Earlier we noted that groups are able to accomplish goals that individuals cannot accomplish on their own. These goals include maximizing performance and productivity, but another important goal of groups is to make decisions—hopefully, good decisions. Indeed, many important decisions are made by groups rather than individuals: the U.S. Senate, the United Nations, congressional committees, juries, political action committees, hiring committees, boards of directors, city councils, award committees, and so forth. Two (or more) heads are better than one, right?

330

In theory, groups have significant advantages over individuals when it comes to decision making. For instance, if four doctors are discussing a difficult medical problem, they can combine their unique knowledge and experience, consider diverse perspectives, and analyze alternative courses of action to determine which is best. Therefore, it seems like an obvious prediction that the doctors would make a better decision together than they would individually. In reality, however, the benefits of group decision making can be subverted by two psychological processes that get in the way of clear thinking: group polarization and groupthink.

Imagine that, as part of a psychology experiment, you are asked to read hypothetical scenarios describing different people making decisions. One scenario describes a man deciding between taking a new job that pays a lot but may not last (a risky alternative) or keeping his current job (a conservative alternative). After reading each scenario, you are asked which alternative you personally would choose. Next, you are asked to discuss the same scenarios with a group of participants and come to a joint decision about each scenario.

When do you think you would make riskier decisions, when thinking about the scenarios alone or discussing them with others as a group? Common sense would say that, during group discussions, people will put the brakes on each other’s extreme views or rash proposals, and as a result the group will make more conservative decisions. But when Stoner (1961) conducted a study like the one just described, he found the opposite result: Participants made riskier decisions as a group than they did on their own, a tendency that came to be known as risky shift (Cartwright, 1971).

The story gets more interesting, though. Researchers who followed up on these findings found that, for some decisions, groups did in fact take more conservative, middle-of-the-road positions than did individuals. Why do groups, compared to individuals, sometimes make riskier decisions and other times make less risky ones? Do you remember how research on social facilitation initially found one thing, then the opposite, and eventually resolved the paradox and arrived at a broader understanding? The same happened with research on group decision making.

Researchers eventually discovered that when people discuss their opinions with like-minded others, the discussion exaggerates their initial leanings, leading the group to take a more extreme position than any of the individual group members held initially. This broader phenomenon (which came to replace the risky shift) is known as group polarization, meaning that group discussion shifts group members toward an extreme position, or pole (Moscovici & Zavalloni, 1969; Myers, 1982).

This discovery reveals that group discussion amplifies the original leanings of individuals in the group. If each individual member of the group leans toward a risky alternative prior to the group discussion, they shift toward an even riskier position after group discussion. And, conversely, if each individual initially prefers a more conservative alternative, group discussion shifts them toward extreme caution (Lamm et al., 1976).

If group polarization amplifies group members’ initial leanings, then we would expect group discussion to intensify initial attitudes about a variety of topics, not just decisions about risk. Numerous studies show that this is the case: When women who were moderately feminist discussed gender issues with each other, they became strongly feminist (Myers, 1975). French students who initially liked their president and disliked Americans felt stronger in both directions after group discussion (Moscovici & Zavalloni, 1969). High school students who leaned toward little racial prejudice before a group discussion became even less prejudiced after discussing racial issues with like-minded students, whereas students who initially held somewhat prejudiced attitudes became even more prejudiced after discussing racial issues together (Myers & Bishop, 1970).

Part of what makes group polarization interesting is that it seems to contradict the research on norm formation. If you look back at our discussion of conformity in chapter 7, you’ll recall Sherif’s (1936) finding that when individuals were put together to voice their opinions about something (in that case, the movement of a single point of light), they made middle-of-the-road judgments that canceled out their own, sometimes extreme, judgments. Group polarization research seems to show the exact opposite: Groups produce more extreme decisions, sometimes even more extreme than any of the group members’ initial individual ratings. Sherif’s studies are different because there was no group discussion to arrive at some consensus. This raises the question, What happens during group discussions that shifts the group toward more extreme positions? Two theories, described next, have stood the test of time (Isenberg, 1986; Myers & Lamm, 1976).

331

Exposure to New Persuasive Arguments

The persuasive arguments theory (Burnstein & Vinokur, 1977) explains group polarization through the concept of informational influence, which occurs when you conform to others’ actions or attitudes because you believe they know something that you don’t (see chapter 7). The theory assumes that people begin with at least one good argument to support their initial opinion or attitude (e.g., Mary likes Candidate X because of his immigration policy), but that they probably have not considered all the relevant arguments (e.g., Candidate X’s environmental policy). During group discussion, group members learn new arguments from each other that reinforce the position they already preferred. As a result, the group as a whole adopts a more extreme position.

This process of polarization by persuasive arguments doesn’t even require face-to-face interaction. People adopt more extreme positions than they initially held after simply overhearing a group discussion among like-minded others (Lamm, 1967) or even after merely reading a list of arguments generated by others (Burnstein et al., 1973).

