="http://www.w3.org/2000/svg" viewBox="0 0 512 512">

Module 8: Group Influence

Module Overview

So far, we have seen the influence of the message on our attitudes and behavior, the power of the situation to result in conformity and this module is going to take it to the next level and examine how groups impact the individual. We will start by first defining what we mean by a group, and then why groups are so important to us. We spend a large proportion of our lives in groups. This module is structured to first examine the impact of the presence of others on our behavior. We will examine how it can increase arousal and result in social facilitation. We will look at how it can lower motivation to work on a group task and then how it can result in anonymity leading to conformity to group values over individual values, arouse us and change our individual performance, how these others can demotivate us and they can make us feel anonymous resulting in behaviors that are more in line with the group values over individual values. The second section will look at the effect of extremity in interacting groups through decision making and discussion by first, examining the concept of groupthink in decision making and second, exploring the process of group polarization during homogeneous group discussion.


Module Outline


Module Learning Outcomes

  • Define a group and clarify why groups are important to us.
  • Clarify the effects of social attention through classic social facilitation and current work looking beyond classic social facilitation.
  • Contrast social loafing and free riding.
  • Explain classic deindividuation theory and the SIDE Model.
  • Describe the work on groupthink.
  • Define group polarization.


8.1. The What and Whys of Groups


Section Learning Objectives

  • Describe what constitutes a group.
  • Discuss why groups are important to us.


8.1.1. What constitutes a Group?

Let’s start by responding to the following four descriptions. Do they represent a group?

  • 5 people waiting at a bus stop
  • People attending a worship service
  • Lady Gaga fan club
  • Students in an online course

How did you respond? What were your criteria for a group? Was it just two or more people together? Can you think of times where you were around other people, but you would not have considered yourself in a group? As we define a group, we will determine whether these four would meet the criteria to be called a group.

It may have seemed silly to have a whole section devoted to defining a group, but at this point in the textbook, you have likely noticed it’s important for us to operationally define terms, especially those that are used in our everyday language. In psychology, they don’t always have the same meaning. Interestingly, even in the field of psychology, not everyone can agree on the exact definition of a group. Some define groups with very rigid conditions that must be met. For example, groups must be stable, permanent, have a structure and the members need to feel the group is part of their identity. A more flexible definition was proposed by Shaw, Robbins & Belser (1981) & Forsyth (2010): in order to be considered a group, two or more people must be interdependent, interact and influence one another. So, let’s take another look at our four descriptions. In most of the ways we could imagine these 5 people waiting at a bus stop we should find they are not a group, but just a collection of individuals. They are not interdependent, interacting or influencing each other. You might be able to imagine a scenario where the people at this bus stop do depend on each other. Maybe they talk and get to know each other and possibly even influence each other in their daily lives. In fact, the movie Speed turned the people from a city bus into a group when a terrorist strapped a bomb to a bus that would blow up if their speed slowed to under 55 mph. These individuals needed each other to survive this trauma, they definitely interacted and influenced one another. Of course, you can see how each description depends on your perception. There are probably worship services where people are a group, but there are also probably services where people just come listen passively to the sermon and leave, never depending on each other, interacting or influencing each other. The same goes for the fan club and the online course. Depending on the specifics of the situation, it may or may not be defined as a group. In most cases, the fan club is probably not a group. And in most cases, if the online course is organized like the courses through the Washington State Online Psychology Program, it is definitely a group.


8.1.2. Why are Groups Important to Us?

For most of us, from the moment we enter the world, we are part of a group. We have a family that we are dependent on, we interact with and are influenced by. Families are crucial for our survival and successful development. As we age, we join other groups: school-related groups, those with our friends, groups for our hobbies, sports teams, dance, etc. We enter the workforce and there are more group opportunities. We have been called herd animals because of our need to belong with others. Researchers have studied this from several different theoretical perspectives. The first one we will look at is the conditioning perspective. This perspective suggests that we learn early in our lives to associate positive outcomes with group membership. As mentioned previously, our first group is our family. These people typically give us physical and psychological support. We flourish in the presence of a consistent, caring family. A large portion of developmental research supports the need for contact and love to have healthy brain development. Throughout our lives, we use groups to get food, shelter, love and friendship (Baron & Kerr, 2003).

