So far we’ve outlined the broad motives and systems that guide our thinking about the social world. Let’s turn now to consider some of the more specific thought processes that people use to understand the world. The first thing to notice is how quickly and effortlessly the mind classifies stimuli into categories. Categories are like mental containers into which people place things that are similar to each other. Or, more precisely, even if two things are quite different from one another (two unique individuals, for instance), when people place them in the same category (“frat boys”), they think about those two items as though they were the same. This makes life easier.

Categorization is an interesting process in its own right, but it is just the starting point of our mind’s active meaning making. That’s because as soon as people classify a stimulus as an instance of a category, their minds quickly access knowledge about that category, including beliefs about the category’s attributes, expectations about what members of that category are like, and plans for how to interact with it, if at all. All of this knowledge is stored in memory in a mental structure called a schema. For example, if you are at the library and you categorize a person behind the desk as a librarian, you instantly access a schema for the category librarian that contains beliefs about which traits are generally shared by members of that group (e.g., intelligence), theories about how librarians’ traits relate to other aspects of the world (e.g., librarians probably do not enjoy extreme sports), and examples of other librarians you have known. Bringing to mind schemas allow the person to “go beyond the information given” (Bruner, 1957), elaborating on the information that strikes their senses with what they already know (or think they know). We can demonstrate this with a simple example. Read the following paragraph:

94

The procedure is quite simple. First, you arrange things into different groups. Of course, one pile may be sufficient, depending on how much there is to do. If you have to go somewhere else due to lack of facilities, that is the next step; otherwise you are pretty well set. It is important not to overdo things. That is, it is better to do too few things at once than too many. At first the whole procedure will seem complicated. Soon, however, it will become just another facet of life (Bransford & Johnson, 1973, p. 400).

You might be scratching your head right now, wondering what these instructions are referring to. If you close your textbook and five minutes later try to remember all of the points in the paragraph, you will probably run into difficulty. What if we tell you that the paragraph is about laundry? Now, reread the paragraph and you will see that the information makes much more sense to you than it did initially. After five minutes, you might do a reasonable job of remembering each of the steps described. The mere mention of the word laundry activated your schema of this process and made it a template for understanding the information you were reading.

Schemas are given special names depending on the type of knowledge that they represent. Schemas that represent knowledge about events are called scripts. These types of schemas (like the laundry example) always involve a temporal sequence, meaning that they describe how events unfold over time (first you sort, then you put one pile into the machine, then you add the soap, and so on). Scripts make coordinated action possible. Playing a game of tennis requires that both you and your partner have a schema of the game, so that you can coordinate your actions and follow the rules of the game, even though you are playing against one another. They also allow you to fill in missing information. If I told you that I got a sandwich at the student union, I don’t need to tell you, for example, that I paid for it. You can fill that detail in because you have the same basic “getting food at a restaurant” script as I do. Our reliance on scripts becomes embarrassingly apparent when we find ourselves without a script for a new situation. Imagine being invited to a Japanese tea ceremony but not knowing where to sit, what to say and when to say it, and how to sip the tea—when everyone else in attendance seems thoroughly acquainted with this very complex ritual of great importance for maintaining respectful social relations.

Schemas that represent knowledge about other people are called impressions. Your schema of the cyclist Lance Armstrong might include physical characteristics (athletic, good looking), personality traits (charismatic, courageous), and other beliefs about him (philanthropist, cancer survivor, stripped of titles after doping scandal). Similarly, we can also have a schema for a category of people (e.g., sports superstars), called a stereotype. You can see that your impression of Armstrong contains many traits (e.g., wealthy, athletic, courageous) that are also part of your stereotype for sports superstars. Finally, as we will discuss in chapter 5, we also have a schema about ourselves—our self-concept.

95

Regardless of their type, the content of our schemas consists of a pattern of learned associations. These patterns of associations can change and expand over time. You first might have learned about Lance Armstrong as an incredible athlete and cancer survivor and only later had to update this positive view of him after repeatedly encountering media reports about his use of performance-enhancing drugs, which eventually led to his being stripped of his seven Tour de France titles. Some of our associations with Armstrong might be stronger than others, because we more frequently think about or hear about Armstrong in terms of those aspects. The learned associations stored in our schemas profoundly shape our perception, judgment, and behavior.

But it’s also important to realize that schemas are not passively filled up with information from the outside. Because of our need for specific closure—again, the motive to maintain particular beliefs and attitudes—we often tailor our schemas to include only some pieces of knowledge. Think about it this way: On your computer you probably have file folders that contain documents, pictures, and sound files that are related in some way, and you label those file folders accordingly, such as “Social Psychology Class” and “Summer Vacation.” The schemas stored in your long-term memory are like those file folders in the sense that they contain all the bits of knowledge you have about a given category, from Nazis and pedophiles to doorknobs and stickers. But the similarities end there. Computer file folders usually don’t magically acquire or lose documents, and they never insist that you open this picture and get nervous if you open up that picture. But that is exactly what schemas do, even without our conscious awareness. For example, if you are the president of the Lance Armstrong fan club, your schema for Lance likely will emphasize the bits of knowledge that flatter the athlete (great cyclist, charity sponsor) and will downplay anything that casts him in a negative light.

