Social thinking is concerned with how we view our own attitudes and behavior and those of others. We will discuss two of the major areas of research on social thinking—attributions and attitudes. First, we will examine attribution, the process (briefly discussed in Chapter 8) by which we explain our own behavior and the behavior of others. In other words, what do we perceive to be the causes of our behavior and that of others? Is the behavior due to internal causes (the person) or external causes (the situation)? Remember, we show a self-serving bias when it comes to explaining our own behavior. In this section, we will revisit this bias and examine two other biases in the attribution process (the fundamental attribution error and the actor-observer bias). The second major topic to be discussed is the relationship between our attitudes and our behavior. For example, do our attitudes drive our behavior, does our behavior determine our attitudes, or is it some of both? We will also consider the impact of role-playing on our attitudes and behavior.

408

Imagine it’s the first cold day in the fall, and you’re standing in a long line for coffee at the student union coffee shop. All of a sudden, a person at the head of the line drops her cup of coffee. You turn to your friend and say “What a klutz!” inferring that this behavior is characteristic of that person. You are making an internal (dispositional) attribution, attributing the cause of dropping the cup of coffee to the person. But what if it had been you who had dropped the cup of coffee? Chances are you would have said something like, “Boy was that cup hot!” making an external (situational) attribution by inferring that dropping the cup wasn’t your fault. We have different biases in the attributions we make for behavior we observe versus our own behavior. Let’s look first at being an observer.

Attributions for the behavior of others. As an observer, we tend to commit the fundamental attribution error (Ross, 1977). The fundamental attribution error is the tendency as an observer to overestimate internal dispositional influences and underestimate external situational influences on others’ behavior. Simply put, observers are biased in that they tend to attribute others’ behavior to them and not the situation they are in. In the coffee example, we tend to make an internal attribution (the person is a klutz) and ignore possible external situational factors, such as the cup being really hot or slippery. The fundamental attribution error tends to show up even in experiments in which the participants are told that people are only faking a particular type of behavior. For example, in one experiment participants were told that a person was only pretending to be friendly or to be unfriendly (Napolitan & Goethals, 1979). Even with this knowledge, participants still inferred that the person they met in the experiment was actually like the way he or she acted, either friendly or unfriendly.

Think about the participants in Milgram’s obedience experiments. On first learning of their destructive obedient behavior, didn’t you think the teachers were horrible human beings? How could they treat fellow humans in that way? Or consider the teachers themselves in these experiments. When the learners kept making mistakes, the teachers may have thought the learners were incredibly stupid and so deserved the shocks. People who have been raped are sometimes blamed for provoking the rape, and people who are homeless are often blamed for their poverty-stricken state. Placing blame in this manner involves the just-world hypothesis, the assumption that the world is just and that people get what they deserve (Lerner, 1980). Beware of just-world reasoning. It is not valid, but is often used to justify cruelty to others.

409

The fundamental attribution error impacts our impressions of other people. There are two related concepts that you should also be aware of when forming impressions of others—the primacy effect and the self-fulfilling prophecy. In the primacy effect, early information is weighted more heavily than later information in forming an impression of another person. Beware of this effect when meeting someone new. Develop your impression slowly and carefully by gathering more information across time and from many different situations. Also be careful about the initial impressions that you make on people. Given the power of the primacy effect, your later behavior may not be able to change their initial impression of you. Try to be your true self when you first meet someone so that the primacy effect will be more accurate.

In the self-fulfilling prophecy, our expectations of a person elicit behavior from that person that confirms our expectations. In other words, our behavior encourages the person to act in accordance with our expectations (Jones, 1977; Rosenthal & Jacobson, 1968). For example, if you think someone is uncooperative, you may act hostile and not be very cooperative in your interactions with that person. Given your hostile behavior, the person responds by being uncooperative, confirming your expectations. The person may not really be uncooperative, however, and may only act this way in response to your behavior. However, your expectations will have been confirmed, and you will think that the person is uncooperative. Self-fulfilling prophecy is related to our tendency toward confirmation bias in hypothesis testing (see Chapter 6). Rather than acting in a manner that confirms your expectations, act in the opposite way (in the example, instead of being uncooperative, be cooperative) and see what happens. You may be surprised.