Another theory explaining our desire to be a part of groups is Festinger’s (1954) social comparison theory. You will remember from module 3 that this theory explains how we compare ourselves to those around us to see how we fit. We have discussed at several points in the text the idea that our realities are subjective and we are searching for frameworks to better understand ourselves and those around us. Social comparison gives us that information to build those frameworks. If we are in groups, we have access to the comparative information we need to create a social reality, especially in times of ambiguous physical realities. We can also use the people in our groups through comparison to protect ourselves from inappropriate behavior and embarrassment (Baron & Kerr, 2003). This looks a lot like what we learned about when we conform for informational reasons in the previous module. We learn that you shouldn’t chew with your mouth open, you should shower and groom regularly, picking your nose in public is not appropriate. All the social norms that we follow by looking at similar others would fit here, personal space norms, norms for courtship and sex, etc.

Social identity theory (Tajfel & Turner, 1986) and self-categorization theory (Turner, 1991) can also help explain the importance of groups. Social identity theory specifically addresses how we first put ourselves into a group, then we see ourselves as a member of that group and then we want to feel good about being a member, which can sometimes lead to derogating other groups (see module 9 on prejudice). In this description, you can see that the groups we identify with become very important to how we see and evaluate ourselves, self-image and self-esteem, respectively. In other words, we are choosing to belong to groups to feel good about ourselves. Choosing to be a student is a great example of a group that you could use to increase your self-esteem and improve your self-image. You are becoming more educated and increasing your chances of employment. You may even have chosen the college you attend to further increase self-esteem. It is possible you might derogate other colleges, they aren’t as good as your school, to further increase self-esteem from being a student at your selected college. The theory of self-categorization explains how we choose and place ourselves into groups. The group must be noticeable or important to us in order for it to have an effect on the self. The groups we choose give us an idea of who we are and how we see the world. The groups define what we believe, what we should and shouldn’t do, and the customs we will follow (Baron & Kerr, 2003).

The final perspective we will discuss is one that some of you might have thought about when we started this section. One very important reason that humans have behaved as herd animals is survival. The sociobiological theory (Bowlby, 1958) explores the idea that survival is more likely when humans group together. They can protect themselves more easily against predators/enemies and they can cooperate to create more group members, find and share food, build shelters and care for the sick and injured.

8.2. The Presence of Others & Its Impact on the Individual


Section Learning Objectives

  • Clarify the effects of social attention.
  • Describe the loss of motivation from presence of groups.
  • Clarify the processes of deindividuation through the classic and alternative explanations.


8.2.1. Effects of Social Attention Classic work on social facilitation. Picture, if you will, a ballerina. You can find her in the studio most days, weeks and months of the year practicing ballet alone. She has been practicing ballet since she was three years old. She is currently practicing for an upcoming show. Based on the above information, will the audience help or hurt her performance? Is she likely to do better or worse than if she did her routine alone? To understand the answer, we need to first travel back in time.

One of the first experiments in social psychology was done by Triplett in 1898. He observed that when competitive cyclists raced against others (compared to alone) they would have faster times. He believed that the presence of others would result in a better performance than when alone. So, he tested this prediction by timing children wind a fishing line apparatus in the presence of other children or winding the same apparatus without any children present. He found support for his prediction. In the presence of others, the children did wind the fishing line faster than when they were alone. Unfortunately, the findings investigating performance in the presence of others weren’t always conclusive. Researchers found contradictory findings. Sometimes others improved performance and sometimes the opposite occurred and performance worsened compared to them performing alone. Research like this continued until 1940 and from there, the topic lay dormant for 25 years. It was resurrected by Zajonc (zy-ence) in 1965. He was able to make sense of this puzzle by bringing in another field of research. The well-established principle that arousal enhances the dominant response helped solve the mystery of all those contradictory studies. He established social facilitation theory: when we experience arousal from the presence of others, we should expect to see improved performance on easy or dominant tasks (these are things we do often) and we should expect to see decreased performance on difficult or non-dominant tasks (these are things we have never done or don’t do often) (Baron & Kerr, 2003; Blascovich, et al., 1999). Based on this, if you look back at our ballerina example, you might predict that she will perform better in the presence of an audience than alone. Ballet is her dominant response and something she does often. If she had just started learning to dance, then her first recital performance would be hindered by the audience. It would be considered a difficult task or non-dominant. Why does this happen? The presence of others increases our arousal by increasing our worry over being evaluated and hurting our reputations. The next section discusses this in greater depth. Current work on social attention. The work on social attention through the theory of social facilitation focuses on how the presence of others impacts our interpersonal behavior. How do people perform on a task in the presence of others? In recent years, research has changed focus based on findings that demonstrate that social attention can impact behavior even when the others aren’t present or capable of observing or evaluating their behavior. For example, when observers are blindfolded and wear earplugs the presence of the audience still impacts the person performing the task (Platania & Moran, 2001). There are some newer cognitive models that suggest the possibility that the influence of others could be fully automatic. This means we take other people’s thoughts, ideas and feelings and internalize them, so even when they aren’t present, we unconsciously are influenced by their possible evaluation of us (Smith & Mackie, 2016a)

