3%) with just one participant and one confederate, 14% with two confederates and rates being highest with three confederates but with bigger groups the rate of conformity went down because participant got suspicious.

Aim

The aim of the study is to do an adaptation of Asch’s experiment into conformity as a means of demonstrating how group size effects conformity. Like Asch the task that I will set will be unambiguous whereby the answer is obvious. I will use independent group design; in the first condition I will be testing the rate of conformity where there are three confederates and one participant and in the second condition where there are seven confederates and one participant thus increasing the size of the pressurising group. Due to previous research where the maximum rate of conformity was achieved with three confederates I expect to find the highest rate of conformity with the first condition because previous research has shown that participants are less likely to conform with groups bigger than three because they get suspicious for this reason I think a directional hypothesis is appropriate.

Experimental hypothesis: Participants are less likely to conform to the view of the larger group due to suspicion and the onset of demand characteristics

Null hypothesis: the group size of confederates has no effect on the rate of conformity

Alternative hypothesis: Participants will conform to the large group as they will feel intimidated

 

Level of significance: p< 0.05        

Method

Design

I will be doing a field experiment in the college café this would be appropriate because it would be easy to gain my opportunity sample and participant would be more comfortable. I will be using independent group design where I have different participants in each condition because all my participants must be naïve in each condition in order for me to achieve viable results. The independent variable will be the wrong answer that I tell the confederates to give in each the dependent variable will be whether or not participants conform. I will be collecting my data by the noting the observed answers given by the participants. I will be using a significance level of p < 0.05 because it is the standard probability used by most social scientists and it makes a good balance between type one and type two errors. I will be conducting my experiment with Paulina.

Participants

We will be using an opportunity sample of AS /A level students from a South East London 6th form college. Our sample will be heterogeneous due to the fact that we will using both male and females from different ethnic/cultural backgrounds. We will use 25 participants as well 7 confederates (3 will be involved in the first condition and all 7 will be involved in the second condition) so in total we will be using 10 participants in condition one and 15 in condition two which will allow us to get results with a high degree of accuracy. Participants were allocated to each condition by using the first 10 participants who agreed to be in the experiments in the condition one and the next 10 in condition two.

Materials/apparatus

We will be using a set of standardised instruction (see appendix).  We chose a peach flavoured drink, which was clear in appearance the flavour of the drink was very obvious as it tasted and smelled very much like peach, we also used clear plastic cups so as to not to influence the experiment in any way.      

Procedure

1. We set aside a table in the college café placing the required number of cups on the table

2. Selected confederates by using an opportunity sample, whereby we asked if they would like to take part in our psychology experiment, we then informed them of what we were intending to find out and what they had to do and that they had to follow the instructions very strictly in order not to raise the suspicions of the naïve participant

3. We then went about the café again using an opportunity sample of people who we were willing to take part in our experiment we took them to the table we had set aside and introduced ourselves and our experiment without giving away our aim and objectives, following what was outlined in our standardised.

4.  Using our standardised instructions we instructed the participants of what they had to do

5.  After each participant had a taste of the drink we asked each one what flavour they thought the drink was informing them that is was not a mixture of flavours but one definite flavour. We made sure in each condition that we asked the opinion of the naïve participant last in order to exert maximum group pressure.

6. Once the experiment was complete we then debriefed participants and asked them why they had or hadn’t conformed to the view of the majority.  

Ethical considerations

     I have to maintain a balance between the interest of the participants and value of the research. To do this I have to make sure that I keep in line with the ethical guidelines. I have to first ask my participants for their consent to take part in my experiment I will have to use presumptive consent, as they will not be able to give informed consent. My participant’s will have to be deceived in order for them to be completely naïve but they will be debriefed. They will also be informed of their right to withdraw from the experiment. By having a colleague we can insure that we are both keeping in line with the ethical guidelines.  

Results 

Below is a graph of my findings (for raw data see appendix):

Key

       Conformed

       Didn’t conform

                             

Results Table 

Conclusion 

The table and graph show the findings from each condition (for contingency table see appendix). It shows that in condition one where there were 3 confederates and one participant a great number of participants conformed compared to condition two where there were 7 confederates and 1 participant where a greater number of people didn’t conform. I will be using the chi-squared test of association because I will be testing the association between group size and conformity. I found that the chi-squared statistic (observed value) was greater (3.79) than the critical value (2.71) for 0.05% significance level at one degree of freedom, so the null hypothesis can be rejected and the directional experimental hypothesis that people are less likely to conform in large groups retained. My original hypothesis proved to be correct by a significant measure showing that my findings were not due to chance (See table and graph above.)

