Social Influence: Group Dynamics

Group Dynamics

So far, we have described social influence as though it were a “one-way street.” The group presses you toward conformity. A salesperson leads you toward a concession. But, of course, social interactions often involve mutual influence—with each person in the group having an impact on every other person in the group. The study of this sort of interaction is the study of group dynamics.

MERE PRESENCE EFFECTS

More than a century ago, Triplett noticed that cyclists performed better when they competed against others than when they competed against the clock (Triplett, 1898; Figure 13.20). This observation inspired him to conduct one of social psychology’s first experiments, in which he told children to turn a fishing reel as quickly as they could, either alone or with another child. Triplett found that children turned the reel more quickly when they were with others than when they were alone.

This finding was subsequently replicated many times, and initial results suggested that this mere presence effect was uniformly beneficial. For example, when working alongside others who are engaged in the same task, people learn simple mazes more

quickly and perform more multiplication problems in the same period of time— illustrations of a pattern known as social facilitation (F. H. Allport, 1920). Other stud-ies, however, show that the presence of others can sometimes hinder rather than help—an effect called social inhibition. While college students solve simple mazes more quickly when working with others, they are considerably slower at solving more complex mazes when others are around (Hunt & Hillery, 1973; Zajonc, 1965, 1980). How can we reconcile such divergent results?

Zajonc (1965) argued that the presence of other people increases our level of bod-ily arousal, which strengthens the tendency to perform highly dominant responses— the ones that seem to come automatically. When the dominant response is also the correct one, as in performing simple motor skills or learning simple mazes, social presence should help. But when the task gets harder, as in the case of complex mazes, then the dominant response is often incorrect. As a result, performance gets worse when others watch, for in that case the dominant response (enhanced by increased arousal) inhibits the less dominant but correct reaction.

Evidence supporting this view comes from a wide array of studies. In one study, researchers observed pool players in a college union building. When good players competed in front of an audience of four others, their accuracy rose from 71 to 80%. But when poor players were observed, their accuracy dropped from 35 to 25% (Michaels, Bloomel, Brocato, Linkous, & Rowe, 1982). Similar effects can be observed even in organisms very different from humans. In one study, cockroaches learned to escape from a bright light by running down a simple alley or by learning a maze. Some performed alone; others ran in pairs. When in the alley, the cockroaches performed better in pairs than alone—for this simple task, the dominant response was appro-priate. In the maze, however, they performed better alone; for this more complex task, the dominant response was incorrect and inappropriate (Zajonc, Heingartner, & Herman, 1969).

SOCIAL LOAFING

The studies just described are concerned with people working independently of each other or people working in the presence of an audience. But what about people work-ing together—such as a committee working on an administrative project, or a group of students working on a class project? In cases like these, everyone is a performer and an audience member, because every member of the group is contributing to the overall product. Likewise, everyone is able to see and perhaps evaluate others’ contributions. How do group members influence each other in this setting?

In this situation, we are likely to observe a phenomenon known as social loafing (Latané, 1981), a pattern in which individuals working together in a group generate less total effort than they would if each worked alone. In one study, individual men were asked to pull on a rope; the average force for these pulls was 139 pounds. When groups of eight pulled together, the average was 546 pounds—only about 68 pounds per per-son. In another study, students were asked to clap and cheer as loudly as they could, sometimes alone, sometimes in groups of two, four, or six. Here, too, the results showed social loafing. Each person cheered and clapped less vigorously the greater the number of others she was with (Latané, Williams, & Harkins, 1979). This general finding that individuals work less hard in groups has now been replicated many times in the United States, India, and China (Karau & Williams, 1993).

Why do individuals work less hard in groups? One reason is that they may feel less accountable and therefore are less motivated to try as hard as they can. Another reason is that they may think that their contribution is not crucial to group success (Harkins Szymanski, 1989). There is an old adage: “Many hands make light work.” The trou-ble is that they do not always make it as light as they could.