How does social attention affect our behavior? One way is through an increase in our public self-awareness. The possible observation and evaluation by others result in a worry or concern over our reputation. We see this best in work on prosocial behavior (see more about this in module 11). Research finds people are more charitable in the presence of others and that the bystander effect disappears when people consider what others might think of them. The bystander effect, discussed in more detail in module 11, is the idea that when there are other people present, we are less likely to help. This occurs because our share of the responsibility is spread out among the other people there. In other words, we assume someone else will help, call 911, or stop and help the person with the flat tire. However, a great way to remove the bystander effect is to make others believe that they will be evaluated for their behavior and people will know if they helped or not and this will impact their reputation. My sister’s children are school age and she mentioned that she always wants to avoid helping for school related functions but then she is influenced by what she thinks the other mothers will think if they find out she didn’t help. We need to believe that others will evaluate our behavior and then can spread that information to others. Socially desirable behavior is driven by our concern to maintain a good reputation, which leads to successful social interactions. Strangely, even subtle cues that someone is watching (human eyes or surveillance camera) can trigger the effect of social attention. We become self-aware, worry about the evaluation that could damage our reputation. Studies with these cues saw a reduction in bicycle theft, littering and increase in desirable behaviors like donating to a charity and group cooperation (Steinmetz & Pfattheicher, 2017).


8.2.2 Presence of Others Can be Demotivating Classic studies on social loafing. Our discussion of social facilitation examined how the presence of others causes arousal, and that arousal results in a change to our individual performance. For this next section, we are going to see that when we are working in a group toward a common goal rather than for ourselves, group presence often has a demotivating effect. We will start this discussion with another one of the first experiments in social psychology. This study was conducted by Ringelmann (1913) and involved having male volunteers in various group sizes engage in a tug-of-war style rope pull. The group’s total effort was measured by a strain gauge. The larger the group, the researcher found the total effort was smaller than if they were to total the individual efforts of each group member (Kravitz & Martin, 1986). Steiner (1972) determined Ringelmann’s finding occurred for two possible reasons. The first is that when others are present, we don’t feel like we have to work as hard — a reduction in motivation. The second is that in situations like a rope pull, everyone has to coordinate their pull with everyone else. So, was everyone pulling their hardest at the same time, feet in right place, etc? This second possible reason was further explored by Ingham et. al., (1974). They had an ingenious idea to determine if coordination loss was really an issue. What if there were no other people, you just believed there were other people? So, they again used the tug-o-war style rope pull. This time though they blindfolded the participants, telling them it was to prevent distraction. They were put in the first position on the rope-pulling apparatus and told the others would be behind them in their spot. They compared participants who were pulling alone with participants who believed they were pulling with a group, but were actually pulling alone. The belief that others were present resulted in participants not pulling as hard, just like what Ringelmann found. So, the conclusion was that the presence of others was demotivating and we put in less effort, not the coordination losses that occur with group tasks.

