Running head: The Effect of Group Decision on Individual Choice 1

Running head: The Effect of Group Decision on Individual Choice
The Effect of Group Decision on Individual Choice
Marissa Disbrow
Hanover College
PSY 344: Social Psychology
Winter 2013
1
Group Decision on Individual Choice 2
Abstract
This study was designed to examine the effects of group decision on individual choice. Participants (N =
34, 62% female, 35% male, and 3% unknown) were asked to read a selection in which they were to
imagine themselves as being a juror deciding how much a company should pay in damages. They were
then placed in one of three conditions: an estimate condition where they just read the selection then
provided an estimate, a group-estimate condition in which they were given an average then asked to
provide an estimate, and a commitment-group-estimate condition in which they first provided an
estimate, then were given the average, and finally asked to provide one last estimate. The means of the
three conditions were significantly different according to a one-way ANOVA, F(2, 30) = 4.906, p = 0.014.
However, there was no significant change for the third condition where the participants first provided
an answer, saw the average, then provided a second answer, where p = 0.233. For future research, it
would be wise to include more detail for the selection that the participants read. This will hopefully clear
up any confusion for the participants.
Group Decision on Individual Choice 3
The Effect of Group Decision on Individual Choice
What goes into making a decision? First, let’s look at Asch’s conformity study. Participants were
placed in a room with other people (the confederates) and asked to pick the comparison line which
matched the standard line. As a reminder of his results, he found that people conformed to the group
about one third of the time. However, what if we changed it up a little before, so that before the
subjects heard the opinions of the other subjects, they had to write down their own answer? Would this
protect them from the conformity effects of the majority?
One specific aspect in regards to decision making is how a group’s decision affects our own,
individual decision making. Social proof is a concept that provides people with a way to look at other’s
decisions in order to help make individual decisions. Social proof is when people infer that the behavior
of others indicates the right choice. An example of this would be in advertising. When they say 7 out of
10 college students stop buying hard copies of their textbooks. Instead, they buy their books for much
less using their Nook or Kindle. What is happening here is that the advertiser is hoping you will be
persuaded to choose to buy your books on the Nook or Kindle. This seems like the right thing to do.
Deutsch and Gerard (1955) talk about two social influences, normative and informational. Normative
social influence is defined by Deutsch and Gerard as “an influence to conform with the positive
expectations of another” while informational social influence “may be defined as an influence to accept
information obtained from another as evidence about reality” (1955).
Latané and Darley (1968) address in their article the lack of emergency when there are certain
conditions being applied to a situation. For example, if three people walk into a room which is filled with
smoke and the first two people do not act in a state of emergency, the third person is unlikely to act in a
state of emergency as well. It can be said then that the two individuals acted as evidence for the third
person, meaning the third person acted in an appropriate behavior according to the other people in the
room. This appropriate behavior shows that people may be highly influenced by others around them,
Group Decision on Individual Choice 4
which in turn can be connected back to the social proof idea. They can use others as information, which
can help them in turn to interpret the situation as one in which no emergency is present.
Viscusi, Phillips, and Kroll (2011) looked at how individuals are influenced by their groups. They
were able to observe participants making risky investment decisions. The participants in the group
condition were able to watch others make their decision and then base their own decision off of the
others. The authors’ results showed that just being able to look at what others were choosing had a
significant impact on others’ behavior; there did not need to be any communication between the
individual and the group. When it came to a person’s investment level, 58% raised their levels when
they observed what others did. They also found that “90% of the people in their majority group changed
their values from when the decision was personal, even though there was no reason to change if others’
choices have no information value” (2011). People who had to make a decision before the knowledge of
other’s responses changed their original responses even though there was no reason to change. The
change did not make the situation better or worse. They were, in fact then, conforming to the general
group knowledge.