Attributions for our own behavior. Now let’s consider our own behavior and making attributions. Our attribution process is biased in a different way when we are actors and not observers. Think about the example where you dropped the cup of coffee (you were the actor). You don’t make a dispositional attribution (“I’m clumsy”). You make a situational attribution (“The cup was slippery”). As actors we tend to have what’s called the actor-observer bias, the tendency to attribute our own behavior to situational influences, but to attribute the behavior of others to dispositional influences. Why this difference in attributional bias? As observers, our attention is focused on the person, so we see him as the cause of the action. As actors, our attention is focused on the situation, so we see the situation as the cause of the action. We are more aware of situational factors as actors than as observers. This explanation is supported by the fact that we are less susceptible to the bias toward dispositional attributions with our friends and relatives, people whom we know well.

410

We described another attribution bias in the last chapter, the self-serving bias—the tendency to make attributions so that one can perceive oneself favorably. As actors, we tend to overestimate dispositional influences when the outcome of our behavior is positive and to overestimate situational influences when the outcome of our behavior is negative. In the last chapter, we were discussing this bias’s role as a defense mechanism against depression. In this chapter, let’s look at how self-serving bias qualifies the actor-observer bias. It defines the type of attribution we make as actors based on the nature of the outcome of our actions. Think about our reaction to a test grade. If we do well, we think that we studied hard and knew the material, both dispositional factors. If we don’t do well, then we may blame the test and the teacher (“What a tricky exam,” or “That test was not a good indicator of what I know”). We take credit for our successes but not our failures. Teachers also have this bias. For example, if the class does poorly on a test, the teacher thinks that the test she made up was just fine, but that the students weren’t motivated or didn’t study. It is important to realize that the self-serving bias and the other attribution biases do not speak to the correctness of the attributions we make, but only to the types of attributions we tend to make. This means that our attributions will be sometimes correct and sometimes incorrect.

The self-serving bias also leads us to see ourselves as “above average” when we compare ourselves with others on positive dimensions, such as intelligence and attractiveness. This tendency is exemplified in Garrison Keillor’s fictional Lake Wobegon community, where “All the women are strong, all the men are good-looking, and all the children are above average.” We tend to rate ourselves unreasonably high on dimensions that we value (Mezulis, Abramson, Hyde, & Hankin, 2004). Think about it. If you were asked to compare yourself with other people on intelligence, what would you say? If you respond like most people, you would say “above average.” However, most people cannot be “above average.” Many of us have to be average or below average. Intelligence, like most human traits, is normally distributed with about half of us below average, about half of us above average, and the rest of us average.

The self-serving bias also influences our estimates of the extent to which other people think and act as we do. This leads to two effects—the false consensus effect and the false uniqueness effect (Myers, 2013). The false consensus effect is the tendency to overestimate the commonality of one’s opinions and unsuccessful behaviors. Let’s consider two examples. If you like classical music, you tend to overestimate the number of people who also like classical music, or if you fail all of your midterm exams, you tend to overestimate the number of students who also failed all of their midterms. You tend to think that your opinion and your negative behavior are the consensus opinion and behavior.

411

The false uniqueness effect is the tendency to underestimate the commonality of one’s abilities and successful behaviors. For example, if you are a good pool player, you tend to think that few other people are. Or, if you just aced your psychology exam, you think few students did so. You think your abilities and successful behaviors are unique. The false consensus effect and the false uniqueness effect relate to the self-protective function of the self-serving bias. We want to protect and enhance our view of ourselves, our self-esteem.

All three major attributional biases (the fundamental attribution error, the actor-observer bias, and the self-serving bias) are summarized in Table 9.4.

In this section, we are going to consider our attitudes. In simple terms, attitudes are evaluative reactions (positive or negative) toward things, events, and other people. Our attitudes are our beliefs and opinions. What do you think of the Republican political party, abortion, Twitter, or rap music? Have you changed any of your attitudes during your life, especially since you have been in college? Most of us do. Our attitudes often determine our behavior, but this is not always the case.

When our behavior contradicts our attitudes. Our attitudes tend to guide our behavior when the attitudes are ones that we feel strongly about, when we are consciously aware of our attitudes, and when outside influences on our behavior are not strong. For example, you may think that studying is the top priority for a student. If you do, you will likely get your studying done before engaging in other activities. However, if there is a lot of pressure from your roommates and friends to stop studying and go out partying, you may abandon studying for partying. But what happens when our attitudes don’t guide our behavior and there isn’t a lot of outside influence on our behavior? To help you understand the answer to this question, we’ll consider a classic study (Festinger & Carlsmith, 1959).