Another early study that confirmed these results was done by Latane et al., (1979) and also had an extremely creative methodology. Have you ever been to a sporting event where the whole crowd is screaming and cheering? It can be so loud. Have you ever felt empowered to yell even louder than you normally would because no one will know that it is you? If you had to predict, in this instance, would you think you would yell louder in a group or alone? Well, I think most of us would guess that we yell loudest in a group. We would lose inhibitions and join the group. Surprisingly, this experiment showed that we are wrong. We actually yell louder when we are alone than when we believe we are with a group. In this study, once again you are blindfolded and asked to wear a pair of headphones to prevent distraction. You yell alone and then alone but hear others yelling through headphones (so you believe you are not alone). The results support previous research that when others are present, we don’t work as hard. It is from these results that Latane comes up with the term social loafing. When we are working together toward a common goal, the presence of others will have demotivating effects on us. These results have been replicated in many different countries and with many different types of tasks, including cognitive or perceptual tasks. Social loafing tends to be a little stronger in western over eastern countries, and men are somewhat more prone than women to social loaf. Those that identify more strongly with individualism and value individualistic traits will be more likely to social loaf, which is why western countries and men are a little more likely to do this. When a group is more important to your identity, your motivation won’t be impacted as much (Karau & Williams, 1993).

Why does this happen? First, it is task-specific. It doesn’t always happen just because you are in a group. There are types of tasks where this doesn’t occur and things we can do to eliminate or reduce social loafing. However, let’s first focus on when it does happen. It happens when there is no way to know what the individual group member contributions are. So, when individual performance can be recorded, we see social loafing go down or go away completely. A study tested this idea by taking the yelling methodology from Latane et al., (1979) and this time you either yell thinking that no one will know what you contribute (how loud you are yelling) or you wear a headset with a microphone that records your individual yell level (Williams et al., 1981). In this study, the results showed that when individual performances (amount each person yelled) could be recorded and identified, the social loafing goes away. Free riding. As we mentioned, the group task is a very important indicator of how the group will impact the group member’s behavior. There are group tasks where the group shares the success or failure. This shared responsibility makes it possible for everyone to contribute different amounts and the group still succeeds. In fact, as long as one member completes the task well, the group can succeed. It is this situation that will lead to free riding. This means you can decrease your effort and benefit from the efforts of others group members. In this situation, it is important to make sure that members are evaluated and identifiable. Group members will weigh the factors of amount of effort, feelings of being needed for the group to succeed, and violating social norms for mooching to determine if they free ride or not (Baron & Kerr, 2003). Reducing social loafing and free riding. If you are like me and found yourself in groups where you seemed to care the most about the outcome, then you were always worried that you would do more of the work and the other group members would just free ride or social loaf, depending on the task that was assigned. They would get credit for all your work. We don’t want to live life without groups, so then it would be helpful to know how we can increase group motivation and how we can eliminate or reduce social loafing and free riding. It is important to not use groups or teams if an individual can do the task easily. Tasks that involve or require a lot of effort and work, are the kind that should be assigned to groups or teams. Similarly, you want the task to be something the group is interested in and stimulated by. Research has also found that when group members feel close to each other, are punished for poor performance or when the group sets their own goals, they are less likely to lack motivation. It is also helpful for everyone’s work to be identifiable and easy to evaluate. This helps to prevent social loafing and free riding (Baron & Kerr, 2003).


8.2.3 Deindividuation Processes – Classic & Alternative Explanations Classic and contemporary deindividuation theory. Let’s start with a demonstration. David Dodd (1985) created this classroom demonstration to allow us to experience deindividuation from the classic theory perspective. He also hoped to show that even students and a few prisoners would respond with anti-normative behavior. He used 229 college students and 29 prisoners taking college courses in prison. The prompt that should be responded to is below:

“If you could do anything humanly possible with complete assurance that you would not be detected or held responsible, what would you do?”

How do you think your answers compare with the students and prisoners from Dodd’s work? I have used this exercise in my classroom over the years, and my student’s responses are always in line with the results from Dodd’s demonstration. The responses were categorized according to content and social desirability. Eleven content categories were established by the author: aggression, charity, academic dishonesty, crime, escapism, political activities, sexual behavior, social disruption, interpersonal spying/eavesdropping, travel and “catch-all” other categories. These categories were then rated on social desirability. Prosocial behaviors were those that benefited others, antisocial behaviors were those that resulted in harm to others or taking away of their rights, non-normative behavior was described as going against social norms but didn’t benefit or hinder others, and finally neutral for behaviors that didn’t fit into the above three definitions. Blind raters found that 36% of the behaviors were antisocial, 19% nonnormative, 36% neutral, and 9% prosocial. Interestingly, he didn’t find any differences between the prisoners and students in the kind of responses. The most common response (15%) was to rob a bank. Every time I have used this demonstration, this has been the single most common response. Between 25-75% of my students say they would rob a bank. Sometimes, they specify to help others and sometimes just the three simple words: rob a bank. In Dodd’s work, a few students said murder, rape and assassination, but in my time in a large classroom of 100-220 students where the anonymity was greater, I never had anything like that written down. The worst responses have been to slash an ex-boyfriend’s tires or beat up a cheating partner. So, how do you fit and what does this demonstration have to do with deindividuation theory?