DEINDIVIDUATION

Apparently, then, the presence of others can influence us in multiple ways—in some circumstances facilitating our behavior, and in other circumstances inhibiting us. But the presence of others can also dramatically change how we act. In a riot or lynch mob, for example, people express aggression with a viciousness that would be inconceivable if they acted in isolation. A crowd that gathers to watch a disturbed person on a ledge atop a tall building often taunts the would-be suicide, urging him to jump. What does being in a crowd do to people to make them act so differently from their everyday selves?

One perspective on these questions describes crowd behavior as a kind of mass mad-ness. This view was first offered by Le Bon (1841–1931), a French social psychologist who contended that people in crowds become wild, stupid, and irrational and give vent to primitive impulses. He believed their emotion spreads by a sort of contagion, rising to an ever-higher pitch as more and more crowd members become affected. Thus, fear becomes terror, hostility turns into murderous rage, and each crowd member becomes a barbarian—“a grain of sand among other grains of sand, which the wind stirs up at will” (Le Bon, 1895).

Many modern psychologists believe that although Le Bon may have overstated his case, his claims contain an important truth. To them, the key to crowd behavior is dein-dividuation, a state in which an individual in a group loses awareness of herself as aseparate individual (Figure 13.21). This state is more likely to occur when there is a high level of arousal and anonymity—just as would be the case in a large and angry crowd or a large and fearful gathering. Deindividuation tends to release impulsive actions that are normally under restraint, and what the impulses are depends on the group and the situation. In a carnival, the (masked) revelers may join in wild orgies; in a lynch mob, the group members may torture or kill (Diener, 1979; Festinger, Pepitone, & Newcomb, 1952; Zimbardo, 1969).

To study deindividuation, one investigation had college students wear identical robes and hoods that made it impossible to identify them. Once in these hoods— which, not coincidentally, looked just like Ku Klux Klan robes—the students were asked to deliver an electric shock to another person; they delivered twice as much shock as those not wearing the robes (Zimbardo, 1970). In the robes, it seemed, the students felt free to “play the part”—and in this case the result was ugly. Other stud-ies, though, reveal the good that can be produced by deindividua-

tion. In a different experiment, students were asked to wear nurses’ uniforms rather than KKK costumes; dressed in this way, students delivered less shock than a group without costumes (R . D. Johnson & Downing, 1979). Thus, deindividuation by itself is not bad—it simply makes it easy for us to give in to the impulses cued by the sit-uation, and the nature of those impulses depends on the circum-stances.

Notice also that deindividuation can happen in several different ways. Being in a large crowd produces deindividuation; this is part of why mobs act as they do. Wearing a mask can also produce dein-dividuation, largely because of the anonymity it provides. But dein-dividuation can also result merely from someone’s wearing a uniform and having an assigned role—in essence, he “becomes” the

role. This third factor was plainly revealed in a classic study known as the StanfordPrison Experiment, in which Philip Zimbardo transformed the basement of StanfordUniversity’s psychology department into a mock prison and randomly assigned male undergraduate participants to the role of either guards or prisoners (Figure 13.22; Haney, Banks, & Zimbardo, 1973; Zimbardo, 1973; see also Haney & Zimbardo, 1998). Guards and prisoners wore uniforms appropriate to their roles, and prisoners were called by assigned numbers instead of their names. The experimenter gave the partic-ipants few instructions, and placed few constraints on their behavior. What rapidly evolved was a set of behaviors remarkably similar to those sometimes observed in actual prisons—with cruelty, inhumane treatment, and massive disrespect evident in all the participants. The behaviors observed were sufficiently awful that Zimbardo ended his study after only 6 days, before things got really out of hand, rather than let-ting it run for 2 weeks, as was originally planned.

Sadly, the powerful effects of deindividuation and stepping into a role extend well beyond the confines of the laboratory setting. As we saw, one now-infamous real-world example is the abusive behavior exhibited by military personnel at Abu Ghraib prison. In the face of worldwide condemnation, Americans of all stripes struggled to understand how their own countrymen and countrywomen could behave in such an unconscionable fashion. Mindful of the lessons of the Stanford Prison Experiment, though, Zimbardo and other social psychologists have argued that powerful social forces were at work here that included—among others— the power of deindividuation through reducing people to their roles (Zimbardo, 2007). As a result, the situation itself may have done far more to create these abuses than the personal qualities of any of the soldiers involved.