412

Imagine that you are a participant in an experiment. You show up at the laboratory at the assigned time, and the experiment turns out to be incredibly boring. For an hour, you perform various boring tasks, such as turning pegs on a pegboard over and over again or organizing spools in a box, dumping them out, and then organizing them again. When the hour is over, the experimenter explains to you that the experiment is concerned with the effects of a person’s expectations on their task performance and that you were in the control group. The experimenter is upset because his student assistant hasn’t shown up. She was supposed to pose as a student who has just participated in the experiment and tell the next participant who is waiting outside that this experiment was really enjoyable. The experimenter asks if you can help him out by doing this, and for doing so, he can pay you. His budget is small, though, so he can only pay you $1. Regardless, you agree to help him out and go outside and tell the waiting participant (who is really a confederate of the experimenter and not a true participant) how enjoyable and interesting the experiment was.

Before you leave, another person who is studying students’ reactions to experiments asks you to complete a questionnaire about how much you enjoyed the earlier experimental tasks. How would you rate these earlier incredibly boring tasks? You are probably thinking that you would rate them as very boring, because they were. However, this isn’t what researchers observed. Participants’ behavior (rating the tasks) did not follow their attitude (the tasks were boring). Participants who were paid only $1 for helping out the experimenter (by lying about the nature of the experimental tasks to the next supposed participant) rated the tasks as fairly enjoyable. We need to compare this finding with the results for another group of participants who received $20 for lying about the experiment. Their behavior did follow their attitude about the tasks. They rated the tasks as boring. This is what researchers also found for another group of participants who were never asked to help out the experimenter and lie. They also rated the tasks as boring. So why did the participants who lied for $1 not rate the tasks as boring?

Before answering this question, let’s consider another part of this experiment. After rating the task, participants met again with the experimenter to be debriefed, and were told the true nature of the study. Following the debriefing, the experimenter asks you to give back the money he paid you. What would you do if you were in the $20 group? You are probably saying, “No way,” but remember the studies on obedience and the high rates of obedience that were observed. What actually happened? Just like all of the participants agreed to lie (whether for $1 or $20), all of the participants gave back the money, showing once again our tendency to be obedient and comply with requests of those in authority. Now let’s think about why the $1 group rated the tasks differently than the $20 group.

413

Festinger’s cognitive dissonance theory. Their behavior can be explained by Leon Festinger’s cognitive dissonance theory, which proposes that people change their attitudes to reduce the cognitive discomfort created by inconsistencies between their attitudes and their behavior (Festinger, 1957). Let’s consider a real-life example before applying this theory to the participants who were paid $1 for lying. Think about people who smoke. Most smokers have the attitude (believe) that smoking is dangerous to their health, but they continue to smoke. Cognitive dissonance theory says that smokers feel cognitive discomfort because of the inconsistency between their attitude about smoking (dangerous to their health) and their behavior (continuing to smoke). To reduce this cognitive disharmony, either the attitude or the behavior has to change. According to the cognitive dissonance theory, many smokers change their attitude so that it is no longer inconsistent with their behavior. For example, a smoker may now believe that the medical evidence isn’t really conclusive. The change in the attitude eliminates the inconsistency and the dissonance it created. So now let’s apply cognitive dissonance theory to the participants in the boring tasks study who lied for $1. Why did they rate the tasks as enjoyable? Their attitude was that the tasks were incredibly boring, but this was inconsistent with their behavior, lying about the tasks for only $1. This inconsistency would cause them to have cognitive dissonance. To reduce this dissonance, the participants changed their attitude to be that the tasks were fairly enjoyable. Now the inconsistency and resulting dissonance are gone.

A key aspect of cognitive dissonance theory is that we don’t suffer dissonance if we have sufficient justification for our behavior (the participants who were paid $20 in the study) or our behavior is coerced. Also, cognitive dissonance sometimes changes the strength of an attitude so that it is consistent with past behavior. Think about important decisions that you have made in the past, for example, which college to attend. Cognitive dissonance theory says that once you make such a tough decision, you will strengthen your commitment to that choice in order to reduce cognitive dissonance. Indeed, the attractiveness of the alternate choices fades with the dissonance (you don’t understand why you ever were attracted to the other schools) as you find confirming evidence of the correctness of your choice (you like your classes and teachers) and ignore evidence to the contrary (such as your school not being as highly rated as the other schools). Carol Tavris and Elliot Aronson (2007) provide numerous real-world examples of this cognitive dissonance–driven justification of our decisions, beliefs, and actions in their illuminating book, Mistakes Were Made (But Not By Me). Recent research by Egan, Santos, and Bloom (2007) indicates that such decision rationalization even appears in children and nonhuman primates.