The classic theory of deindividuation was first introduced by Festinger, Pepitone, & Newcomb (1952) looking to take a more scientific approach to the study of the crowd’s impact on the individual. They coined the term deindividuation to explain the effects of losing your own personal identity in a crowd, which allows you to engage in behaviors you wouldn’t normally do alone. These ideas were expanded on by Zimbardo (1969) and he specified that there were conditions that must be present for deindividuation to occur in a group setting. He suggested quite a few: anonymity, not feeling personally responsible, arousal, sensory overload, novel or unstructured situations, and conscious-altering substances such as drugs and alcohol could lead to deindividuated behaviors. He defined deindividuated behaviors to be those that went against what was considered appropriate. He did believe that they could be prosocial, but his primary focus was antisocial behavior. If you look back at our demonstration, you can see that the focus here is on the condition of anonymity and lack of personal responsibility, possibly higher arousal. It is exciting to think of situations where we don’t have the pressure of social norms and expectations, where we could be free of all those social restraints and the norms we follow to be accepted and belong. We won’t be sanctioned or punished for violating these norms of appropriateness (Postmes & Spears, 1998).

To better understand this original way of looking at the crowd’s impact on the individual, it is important to examine a few of the studies that were conducted. First, Zimbardo (1969) did a set of three studies that are fairly well known. In one study, he placed participants in oversized lab coats and hoods. The control group wore name tags and normal clothes. The idea was to see if anonymity would result in an increase in anti-normative behavior. Those wearing the hoods and lab coats did shock others (seen as antinormative to inflict pain) longer than the control condition, supporting Zimbardo’s hypothesis. We prefer our theories and research findings to all be straightforward and unfortunately, one of the things you will see as we move through the different explanations of crowd impact on the individual is that the results aren’t always straightforward and don’t fully support the different explanations. In Zimbardo’s second experiment, he used soldiers wearing their uniforms in one condition (anonymity condition) and then soldiers wearing their uniform and a name tag in the control condition (identifiable condition). In this experiment, he found that the soldiers with name tags shocked more than those in anonymous condition, which doesn’t support the prediction that anonymity will lead to anti-normative behavior. One of our later explanations will help us to understand this contradiction that isn’t explained by classic deindividuation theory alone (Postmes & Spears, 1998). Another study examining the impact of anonymity looked at aggressive driving behaviors. This field study examined the horn-honking behavior of either convertibles or 4 X 4s with top up (identifiable condition) or top down (anonymous condition). The confederate would pull in front of the car and when the light changed, they would hesitate to go. The horn-honking was measured in the first 12 seconds after the light changed. They looked at how quickly they honked when light changed, how long they pressed on the horn and the number of times they honked. The results again supported the anonymity leading to anti-normative behavior — more aggressive driving by horn honking (Ellison, et al., 1995).

There is one more important contribution to this classic theory. Diener (1979) refined the theory a bit and added that deindividuation was occurring because of the psychological mechanism of self-awareness reduction. It concluded that the less self-aware we are, the more deindividuated and the less likely we are to adhere to our personal norms and values. The well-known study done with children and Halloween candy helped him illustrate his point. In one version of the study, he has children wearing costumes concealing their identity completely or in large groups, which increase anonymity as well and he compares them to children who were alone or wearing costumes that didn’t conceal their identity. The study was done on Halloween and the house has a bowl of candy with a sign that says: “Please take one.” The measurement is how much candy is taken. Taking more than one would be considered a violation of the norm that is presented. Results support the prediction that kids who were more anonymous would engage in more anti-normative behavior and take more candy (Diener, et al., 1976). They are less self-aware, which means they aren’t thinking about their personal norm that stealing is wrong. There is a variation where there is a mirror behind the candy bowl and they are asked their name and address, and when made more self-aware, they take less candy. Even with the mirror, those in the anonymity condition weren’t affected. Researchers attributed this to the anonymity reducing self-awareness when wearing a disguise (Beaman, et al., 1979).

