Group Polarization - Major Empirical Findings

Major Empirical Findings

A number of studies have explored the effects that discussion between groups consisting of like-minded individuals has on members' prevailing opinions.

In 1970, Myers and Bishop selected groups of highly racially prejudiced students and groups of less racially prejudiced students to discuss a number of racial issues. Results demonstrated that racial prejudice decreased for already low-prejudice individuals and increased for already high-prejudice individuals after individuals engaged in their respective group discussion. Thus, their study supported the claim that discussion among like-minded individuals tends to increase and intensify pre-existing attitudes, thereby demonstrating group polarization.

In 2009, an interesting occurrence of group polarization was found in a study conducted by Luhan, Kocher, and Sutter, in which subjects played a ‘dictator game’. In this game, both individual and group decision making was observed to see how individual preferences with respect to the allocation of money between a dictator and a recipient are transformed into a team decision. Their main finding was that team decisions were more selfish and competitive, less trusting and less altruistic than individual decisions. This study therefore offers evidence of group polarization in that the actions of individuals when in a group were more extreme than when the individual acted individually.

Social Comparison Theory

In their 1978 study, Bray and Noble explored the effects of authoritarianism on juror and jury decisions and further investigated the generalizability of the group polarization for a simulated jury task. The underlying assumption of their study was that group interaction tends to enhance choice tendencies initially favored in the subject population. Thus, if predominant sentiment among individual jurors is toward a guilty (not guilty) verdict, that sentiment should be more prevalent following discussion. Their hypothesis regarding group polarization was supported in that they found that deliberations produced a shift toward more severe punishment for high authoritarians but toward more lenient punishment for low authoritarians. Thus, this study provided strong evidence for group polarization in that the groups made decisions that were more extreme than the initial inclination of its members. Mackie (1986) conducted an experiment to explore the Social Comparison Theory. In the experiment, participant’s perceptions and polarization was measured in response to information presented in one of three different conditions: in-group, unrelated group, and uncategorized individual. In accordance with the Social Comparison Theory, researchers expected participants to report greatest polarization in the in-group condition. Polarization should only occur when an individual is exposed to information from their own group. Therefore, researchers expected to see a lower degree of polarization in the unrelated group and uncategorized individual conditions. Mackie (1986) found that attitude polarization occurred only when the participants received information about a group that they considered themselves a member of; supporting the Social Comparison Theory. In their research, Ledgerwood and Chaiken (2007) primed participants with in-group, out-group, or neutral statements. Ledgerwood and Chaiken (2007) used political parties as a way of creating in-groups and out-groups. After being primed, participants were asked to rate how much they agreed with multiple statements that took on either Democratic or Republican views. Researchers hypothesized that participants primed with their political in-group and participants primed with their political out-group would show the highest levels of polarization (strongly agreeing with their in-group and strongly disagreeing with the out-group), compared to the neutral priming condition (or control group). Ledgerwood and Chaiken (2007) found that individuals who identified as Democrats and Republicans, both reported more extreme agreement with their in-group and more extreme disagreement with the out-group, in comparison to the control group, after being primed

Informational Influence Theory

One of the original studies reporting findings of informational influence was conducted by Stoner (1961). In his study, participants had to advise an imaginary person, who had to make a decision. The imaginary person was choosing from two alternatives. One option was risky but had the potential to have a very positive outcome, where the second option was less risky but had only a moderately positive outcome. Stoner (1961) found that individuals were riskier after discussing the options with the group. A follow-up study was conducted by Myers and Arenson (1972), where they concluded that group discussion does not always result in riskier decision making, however, the individuals original position becomes more extreme after group discussion. After hearing the opinions of others and learning new arguments in favor of their opinion, the individual will become more confident in their decision and their viewpoint becomes more extreme, resulting in group polarization.

In regard to the Informational Influence Theory, Hinsz and Davis (1984) found that it is not simply the sharing of information that is important; however, the amount of information and persuasiveness of the arguments mediate the level of polarization experienced. In their study, Hinsz and Davis( 1984) manipulated the number of arguments and persuasiveness of those arguments. Participants were randomly assigned to a room either by themselves or with 3 other people. Individuals in the alone condition were asked to read through a dilemma and respond. Individuals in the group condition were asked to read through a dilemma and come up with a group consensus. After the individuals and groups were done, they were given a few more minutes to look over the information again, and given the opportunity to change their responses. Researchers came up with a polarization score that was the difference between an individual’s initial response and final response. They found that the higher number of arguments and the higher level of persuasiveness significantly increased polarization. In 1977, Kaplan conducted three experiments in which jurors first heard an incriminating or exonerating trial and then discussed it by exchanging notes with "bogus discussants". The notes differed in the incriminating appearance of shared facts, the redundancy of shared facts, the number of bogus jurors, and the prediscussion judgments of bogus jurors. In instances where the incriminating/exonerating proportion of facts was the opposite of both the trial incriminating/exonerating appearance and the facts cited by the participant, postdiscussion judgment moved away from extremeness. When proportion matched the participant's, judgment became more extreme, yet less so when shared facts were redundant. Kaplan found that neither the number of bogus discussants nor their communicated judgments had an effect. These results did not demonstrate substantial evidence for normative influences, in which judgment shifts in response to normative pressure exerted by knowledge of others' positions. Rather, the result indicate that participants appear to evaluate and integrate information provided by others into their final judgment, in the same manner as they would from any source. This finding provided profound evidence for the informational influence theory explanation.