414

Bem’s self-perception theory. There have been hundreds of studies on cognitive dissonance, but there is an alternative theoretical explanation for the results of some of these studies—Daryl Bem’s self-perception theory, which proposes that when we are unsure of our attitudes we infer them by examining our behavior and the context in which it occurs (Bem, 1972). According to Bem, we are not trying to reduce cognitive dissonance, but are merely engaged in the normal attribution process that we discussed earlier in this chapter. We are making self-attributions. Self-perception theory would say that the participants who lied for $1 in the boring tasks experiment were unsure of their attitudes toward the tasks. They would examine their behavior (lying for $1) and infer that the tasks must have been fairly enjoyable and interesting or else they would not have said that they were for only $1. Those paid $20 for lying would not be unsure about their attitude toward the boring tasks, because they were paid so much to lie about them. According to self-perception theory, people don’t change their attitude because of their behavior but rather use their behavior to infer their attitude. People are motivated to explain their behavior, not to reduce dissonance. According to self-perception theory, there is no dissonance to be reduced.

So which theory is better? Neither is really a better theory than the other. Both theories have merit and both seem to operate—but in different situations. This is similar to our earlier discussion of color vision theories, the trichromatic-color theory and the opponent-process theory in Chapter 3. Remember that trichromatic-color theory operates at the receptor cell level and opponent-process theory at the post-receptor cell level in the visual pathways. Cognitive dissonance theory seems to be the best explanation for behavior that contradicts well-defined attitudes. Such behavior creates mental discomfort, and we change our attitudes to reduce it. Self-perception theory explains situations in which our attitudes are not well-defined; we infer our attitudes from our behavior. As with the color vision theories, both cognitive dissonance theory and self-perception theory operate, but at different times.

The impact of role-playing. Now let’s consider one final factor that impacts the complex relationship between our attitudes and our behavior—role-playing. We all have various social roles that we play—student, teacher, friend, son or daughter, parent, employee, and so on. Each role is defined by a socially expected pattern of behavior, and these definitions have an impact on both our behavior and our attitudes. Given the power of roles on behavior, think about how each of the various roles you play each day impacts your own attitudes and behavior. They are powerful influences. As social psychologist David Myers has observed, “we are likely not only to think ourselves into a way of acting but also to act ourselves into a way of thinking” (Myers, 2002, p. 136).

415

The Stanford prison experiment (SPE) conducted in 1971 at Stanford University by Philip Zimbardo has often been viewed as a dramatic example of the power of roles, specifically prison guard and prisoner (Haney, Banks, & Zimbardo, 1973; Zimbardo, 2007; Zimbardo, Maslach, & Haney, 1999). Like Milgram’s obedience study, the SPE is another famous, controversial psychological study that has become a contentious classic. Thus, as we did for Milgram’s study, we will cover the SPE in more detail than usual. Also, like Milgram’s study, the SPE has remained in the public eye from the initial publications about it until the present. Curiously, it was also dramatized in 2015 in a theatrical film, The Stanford Prison Experiment, starring Billy Crudup as Zimbardo. Even stranger, the creators of these two contentious classics, Milgram and Zimbardo, were high school classmates at James Monroe High School in the Bronx in the late 1940s. It is highly unlikely that anyone back then could have predicted that Milgram and Zimbardo would both become famous social psychologists and conduct such controversial research. As with Milgram’s study and his interpretation of its results, much criticism has surfaced in response to both the SPE itself and Zimbardo’s interpretation of its results, casting doubt on the validity of the study, its findings, and the conclusions that have been drawn from the findings. In addition, a more recent prison study, known as the BBC Prison Study, has been conducted (Haslam & Reicher, 2005; Reicher & Haslam, 2006), and its results differ greatly from those of the SPE. We will first briefly describe the SPE, consider some of the methodological criticisms of it, and then describe the BBC Prison Study, its results, and the interpretation of these results.