These studies are fascinating and it seems to make sense that anonymity, large groups, and lack of self-awareness would lead someone to feel they could violate norms and go against the group. However, the results are mixed and we don’t always see these conditions resulting in deindividuation as explained by this classic theory. In particular, from above, one of Zimbardo’s variations in the 1969 work. A meta-analysis of 60 independent deindividuation studies was conducted to better understand all the work that had been done on the topic and to determine where support exists for these different explanations (Postmes & Spears, 1998). These researchers first collected all the research on deindividuation that met their criteria for deindividuation and then averaged across the findings to determine what parts of the explanations were supported by all the research and which parts weren’t supported. For the classical theory of deindividuation, researchers findings were overall inconclusive. Only very small effects were found for deindividuation causing antinormative behavior: specifically, in the condition where there was a group present and reduced responsibility. This should leave us with some concern as to whether deindividuation as described in classical theory exists. It doesn’t seem that anonymity, large groups and lack of self-awareness actually cause antinormative behavior (Postmes & Spears, 1998). Alternative explanations to deindividuation effects – SIDE theory. So, how can we explain what these researchers found or what we see when large groups of people get together? Why does it seem like every time a large group gets together they do something wrong, like looting, acts of aggression between protestors, or tearing down goal posts? We definitely see these behaviors as inappropriate and violations of our societal social norms. It is wrong to hurt people and their property. Some of you might still be thinking about the last module and all the ways the group influenced us to conform, to go along for both normative and informational reasons (Deutsch & Gerard, 1955). In fact, you might be saying to yourself, the idea of deindividuation sounds cool, but didn’t we just learn that groups have powerful abilities to cause conformity? The real or imagined pressure of others usually results in us following the group social norms, not going against them. Violating norms is extremely difficult and done rarely by most of us. When we go along, we ensure that we won’t receive group sanctions. We learned above that we need the group to survive, feel good about ourselves, etc.

Before we introduce the social identity model of deindividuation effects model (SIDE), let’s look at the first study that suggested that local group norms could explain findings from Zimbardo’s (1969) study. This study, if you remember from the discussion above, had participants wearing a disguise to create anonymity and show that people will be more likely to engage in the antinormative behavior of shocking more than when they were identifiable. This first study used a variation of Zimbardo’s study by comparing a group that wore overalls and a mask (similar to the KKK-like clothing from Zimbardo’s study), to a group wearing nurses uniforms, and then a control that was identifiable. We would expect based on the classic theory of deindividuation that both groups that are dressed in uniform/disguises should be feeling anonymous. This anonymity should result in increased antinormative behavior, which in this study is shocking another human. It is expected that both of these conditions will lead you to shock more than the condition where you are identifiable. This again is in line with the classic deindividuation theory. The results found only a small increase in shocking from the KKK-like clothes and then, surprisingly, the nurses went in the other direction and shocked less compared to our control condition. They were more prosocial in their behavior. What does this tell us? What does it mean? It means that it is possible that we found these findings based on situational, local group norms. Nurses are supposed to help so it triggers a norm of not hurting. They shock less than those in the control (Johnson & Downing, 1979). Earlier I mentioned that the second study in Zimbardo’s (1969) set found that participants dressed in military uniforms and wore nametags (this made them identifiable) shocked more than participants dressed in military uniforms but were anonymous. It is possible that by placing a nametag on the participant wearing the uniform that it made the group salient for them and made them consider the norms associated with the group, possibly norms of aggression. So, again, just following the social norms of the salient group.

It was studies like Johnson & Downing (1979) that sparked researchers to expand beyond classic deindividuation theory and consider other possibilities for the effects from deindividuation manipulations. The social identity theory (Tajfel & Turner, 1986) that was discussed earlier in this module as a reason we desire group affiliation, was determined to be a solid framework for explaining the deindividuation phenomena and is referred to as social identity model of deindividuation effects model (SIDE). They suggested that the setup of these studies and real life situations where we combine anonymity, the closeness of the group and group immersion, actually makes the group importance and norms stronger for the person. They predicted that we should expect people to conform to the important group’s norms in the current context and not the larger, more general societal norms. So, I will steal the candy because everyone in my group took two extra pieces (group norm that is important) and I won’t even be thinking about the larger societal norm against stealing. The one shared aspect with classical deindividuation is the focus on anonymity. They see anonymity from immersion in the group to reduce self-awareness and make group identity more salient (Reicher et al., 1995; Postmes & Spears, 1998).