GROUP POLARIZATION

Being in a group doesn’t just influence our behavior; it also influences our thoughts— and often for the worse. For example, consider the phenomenon of group polarization, a tendency for group decisions to be more extreme than the decisions that would have been made by any of the members on their own. This pattern arises in many different group contexts, such as when juries decide how much money to award a plaintiff at the end of a lawsuit.

Often the polarization takes the form of a so-called risky shift, in which groups appear more willing to take risks, or more willing to take an extreme stance, than the group members would be individually (Bennett, Lindskold, & Bennett, 1973; C. P. Morgan & Aram, 1975; Schroeder, 1973). However, group polarization can also take the opposite form. If the group members are slightly cautious to begin with or slightly con-servative in their choices, then these tendencies are magnified, and the group’s decision will end up appreciably more cautious than the decisions that would have been made by the individuals alone (Levine & Moreland, 1998; Moscovici & Zavalloni, 1969).

What produces group polarization? One factor is the simple point that, during a dis-cussion, individuals often state, restate, and restate again what their views are, which helps to strengthen their commitment to these views (Brauer, Judd, & Gliner, 1995). Another factor involves the sort of confirmation bias that we discussed— the fact that people tend to pay more attention to, and more readily accept, information that confirms their views, in comparison to their (relatively hostile) scrutiny of information that challenges their views. How does this shape a group discussion? In the discussion, people are likely to hear sentiments on both sides of an issue. Owing to confirmation bias, the arguments that support their view are likely to seem clear, persuasive, and well informed. Opposing arguments, however, will seem weak and ambiguous. This allows people to conclude that the arguments favoring their view are strong, while the counterarguments are weak, which simply strengthens their commit-ment to their own prior opinion (for the classic example of this pattern, see C. G. Lord, Ross, & Lepper, 1979; also Kovera, 2002).

Another factor leading to group polarization hinges on two topics we have already mentioned. On the one hand, people generally try to conform with the other mem-bers of the group, both in their behavior and in the attitudes they express. But, in addition, people in individualistic cultures want to stand out from the crowd and be judged “better than average.” How can they achieve both of these goals—conform-ing and excelling? They can take a position at the group’s “leading edge”—similar enough to the group’s position so that they have honored the demands of conform-ity, but “out in front” of the group in a way that makes them seem distinctive. Of course, the same logic applies to everyone in the group, so everyone will seek to take positions and express sentiments at the group’s leading edge. As a result, this edge will become the majority view! In this way, right at the start the group’s sentiments will be sharpened and made a step or two more extreme—exactly the pattern of group polarization.

GROUP THINK

Group decision making also reveals a pattern dubbed groupthink (Janis, 1982). This pattern is particularly likely when the group is highly cohesive—such as a group of friends or people who have worked together for many years—and when the group is facing some external threat and is closed to outside information or opinions. In this setting, there is a strong tendency for group members to do what they can to promote the sense of group cohesion. As a result, they downplay doubts or disagreements, celebrate the “moral” or “superior” status of the group’s arguments, stereotype enemies (“our opponents are stupid” or “evil”), markedly overestimate the likelihood of success, and dis-count or ignore risks or challenges to the group (Figure 13.23).

Arguably, groupthink caused a number of disastrous decisions, including the U.S. government’s decision to invade Cuba in the early  1960s (Janis, 1971) and the National Aeronautics and Space Administration’s decision to launch the Challenger on a cold day in 1986 despite the knowledge that one part of the space shuttle did not perform well at very cold temperatures (Moorhead, Ference, & Neck, 1991). Social psychologists are still working to understand exactly when the groupthink pattern emerges and what steps can be taken to limit the negative effects of groupthink on decision making (Kruglanski, Pierro, Mannetti, & De Grada, 2006; Packer, 2009).