First, you should realize that although Zimbardo’s study is referred to as an experiment (hence, the E in SPE), it really is not an experiment in the sense that we described in Chapter 1. It would be more aptly described as a simulation study, specifically a psychological simulation of prison life. So what did Zimbardo do? He recruited male college students to participate in the study and renovated the basement in the Stanford University psychology building to be a mock prison. Any volunteers with prior arrests or any with medical or mental problems were disqualified from participating. Following psychological assessments and in-depth interviews, 24 volunteers were chosen and then randomly assigned to the role of prisoner or guard (12 as guards and 12 as prisoners). The guards were given uniforms and billy clubs and instructed to enforce the rules of the mock prison. The prisoners were arrested, booked, locked in cells, and had to wear humiliating clothing (smocks with no undergarments). Such clothing was used in an attempt to simulate the emasculating feeling of being a prisoner. How did the roles of prisoner and prison guard impact the attitudes and behavior of the participants? Some of the participants began to take their respective roles too seriously. After only one day of role-playing, some of the guards started treating the prisoners cruelly. Some of the prisoners rebelled, and others began to break down. What was only supposed to be role-playing seemed to become reality. The guards’ treatment of the prisoners became both harsh and degrading. For example, some prisoners were made to clean out toilets with their bare hands. Some prisoners began to hate the guards, and some of them were on the verge of emotional collapse. The situation worsened to such a degree that Zimbardo said that he had to stop the study after six days. According to Zimbardo, due to the power of their situational roles, the participating college students had truly “become” guards and prisoners. The roles transformed their attitudes and behavior. In sum, the abusive guards were not “bad apples.” It was the “bad barrel” of the Stanford prison (the situation) that temporarily transformed them.

416

Let’s consider Zimbardo’s situational account of the SPE results. First, it is not clear that the SPE participants behaved in ways consistent with their roles because of their natural acceptance of situational role requirements. It seems more likely that their behavior was due to the active leadership provided by Zimbardo (e.g., Banyard, 2007; Haslam & Reicher, 2012). Zimbardo served as prison superintendent and gave the guards an orientation that seems, in retrospect, to provide clear guidance about how they should behave. Zimbardo (2007) recounted the following from this orientation:

We can create boredom. We can create a sense of frustration. We can create fear in them, to some degree. We can create a notion of the arbitrariness that governs their lives, which are totally controlled by us, by the system, by you, me. . . . They’ll have no privacy at all, there will be constant surveillance—nothing they do will go unobserved. They will have no freedom of action. They will be able to do nothing and say nothing that we don’t permit. We’re going to take away their individuality in various ways. . . . In general, what all this should create in them is a sense of powerlessness. We have total power in the situation. They have none. (p. 55)

As Banyard (2007) pointed out, notice the use of pronouns in this orientation. Zimbardo puts himself with the guards (“we”) and gives clear instructions for the hostile environment that “we” are going to create for “them.” Thus, Zimbardo’s leadership may have legitimized oppression in the SPE. Banyard concludes that, “It is not, as Zimbardo suggests, the guards who wrote their own scripts on the blank canvas of the SPE, but Zimbardo who creates the script of terror. . . .” (2007, p. 494). But Zimbardo’s guidance continued past the orientation. Via his silence, Zimbardo provided tacit approval of the abusive guards’ behavior, thereby confirming to them that they were behaving as they should. He also specifically instructed his prison warden (a research assistant) to chastise the guards who were not behaving like the “bad” guards (Zimbardo, 2007). With respect to Zimbardo’s active role in the SPE, some recent comments by one the guards in the SPE, John Mark, are relevant (Ratnesar, 2011). Mark said:

I didn’t think it was ever meant to go the full two weeks. I think Zimbardo wanted to create a dramatic crescendo, and then end it as quickly as possible. I felt that throughout the experiment, he knew what he wanted and then tried to shape the experiment—by how it was constructed, and how it played out—to fit the conclusion that he had already worked out. He wanted to be able to say that college students, people from middle-class backgrounds—people will turn on each other just because they’re given a role and given power. Based on my experience, and what I saw and what I felt, I think that was a real stretch. I don’t think the actual events match up with the bold headline. I never did, and I haven’t changed my opinion.