So, is this model supported in the meta-analysis by Postmes & Spears (1998) that we described above? Yes. The most robust finding was that the conditions of anonymity, larger groups and reduced self-awareness (which from classic deindividuation theory, should result in anti-normative behavior) actually resulted in greater conformity to the situational norms. So, we are seeing a specific form of social regulation and not the breakdown of regulations as previously thought. This model gives us a way to move forward, and it gives us explanations for both prosocial and antisocial behavior as well behavior that doesn’t fit either — similar to our original demonstration. If I had made certain group norms salient, would it change how you responded to the initial prompt. For example, family, being male or female, being a student, or parent, etc.? What do you think? Would your answer change?      


8.3. Groups that Interact & Their Impact on the Individual


Section Learning Objectives

  • Clarify research on groupthink.
  • Explain group polarization.


8.3.1. Groupthink

Let’s now move away from the mere presence of groups and move toward groups that are interacting. The first group situation we will examine deals with group decision making. Do groups help us make good or bad decisions? What are the group conditions that lead to better decision making? Some of the most important decisions in our lives are made by powerful people in groups. Currently, our government and judicial system both federally and locally are making decisions that will impact your lives. If you are a member of the United Methodist Church, they just made a decision to not support homosexuality in their clergy or members. Was this a good decision or bad decision? Well, sometimes with real-world examples, we can’t see the valence until time has passed. Some of the most popular examples used with groupthink are ones where history clearly demonstrates the rightness or wrongness, even though at the time, it might have been ambivalent. For example, Pearl Harbor was the result of a poor group decision with a leader who clearly underestimated the Japanese ability to bomb the United States. This didn’t take long for them to see that this was a poor decision with 2,400 lives lost and a large number of ships and planes to fight the war also gone (Janis, 1971, 1982). However, some political decisions may not be determined to be poor right away — it might take time to see the effects.

Irving Janis (1971) was inspired by decisions made by presidents and their advisors to propose the theory of groupthink. Specifically, those that went horribly wrong. He set out to find a theory that could help us understand this poor decision making. As you may have noted in the research module, most psychological theory comes from empirical studies results and these together either create the theory or will, after the fact, support the theory that is presented. In this case, Janis marketed his idea/theory in a more unusual way for scientists by first taking it more mainstream, publishing in Psychology Today (not a peer-reviewed journal), the psychology magazine. His ideas were exciting and interesting and people didn’t seem to mind that they were not supported by empirical evidence. In fact, even with limited empirical support, there are more than 100 citations of this theory and it is discussed in a variety of fields, business, psychology, political science and communication. There are even interventions that are designed to prevent it (Esser, 1998).

Janis proposed that groupthink occurred when group members suppressed dissent toward a poor decision because of a set of antecedent conditions. A review of the research shows that there are three different ways to interpret Janis’s model. First, a ‘strict’ interpretation, requires all the antecedents to be present. These are (Baron & Kerr, 2003):

  • directive leadership style (a leader who clearly states their perspective on the decision from the outset)
  • intense group cohesion (groups like the president’s cabinet are extremely close)
  • similarity of ideology (group polarization can occur – becoming more extreme on a topic)
  • pressure to be unanimous
  • group isolation from critics
  • insecure member self-esteem
  • sense of crisis

The second way to interpret this model is “additive.” In this perspective, as each condition is added, the groupthink experienced by group members is stronger. There are no published studies to support these first two interpretations. There is small support for the perspective that aspects of the groupthink model do lead to poor group decision making (Turner & Pratkanis, 1998).

There are also three different ways that researchers have suggested we think about the model of groupthink. First, one researcher thinks it is time to reject and get rid of the theory. There is very limited empirical evidence, and the historical evidence doesn’t account for all aspects of the theory (Fuller & Aldag, 1998). The second way to think about groupthink is to fix and possibly rethink the model. For example, Kramer (1998) believes that we should be considering the motivation to maintain political power while examining these group decisions as well as some of the other antecedents, but removing others. Finally, there are some that think we should revitalize the theory as Janis envisioned it — in fact, there are groupthink interventions already in existence being used in businesses as we read this module (Turner & Pratkanis, 1998).