It is not always wise to follow the majority however. Punchochar and Fox (2004) found that
groups can be overconfident. They found that students do produce higher exam accuracy scores when
they work in groups than when they work alone. However, they also found that there is an increased
confidence for the groups’ wrong answers. They question the age-old myth that “two heads are better
than one.” There are pros and cons to each side; group work could be useful, for example, when it
comes to staying productive in menial tasks; however, working in a group could inflate confidence. Their
results indicate that groups are not necessarily a good source of information even though they seem
more confident. This suggests a problem with social proof. How do we fix that problem? How do we
make people more resistant to social proof? One strategy is getting them to commit to a personal
decision before hearing the group's opinion.
Group Decision on Individual Choice 5
There is a possible solution that could reduce dependence on a group’s decision. This would be
to have the individual make an individual decision first before learning what the group’s decision is.
Deutsch and Gerard (1955) had two conditions that look at this in particular, 4 and 4a. In the 4
condition, which Deutsch and Gerard called “the self-commitment variation,” subjects would write
down on a sheet of paper their decision. They would not put their names on this paper, and at the end
of the experiment, the papers would be thrown away. Their answers would not be erased either after
receiving the others’ judgments. In the 4a condition, which the authors called “the Magic Pad selfcommitment variation,” the subjects would write their answers down on the magic pad, hear the others’
judgments, and then immediately erase their own decision. It might be thought that these two
conditions would produce about the same results. However, this is not the case. The Magic Pad selfcommitment, 4a, was less effective in reducing socially influenced errors whereas the self-commitment
variation, 4, was more effective. The authors believe that the first self-commitment variation made the
subjects feel as though others could still what they had written. The subjects in the Magic Pad variation
however were wiping their answers away before anyone could see them.
There have been many studies looking at how group decisions can affect individual choice. The
main question I am directing my attention to however is if conformity plays a role. Does making a
commitment to a decision before being involved with a group reduce a person’s susceptibility to
influence from a group’s information? My experiment will look for any differences that might arise from
three possible conditions, an estimate condition, a group-estimate condition, and a commitment-groupestimate condition, to see if a group’s decision actually sways an individual’s choice. The first two
conditions will be compared to see if social proof has an effect, so are people influenced by the opinions
of others. The second and third condition will be compared to see if getting people to make a
commitment will reduce the effect of social proof. This will be different from Deutsch and Gerard’s 4
Group Decision on Individual Choice 6
and 4a conditions because the subjects will write down their responses and their initial decision will be
recorded.
Method
Participants
Participants were acquired through a random selection obtained through the student email
registry of a small liberal arts college in Indiana. Three hundred students received emails asking for their
participation in this study. Of the three hundred students solicited, 34 responded. Of our 34
participants, 35% were male, 62% were female, and 3% were unknown. 91% were Caucasian, 6% were
other, and 3% were unknown. The participants ranged from 18 to 23 years of age, with a mean age of
20.9.
Materials & Procedure
The survey was sent out via email to Hanover College students and the data collection was
conducted using online questionnaires. After clicking on the hyper link in the email, students were taken
to the survey site. The students read the informed consent form. Informed consent was assumed by
subjects clicking a “next” link at the end of the informed consent form. The next step the participants
came upon was filling out a brief demographic survey, consisting of age, gender, school year, and
ethnicity.
After clicking next upon completion of the demographics, they then started the actual survey.
There were three conditions in which participants could be placed. Participants were assigned to the
three conditions by random assignment. An excel sheet with Hanover students’ email addresses
randomized the order in which the students were placed into each condition. The first one hundred
went into the first condition, the second one hundred into the second condition, and the last one
hundred into the final condition. All conditions first read a selection describing a fictional legal case
Group Decision on Individual Choice 7
where a man was harmed by a piece of equipment that malfunctioned. The man sued the company but
the company stated it was not their fault a piece of their equipment malfunctioned. However, the court
decided that the company was responsible for the injury and should have to compensate the victim. The
participants read the following selection:
Jim was injured by a piece of Carwright’s farming equipment that malfunctioned. His injury will
impair his arm movement for the rest of his life. Jim took action against the Carwright Company
and the court decided that the company was responsible for the injury. The company now has
to provide compensation for Jim. Imagine you are on the jury in this legal case.