417

Mark’s sentiments are echoed by Carlo Prescott, an African-American ex-con who was the SPE’s chief consultant on real prisons (Zimbardo, 2007), in his 2005 letter in The Stanford Daily entitled “The Lie of the Stanford Prison Experiment.” He expressed regret for his involvement and disclosed that it was he and not the guards who came up with the abusive and humiliating behaviors displayed by the guards. He wrote:

My opinion, based on my observations, was that Zimbardo began with a preformed blockbuster conclusion and designed an experiment to “prove” that conclusion . . . ideas such as bags being placed over the heads of prisoners, inmates being bound together with chains and buckets being used in place of toilets in their cells were all experiences of mine at the old “Spanish Jail” section of San Quentin and which I dutifully shared with the Stanford Prison Experiment brain trust months before the experiment started. . . . How can Zimbardo . . . express horror at the behavior of the “guards” when they were merely doing what Zimbardo and others, myself included, encouraged them to do at the outset or frankly established as ground rules?

That Zimbardo’s strong involvement and guidance (experimenter bias) may have been critical to the SPE outcome is also supported by some findings for simulated prison environments in a study conducted in Australia by Lovibond, Mithiran, and Adams (1979). These researchers found that changes in the experimental prison regime produced dramatic changes in the relations between guards and prisoners. In a more liberal prison condition (essentially the opposite of what Zimbardo created) in which security was maintained in a manner that allowed prisoners to retain their self-respect and in a participatory condition in which prisoners were treated as individuals and included in the decision-making process, the behavior of the guards and prisoners was rather benign and very dissimilar from the dramatic behavioral outcomes observed in the SPE.

Zimbardo’s guard orientation that essentially told the guards how he wanted them to behave, his tacit approval of the abusive guards’ later abusive behavior, and the prison warden’s talks with some of the nonabusive guards about being more abusive are excellent examples of demand characteristics (Orne, 1962). Demand characteristics are cues in the experimental environment that make participants aware of what the experimenter expects to find and how participants are expected to act. Demand characteristics can impact the outcome of an experiment because participants may alter their behavior to conform to the experimenter’s expectations. Hence, given the blatant demand characteristics of the SPE, it should not be surprising that some of the guards became “bad” guards. Banuazizi and Movahedi (1975) provide data that indicate that the SPE would have likely been confounded by demand characteristics even if Zimbardo had not been so actively involved in the study. They mailed students a questionnaire that included a brief description of the procedures followed in the SPE and some open-ended questions to determine their awareness of the experimental hypothesis and their expectancies regarding the outcomes of the experiment. Of the 150 students responding, a vast majority determined the experiment’s purpose was to show that normal people placed into the positions of guards and prisoners would act like real guards and prisoners and predicted that the behavior of the guards toward prisoners would be oppressive, hostile, etc., and that prisoners would react in passive or defiant ways or both. These findings indicate that the demand characteristics inherent in the SPE’s design were fairly obvious, so it is not surprising that the guards and prisoners acted in the ways they did.

418

Given the strong demand characteristics available in the SPE and Zimbardo’s blatant guidance, what is surprising is that only a few of the guards (about one-third) were “bad” guards (Zimbardo, 2007). The rest were “strict but fair” guards and “good” guards (those that sided with the prisoners). This finding obviously runs counter to Zimbardo’s simplistic situational explanation of the SPE results and shows that dispositional factors (personality traits) contributed to the results. Given that dispositional factors played a role in the behavior of the guards, is it possible that the “bad” guards were, using Zimbardo’s metaphor, “bad apples”? That is, is there anything unusual about the type of person who would volunteer to participate in a study like the SPE? If so, then participant self-selection may have occurred in the SPE. This means that the participants selectively volunteered for the SPE because they were attracted to a prison study because of their personalities. Carnahan and McFarland’s (2007) findings suggest that this is a distinct possibility. They recruited students in two ways: (1) for a psychological study of prison life using a virtually identical newspaper advertisement as used in the SPE, and (2) for just a psychological study, an identical ad but without the mention of prison life. Carnahan and McFarland found that those who volunteered to participate in the prison life study tended to be more aggressive, authoritarian, socially dominant, and Machiavellian and less empathetic and altruistic than those who volunteered for the more innocuous experiment. Thus, counter to what Zimbardo argued, it could be that the barrel was not bad but that some of the “apples” (the “bad” guards) were already “bad” before being put in the barrel (the Stanford prison).

