These three studies appear to show the effects of informational social influence.
Research Aims and Hypothesis
Aims
In this research I aim to discover if or how often people will conform to other (fake) answers when asked to estimate how many rubber bands (the ambiguous stimulus) are contained within a transparent plastic container in an opportunity study.
Hypothesis
My hypothesis is that people who are given a sheets with higher guesses on it will give higher estimates than the people who had the sheets with lower guesses on due to the effects of informational social influence.
I think that the ambiguity of the task group for the participants will affect the degree with which they conform e.g. if I used ten grains of rice, conformity would be much lower than if I used one million, this would be difficult to prove, however. Since participants would be using their own judgement on a small quantity, answers would be close, but not due to conformity. Even if conformity was the cause, I would not be able to tell from asking the participants; in Sherif’s study participants did not feel influenced by other participants even though they did conform.
I predict a higher mean for the group given the higher sheets, a lower mean for the group given the lower sheets and the mean for the controls to be in between the two.
I predict that the range for the high and low groups will be about the same, but that control group will have a significantly larger range than either of them.
Method
Design
My study is an experiment; I am using this type of research because I want to experience first hand what effect certain elements have on the results of the study.
Because this is an experiment, there will several types of variables present (participant variables will also be present, but those are unavoidable).
The independent variables in this study are the guesses written on the sheets before they are given to the participants.
The dependant variables in this study are the guesses written on the sheets by the participants.
Situational Variables
To avoid other factors affecting my results I need to eliminate these situational variables as far as possible.
Location
Because I am using an opportunity sample, the location is likely to vary, therefore I will always ask people in a location with similar qualities.
Distractions
I will need to find a quiet area in order to prevent participants from being distracted. If participants are distracted, they may just put any answer down, without giving their answer as much consideration as they would in a quiet environment.
Other people
Other people may act as a distraction, or the participant may conform with them instead of the guesses on the sheet.
Standardisation
To succeed in eliminating the situational variables I need to standardise the experiment.
- Give written instructions
- Ask participants alone in a quiet room
- Give participants a standard time looking at the container
Participants
I will use an opportunity sample in my study, this means that participants will be gained as and when I can find them, or when I have the opportunity to test them.
I will use 30 participants, all students or lecturers (Lecturers and students in my psychology group were only used as controls) at Worcester College of Technology.
Materials
I will use:
A pen
A transparent container full of rubber bands
20 sheets (10 with high guesses on that will be given to the High group, and 10 with low guesses on that will be given to the Low group)
The high sheets will have these numbers on: 700, 670, 800, 731, 950, 825.
The low sheets will have these numbers on: 400, 470, 550, 342, 535, 380.
Procedure
- I will ask people that I see at college individually if they will guess how many rubber bands are in a container.
- I will ask them to look at the rubber bands and then to write down an estimate of how many there are on a piece of paper. The control group will not be shown any other estimates. The piece of paper will have one of two groups of fake answers already written on it, although the participants will not be told this. One paper will have high answers written on it (given to the high group) and the other low answers (given to the low group).
- Once I have collected my data I will present it in a graph/table.
Ethical Considerations
In this experiment consent will be gained from any participants, however, it is not fully informed consent, as the participants will not be told that this is a study into conformity and that the guesses that they see on the paper are not genuine, this is deception and therefore not ethical. This is unavoidable without explaining that this is an experiment into conformity, potentially altering their behaviour, therefore defeating the entire object of the study.
To make my study as ethical as possible, I will fully inform participants of what I have done, and why I have done it once they have given their estimates in a process called debriefing.
I will then ask their permission to use their data in my study, if that permission is refused, then their data will be omitted from the study.
Results
These are the results, I will now find the mean and range for each group.
Mean
High Low Control
1000 650 450
420 350 1500
510 151 300
767 800 243
392 694 2200
+ 777 + 420 + 321
720 1000 300
650 360 200
525 320 300
900 450 400
6661 5195 8437
/ 10 / 10 / 10
Mean= 666.1 519.5 843.7
The bar chart below shows the mean estimates from the High, Low and Control groups.
I did not predict that the mean estimate of the Control group would have largest mean, however, I did think that the means for High and Low groups would differ significantly, this is clearly indicated in the chart.
Range
High Low Control
-1000 -1000 -2200
392 151 200
Range= 608 849 2000
This, second bar chart shows the ranges of the three groups.
The chart shows that the ranges of the High and Low groups differ greatly from the range of the Control group, this appears to show that the High and Low groups conformed and that the independent variable did indeed influence the dependant variable. This confirms my hypothesis.
Discussion
I have found that the mean of High group is much larger than the mean of Low group and that the ranges of the participants and the controls differ greatly, and have interpreted these results to mean that the independent variable (the “fake” estimates, which were not a present factor for the controls) affected the dependant variable (the participants estimates) thus proving my hypothesis.
These findings appear to confirm my hypothesis to an extent, as the High group did have a higher mean than the Low group, although the Control group had a far higher mean that either of the participant groups, and I did not predict that.
My second prediction was that the range for the High and Low groups will be about the same, but the Control group will have a significantly larger range than either of them. This prediction was correct.
These results seem to reinforce findings by Sherif, Asch and Jenness among others, in the sense that all three found that other peoples’ opinions affected the participants’ responses, when evaluating the ambiguous (or in the case of Asch unambiguous) stimulus.
The reason for conformity in this case is likely to be informational social influence; participants had little, or no idea how many rubber bands were in the container.
Possible improvements
If I carried out the experiment again, these are the things that I would change.
- Make both sets of answers have the same range to see if that means participant answers will have a similar range.
- I found that after answering, some participants said that they had studied GCSE psychology themselves, and were aware of the aims of the experiment, and altered their estimates so that they did not appear to be conforming.
This may mean that although there was evidence of conformity in this study, it would have been less prominent than if all the participants had been unaware of my aims during their participation.
- To prevent this, I would ask participants in the High and Low groups if they have done a psychology course before, if they have, they would not participate in my experiment.
- Because I found gathering all my participants together in one place and asking them into an empty room impractical, the study was not done this way, if I did the study again, I would organise it more rigorously.
- Because I asked students in a college to participate in this study, the sample did not represent the general public, in an ideal situation, I would use people of all ages and backgrounds.
What Do These Findings Mean In Real Life?
The nature of this experiment is quite simplistic; any real life situation where people conform will have more elements in it that may affect a person’s behaviour.
However, the concept of informational social influence may explain situations such as when people watch others in an upmarket restaurant to see which piece of cutlery they are using to eat a particular type of food and copy them so that they do not look odd because they used a fish fork to eat cake, for example.
References
Wood – Psychology First
http://en.wikipedia.org/wiki/Asch_conformity_experiments
The above website drew from the following sources:
Asch, S. E. (1951). Effects of group pressure upon the modification and distortion of judgement. In H. Guetzkow (ed.) Groups, leadership and men. Pittsburgh, PA: Carnegie Press
Asch, S. E. (1955). Opinions and social pressure. Scientific American, pp. 31-35.
Asch, S. E. (1956). Studies of independence and conformity: A minority of one against a unanimous majority. Psychological Monographs, 70 (Whole no. 416)
Bond, R., & Smith, P. (1996). Culture and conformity: A meta-analysis of studies using Asch’s (1952b, 1956) line judgment task. Psychological Bulletin, 119, 111-137.