The independent variable in the study is commitment to a decision. All responses were collected
using an open-ended questionnaire. The estimate condition had the participants declaring how much
money the company should owe the man by typing in the amount and not seeing what a group had to
say. The second and third conditions were given a group decision. The group decision was said to be an
average from twenty of the participant’s peers. The group-estimate condition had the participants see
an average of group responses, which looked like this: “Please take a quick look at the average provided
from twenty of your peers: AVERAGE = $450,000,” and then provide an estimate. The commitmentgroup-estimate typed their personal estimate, saw an average of group responses, and then provided
another estimate. After the survey was filled out and submitted, the participants were then taken to the
debriefing page.
Results
My first hypothesis was testing the first and second conditions to see if social proof had an
effect. My second hypothesis was testing the second and third conditions to see if getting people to
make a commitment will reduce the effect of social proof. The averages of each condition were as
follows: condition one had an average of $250,909. Condition two had an average of $328,571.
Group Decision on Individual Choice 8
Condition three’s average before hearing the average was $137,875, and after hearing the average of
other’s decision, the average for condition three was $226,250. I ran a one-way ANOVA to test whether
these means were all the same. The means of the three conditions were significantly different according
to a one-way ANOVA, F(2, 30) = 4.906, p = 0.014. However, there was no significant change for the third
condition where the participants first provided an answer, saw the average, then provided a second
answer, where p = 0.233.
Discussion
Conclusion
While there was significance between the means of the three conditions, there was no
significance when it came to comparing conditions one and two. This then leads us to believe that
people really are not influenced by the opinions of others.
Unfortunately, it also did not appear that making an individual decision increased the chance for
participants to stick with their decision even after seeing a group decision. Even though the average for
condition three increased, it was nowhere near the average they saw in the survey, which was $450,000.
We can then say again that social proofing had no effect. However, we cannot say indefinitely that
people are sticking with their individual decision because of the limitations that were in this study.
Limitations and Future Research
Many participants stated that they needed more information in order to make the best decision.
For example, one participant wrote:
I don't have concerns over the survey itself, or instructions that were provided, but I am left
with questions over the selection. I found myself asking what state of mind the worker was in
when he got hurt. All I know is that he was injured for the rest of his life, and the company was
stuck giving him a settlement, thus I gave him the full amount of money. I think if I had been
given more information, my opinion would have changed.
Group Decision on Individual Choice 9
There were many variations of this, stating that they would want to know more information on what
exactly happened to the man, how injured he was, would he be able to work in another field or was he
completely done with any and all working capabilities. In future research, the reading selection would
need to be more fleshed out. If this had happened, then maybe there could have been a stronger
difference.
There were also an unequal number of participants in each condition. There were twelve
participants in the first condition (estimate condition), fourteen participants in the second condition
(group-estimate condition), and eight participants in the third condition (commitment-group-estimate
condition). Perhaps if the subjects had been more evenly distributed, there could have been more
significance found in the results. To go along with this, if there had also been more participants, again
the differences may have been more significant or even really proven the point that there is no
significance.
Group Decision on Individual Choice 10
References
Deutsch, M., & Gerard, H. B. (1955). A study of normative and informational social influences upon
individual judgment. 629-636.
Latané, B., & Darley, J. M. (1968). Group inhibition of bystander intervention in emergencies. Journal of
Personality and Social Psychology, 10(3), 215-221.
MacCoun, R. J. (2012). The burden of social proof: Shared thresholds and social influence. Psychological
Review, 119(2), 345-372. doi:10.1037/a0027121
Puncochar, J. M., & Fox, P. W. (2004). Confidence in individual and group decision making: When 'two
heads' are worse than one. Journal of Educational Psychology, 96(3), 582-591.
doi:10.1037/0022-0663.96.3.582
Viscusi, W., Phillips, O. R., & Kroll, S. (2011). Risky investment decisions: How are individuals influenced
by their groups?. Journal of Risk and Uncertainty, 43(2), 81-106. doi:10.1007/s11166-011-91233