How do we know that informational social influence is at work here? In one study (Liu & Latané, 1998), participants discussed a topic as a group and then were asked two weeks later how they felt about the topic. Their individual position lined up with their group’s extreme position, not with the less extreme position that they endorsed prior to the discussion. This suggests that learning new arguments from the group changed how these individuals perceived the world, not just what they reported believing.

Trying to Be a “Better” Group Member

If we apply social comparison theory (Festinger, 1954) (see chapter 5), we can explain group polarization as the result of normative social influence, which occurs when you conform to others’ actions or attitudes to be liked (Myers et al., 1980).

When individuals get together to make a decision, they often look around to figure out where the other group members stand on the topic at hand. Once it becomes clear what position the group is leaning toward, a cycle of comparison and amplification is set in motion: One person in the group tries to compare herself favorably with other group members. She wants to be a “better” group member, so she advocates the group’s position but takes it a little further than everyone else. (“You guys seem to like this idea, but I love it!”) Seeing this, another group member tries to present himself to the group even more favorably, so he takes the group’s position even further. (“Oh yeah? I will fight tooth and nail for this idea!!”)

332

The net effect of this cycle is that the group shifts toward a more extreme position. As we would expect from this theory, group discussion is more likely to result in polarized positions when group members are highly motivated to present themselves positively to other group members (Spears et al., 1990).

Most of the examples and research findings we’ve used to discuss group decision making deal with informal group contexts and hypothetical decisions. In these contexts, where decisions do not have earth-shattering consequences and group members are not overly concerned with being “right,” it’s not surprising that processes like group polarization can get in the way of clear thinking. What’s more surprising is when smart people have formal discussions about important topics and still end up making really bad decisions that can result in disaster.

Consider this real-world example. On January 28, 1986, NASA launched the space shuttle Challenger. There was special public interest and excitement about the launch because one of the seven crew members was a private citizen and teacher, Christa McAuliffe. She was the first representative of the Teacher in Space Project, a NASA program designed to inspire students, honor teachers, and spur interest in mathematics, science, and space exploration. Consequently, students in schools around the country watched the launch on TV. However, there were warning signs that the shuttle could malfunction. Despite these warnings, NASA insisted on moving forward with the scheduled launch. The shuttle indeed broke apart 73 seconds into its flight, and all seven crew members were killed.

With so much on the line, why would smart people working together make such a bad decision? Irving Janis (1982) tried to answer this question by analyzing notoriously bad foreign policy decisions made by top U.S. officials, including the Bay of Pigs fiasco in 1961 (when President Kennedy and his inner circle launched an ill-fated attempt to overthrow the communist government of Fidel Castro in Cuba) and the decision in 1964 to escalate U.S. military involvement in Vietnam (when Congress passed the Gulf of Tonkin resolution at President Johnson’s request).

Janis concluded that these bad decisions all suffered from a common problem called groupthink, a kind of faulty group thinking that occurs when group members are so intent on preserving group harmony and cohesion that they fail to analyze a problem completely (FIGURE 9.4). Groupthink is similar to group polarization but taken to the extreme, as if the group has become of one mind, completely unchecked by diverse opinions. Group members start to focus their attention on information that supports their position and ignore information that contradicts it, they stop testing their assumptions against reality, and they stop generating new perspectives on the problem at hand. Eventually they become convinced of the absolute truth and morality of their preferred course of action, and they never stop to think what would happen if they made an error in reasoning.

Janis described groupthink using the metaphor of a syndrome that afflicts the group, and he specified several symptoms of groupthink. One hallmark symptom is suppression of dissent: When group members express doubts about the majority’s preferred position, they are harshly criticized and pressured to fall back in line with the majority view. To avoid being reprimanded or excluded, group members begin to censor themselves, meaning that they give the outward impression of agreement even though privately they think that the group is on the wrong track. This results in an illusion of unanimity: It appears that everyone is in agreement, although some group members may harbor serious misgivings. For example, when NASA engineers charged with understanding the safety parameters of equipment pointed out one of the shuttle’s mechanical flaws, they were harshly rebuked and pressured to stay silent by those overseeing the launch (Esser & Lindoerfer, 1989; Vaughan, 1996).

333

Groupthink is especially likely to occur when group members view group cohesion as more important than anything else. In contrast, if group members are less concerned with reaching a consensus or being disliked by other group members, they are more likely to appraise alternative courses of action realistically and express their doubts about the majority view. Janis identified other conditions that make groupthink more likely to occur. One is the isolation of the group from outside sources of information and dissenting voices. Another is the presence of a strong leader who makes his or her opinions and preferences known to the group at the beginning of the discussion. Knowing the leader’s views, the other members want to reinforce those views to win the leader’s approval.

335