It is clear that the SPE had serious methodological shortcomings—the presence of experimenter bias and strong demand characteristics and possibly participant self-selection—that prevent any sound conclusions being drawn from its results. Because of this, some textbook authors do not even cover the SPE in their texts (e.g., Gray, 2013). Unfortunately, textbook authors who do cover the SPE tend not to cover its shortcomings; hence, teachers tend not to cover them in their courses (Bartels, Milovich, & Moussier, 2016). Because of its renown both within and outside of psychology, we discussed the SPE and its methodological shortcomings so that you would be aware of these shortcomings and how they invalidate Zimbardo’s conclusions about the results of the study. It is important, though, that you realize that this does not mean that social roles are not important influences on our behavior; they definitely are (Greenberg, Schmader, Arndt, & Landau, 2015). It is just that the SPE is not a cogent demonstration of their influence. It is also worth noting that Zimbardo served as an expert witness in the trial of one of the abusive prison guards at Abu Ghraib, the U.S. military prison in Iraq (Zimbardo, 2007). He argued that situational pressures, as he claimed in the SPE, led the soldier to commit the abusive acts. His argument, however, failed, and the soldier was reduced to the lowest rank and given a harsh prison sentence and a dishonorable discharge. The prosecutor’s winning argument was that individuals are responsible for their own behavior and that the abusive guards’ immoral, evil behavior stemmed from their dispositions (they were “bad apples”). Next, we would like to make you are aware of a more recent prison study that was conducted, one that you likely have not heard about but that is methodologically sounder than the SPE.

419

Two British social psychologists, Alexander Haslam and Stephen Reicher (the proponents of the engaged-followership explanation of Milgram’s findings that we discussed earlier in the chapter), in collaboration with the British Broadcasting Corporation (BBC), re-created the SPE, but with ethical procedures that ensured that the study would not harm participants (Haslam & Reicher, 2005, Haslam & Reicher, 2012; Reicher & Haslam, 2006). This study has become known as the BBC Prison Study. It was filmed by the BBC and televised in 2002. As in the SPE, the male volunteer participants were randomly assigned to the roles of guards and prisoners within a custom-built prison setting. However, unlike Zimbardo and his colleagues, Haslam and Reicher took no leadership role in the study (in particular, they did not instruct their guards to subjugate the prisoners to their will in the way that Zimbardo did). Hence, the fact that the guards’ and prisoners’ behavior diverged markedly from that in the SPE bolsters the argument that demand characteristics (and Zimbardo’s leadership) were likely responsible for the outcome of the SPE.

So what happened in the BBC Prison Study? The guards failed to identify with their role and were reluctant to impose their authority. The prisoners, however, did form a cohesive group identity, leading them to rebel and overthrow the established regime after six days. Hence, rather than brutal guards and passive prisoners, ambivalent guards and rebellious prisoners materialized. The guards and the prisoners then formed a single self-governing, egalitarian commune, but the commune disintegrated after a short time, because again some participants did not want to discipline those who broke the commune’s rules. Following the collapse of the communal system, some former prisoners and former guards proposed a coup in which they would become the new guards, creating a more hard-line prisoner–guard divide and using force, if necessary, to maintain it. At this point, Haslam and Reicher brought the study to a close on the eighth day. Haslam and Reicher interpreted their findings in terms of social identity theory, in that power resides in the ability of a group to establish a sense of shared identity. This group power can be used for a positive purpose, as illustrated by the prisoners in the BBC prison study, or for a negative one, as illustrated by the guards in the SPE. They concluded that people do not automatically assume roles that are given to them, as was suggested by the Zimbardo’s account of the SPE, and that tyranny arises from a complex process in which group failure and powerlessness lead group members to identify with authoritarian regimes and their leaders (a process that was short-circuited in the SPE because, from the outset, Zimbardo encouraged identification with his own authoritarian leadership). Of course, it may be argued that the guards in the BBC Prison Study failed to display the brutality of the SPE guards because their behavior was filmed and would ultimately be broadcast on television. Against this, Haslam and Reicher (2012) noted that toward the end of the study, once a group of new guards had come to identify with their role, they proved very willing to oppress prisoners. Indeed, here the regime that was in ascendancy closely resembled that in the SPE. However, the participants had arrived at that point because they believed in the authoritarian regime they were implementing and not because they were blindly conforming to role.

420