(For chi-square calculations see appendix)

Discussion

Explanations of findings

I found that there was a 16% conformity rate in the large group situation compared to a 70% conformity rate in the small group situation this shows a significant difference of 54% proving that people were 3 times more likely to conform to a group size of 3. The findings also showed that there was a nonconformity rate of 84% in the large group compared to a 30% nonconformity rate in the small group showing a significant decrease of 54% showing that people were 2 times more likely to not conform in a large group situation than in a small group proving my hypothesis and achieving my aims.

Relationships to background research

I found that people were more likely to conform to a group size of 3 than to a group size of 7, whilst I was debriefing the participants I asked them why they had/hadn’t conformed to the view of the other participants many of the participants in the first condition stated that they had conformed because everyone else was saying ‘orange’ so they had just gone alone with them because they wanted to fit in with the group. Like Asch’s findings this showed the clear presence of normative social influence. Yet one of the participant’s  who did not conform, stated that although he was tempted to go along with the rest of the group he knew that their answers were incorrect and so in order to be right he decided to state that the drink was in actual fact peach (which it was). Abram’s et al (1990) states that we only experience uncertainty when we disagree with those with whom we expected to agree (membership groups i.e. those we see as having similar characteristics to ourselves) this would also explain why there is a high level of conformity in the small group situation because all participants were students of similar age and the naïve participants could have seen the others as a membership group.

Normative social influence had only a small presence in the large group situation where the 4 participants who had conformed to the view of the group stated that they had felt intimidated by the fact that all the other participants had said the same answer so the just went along with them in order to fit in. Turner (1991) states that this is due to referential social influence whereby people want to be acknowledged as part of the membership group so they learn the group’s norms and behave according to those norms. Yet the presence of normative and referential informational influence was not that significant in the large group as most participants did not conform. When asked why most the nonconforming participants stated that they had become suspicious of the situation and had worked out what we were trying to do, plus it was obvious that the drink wasn’t orange flavour. Willis (1963) would state that this refusal to ‘go along with the crowd’ may be an attempt to remain independent as a matter of principal this he calls anticonformity.  

Limitations and modifications

The main advantage of using independent groups design is that it eliminates order effects, but the limitation are that we will need twice as much participants as repeated measures design in order to obtain the same amount of results, this meant it took much longer for us to collect our results and although it was extremely expensive it raised the costs as we needed quite a lots of equipment in terms of plastic cups and the drink itself which we had to ensure was the same for both conditions, also differences may be due to individual difference rather than the effect of the independent variable. Although using a field experiment allowed me to work with participants in a natural condition it meant that I had no control of the environment so anybody could have interfered. To improve our research we could have possibly used repeated measures design by testing the same participants on both condition but having sufficient time such as a week between each test, this would be a way of showing that our findings were not due to individual differences. Yet this would have been difficult as we may not be able to find the same participants, it would be very time consuming and also demand characteristics would be more likely to occur. We could have used a lab experiment which would have enabled us to have tighter control over the environment

Implications and suggestions

The implication of our research is that it provides supporting evidence for the original findings of Asch in relation to group size and conformity. Follow-up studies could look at gender and conforming in group situations and possibly it should be investigated whether our hypothesis can be proved in a natural setting increasing the level of ecological validity.    

References 

1. Abrams, D., (1990) Knowing what to think by knowing who you are: Self-categorisation and the nature of norm formation. British Journal of Social Psychology, 20, 97-119 cited in Psychology a new introduction for A level 2 edition pg 123.

2. Asch, S.E. (1951) Effect of group pressure upon the modification and distortion of judgement. In H. Guetzkow (Ed.) Groups, Leadership and Men. Pittsburgh, PA: Carnegie Press     

3. Deutcsh, M. & Gerald, H.B (1995) A study of normative influence upon individual judgement. Journal of Abnormal & social policy.

4. Myers, D.G. (1994) Exploring Social Psychology. New York: McGraw-Hill. 

5. Sheriff, M. (1935) A study of social factors in perception. Archives of Psychology, 27, whole No.187.

6. Tuner, J.C. (1991) Social influence. Milton Keynes: Open University Press

7. Willis, R.H. (1963) Two dimensions of conformity-nonconformity. Sociometry, 26, 499-513. Cited in Psychology a new introduction for A level 2 edition pg 124.