8.3.2. Group Polarization

The first interacting group situation we examined looked at how a group impacts our decision making. Groupthink demonstrated how a group leader can assert a group norm, biasing the content of discussion and preventing dissent from the group members. This section will focus primarily on group discussion in general. What are the possible effects of talking with like-minded others? Most us of probably think this sounds nice. Remember, so far we have learned that we exist in a subjective reality, searching for frameworks to guide us, and choosing to surround ourselves with similar others. We have discussed, but not formally (this is done in module 12 on attraction), that we prefer similar others because liking people similar to us makes us feel that the way we see the world is the right or correct way. You can probably think of times you gathered with people like you: groups in the college, church groups, political groups, working moms, first time moms, etc. What was the result of meeting with these people? You probably felt better about your reality, but did it change an attitude or perspective that you had prior to entering the group?

Researchers found that when others sharing the same perspective are put into a group and left to discuss, they will move to a more extreme opinion from their initial opinion. This is referred to as group polarization (Moscovici & Zavalloni, 1969). The original work in this area was completed by a graduate student at MIT, Stoner (1961) was testing the common belief that a group would be riskier than an individual. The finding was termed the “Risky Shift” and spurred a ton of research aiming to support the idea that groups were riskier than individuals. It was the work by Moscovici and Zavalloni (1969) that started the support for the idea that what mattered was the attitudes and opinions of the group to start with. For example, if the group had a cautious attitude to start with instead of risky, they would become more cautious and not more risky, if they had a positive attitude toward cats, they would like them more at the end of the experiment examining people with positive cat attitudes. In the study completed by Moscovici & Zavalloni (1969), French students liked French people more and Americans less after discussion, becoming more extreme compared to their original like for their own people and dislike of Americans. An even more interesting finding is that people are unaware that this polarization is happening or has happened. Groups discussing with other like-minded individuals whether President Obama or President Bush was a better president became more extreme in their attitudes and when asked they misremembered ever having a less extreme attitude (Keating et al., 2016).

Why does this happen? What are the psychological processes that underlie and explain our tendency to become more extreme? It is possible as you read about group polarization, some of you were thinking about Sherif’s group norm work — people being placed in groups and then their responses converging. Informational influence is at work here. Instead of an ambiguous situation though, we have group members who share the same attitude or opinion, and they are presenting arguments and reasons for why they feel the way they do. This information is collected by each group member and adds to the reasons that are already held to support their opinion. This, in combination with normative influence (Deutsch & Gerard, 1955), is what causes this effect. Remember, normative influence is why we are concerned about social approval. This group contains members who share our views, they are like us, and will likely be seen as in-group members with greater likeability. As a result, we believe it is important to be a good group member and through social comparison (looking to others to see how we fit) we will want to move our attitude in the direction that is acceptable to the group. Typically, attitude strength above average is a safe bet to make so the group members will find us acceptable or approve of us (Baron & Kerr, 2003). Obviously, there are real-world dangers to this effect. If people are only surrounding themselves with like-minded individuals, they are likely to become more extremist in their ideas. This could account for the political tribalism we see today. It is very easy to only surround yourself with like-minded others, especially with the social-networking sites that are available. We even hear politicians discuss how they don’t interact as much with members of the opposite side as they did in the past. This is reflected in fewer bipartisan efforts. What are some other real-world issues that are currently being affected by group polarization? What do you think we can do to reduce this? Could spending time talking to moderate others help move you more toward the middle?

Module Recap

Group influence research has a long history. Starting in the late 1800s, it is some of the first research we did in psychology. This isn’t surprising given what we learned about the importance of groups to our lives. The long history gives us great examples of how theories change and morph through time. Social facilitation shows how a theory can be revitalized 40 years after research stops when someone comes up with a solution to the problems found. This is science and there is always hope that we can get closer to the truth behind human behavior as we perfect our science and move through time. This is true of the study of deindividuation and groupthink as well, with groupthink having much further to go as a supported theory. Social loafing and group polarization are much more straightforward, but the ever changing online world provides new ways to investigate these phenomena. The next module takes the foundations laid from attitudes, persuasion, conformity and group influence to help us better understand the processes of stereotyping, prejudice and discrimination.

2nd edition


Creative Commons License
Module 8: Group Influence by Washington State University is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License, except where otherwise noted.

Share This Book