Let’s take a closer look at where we acquire the knowledge that makes up our schemas. The example of Lance Armstrong pointed to various sources of knowledge. In some cases, we come into direct contact with people, events, and ideas and form concepts on the basis of that personal experience. But looking at this from the cultural perspective, we also learn a great deal about our social world indirectly, from parents, teachers, peers, books, newspapers, magazines, television, movies, and the Internet. A lot of our general knowledge comes during childhood from the culture in which we are raised. As children learn language and are told stories, they are taught concepts such as honesty and courage, good and evil, love and hate. From this learning, people develop ideas about what people in the world are like, the events that matter in life, and the meaning of their own thoughts and feelings, among other fundamental lessons.

A considerable amount of our cultural knowledge is transmitted to us by our parents, but also by peers, teachers, and mass media sources. For most children in industrialized nations, television and movies provide scripts of situations (workplace interactions: Mad Men; romance: the Twilight series), schemas of types of people (villain, hero, femme fatale, nerd, ingénue) and stereotypes of groups of people (gay men are effeminate; grandmothers are kind; Asians are martial artists) before the child has firsthand experience with such situations, people, and groups. And children intuitively sense that television and books provide a preview of the next steps in their lives: Grade school kids tend to like shows and books about middle school, and middle school kids tend to like shows and books about high school.

96

Finally, the basic way that we categorize information and build schemas is thought to be culturally universal, but as we have described it here, the content of those schemas and how they are organized is shaped by our cultural experiences. This can result in cultural differences in the meaning that concepts can sometimes have. For example, kids who grow up in a rural Native American culture—which values connections with nature—have a concept of “animal” that is most closely linked to those species that become part of their daily lives (Winkler-Rhoades et al., 2010). In contrast, urban-dwelling European American kids asked to list animals bring to mind exotic species such as elephants and lions that populate their picture books. Here we see the both groups develop a schema for the same general category, but the content of that schema differs in important ways, depending on the physical and social environment in which people carry out their daily lives.

Rumors and Gossip

Rumors and gossip are two other common sources of knowledge contained in our schemas. Much of what we learn about other people or events comes from news passed from one person to another. But beware. When information is passed from person to person to person before you get it, it tends to be distorted in various ways.

For one, as people perceive and relay information, it is altered a bit as it is filtered by each person’s schemas and motive for specific closure. Specifically, transmitted information can be biased by processes called sharpening and leveling. Think about how you tell a story to a friend. You’re probably going to emphasize the main features of the story, which is called sharpening, and leave out a lot of details, which is called leveling. The main features are more memorable than the details, and they also make for a more interesting tale. An unfortunate consequence of this bias is that people hearing about a person or event, rather than gaining knowledge firsthand, will tend to form an oversimplified, extreme impression of that person or event (Baron et al., 1997; Gilovich, 1987).

For example, Robert Baron and colleagues (1997) had a participant watch a videotape in which a young man described unintentionally getting drunk at a party, getting involved in a fight, and getting into a subsequent car accident. The man noted this was uncharacteristic of him, that he was egged on by friends, and that he regretted his actions. The participant rated the man on various positive and negative traits. Then the participant, now in the role of storyteller, was asked to speak into a tape recorder to describe the man’s story. Listeners then heard the audiotape and rated the man on the same traits. The listeners rated the man more negatively than the teller did. These effects seem to result both from a tendency of storytellers to leave out mitigating factors and complexities and a tendency of listeners to attend only to the central aspects of the stories they hear.

In addition to this tendency to tell simplified stories, our stereotypes of groups can also make us biased in our recall and retelling of information. Gordon Allport and Joseph Postman (1947) demonstrated this back in the 1940s in a study involving White American participants. They briefly showed a person a picture depicting a scene on the New York subway involving a White man standing, holding a razor, and pointing his finger at a Black man (FIGURE 3.6). They then had that person describe the scene to another person who had not seen the picture. That second person then described the scene to a third person, and so on, until the information had been conveyed to a seventh person. More than half the time, that seventh person reported that the scene involved the Black man, rather than the White man, holding the razor. So when we get our information filtered through lots of people, it’s pretty likely that prevalent schemas (such as stereotypes about a person’s group) have biased the information.

97

Mass Media Biases

Of course, we don’t get information only from having it told to us directly by others; we also learn a great deal from the stories we see and hear in the media. Just as rumors and gossip can distort the truth, media portrayals seldom are realistic accounts of what life is like, although they do provide vivid portrayals of possible scenarios. Think about your own schemas or scripts about dating and romantic love. Your earliest ideas about such matters probably came from fairy tales, television shows, the Internet, and movies. Unfortunately, these media offer biased views of many of these matters. For example, they tend to portray romantic relationships and love in oversimplified ways; to portray men, women, and ethnic groups in stereotypic ways; and to show a lot of violence (Dixon & Linz, 2000). The latter feature may explain why people who watch a lot of television think that crime and violence are far more prevalent in the world than they actually are (Shanahan & Morgan, 1999).

News programming is based on reality, and so people tend to assume it paints a fairly realistic, accurate, and less biased picture of events and people. But the news is created at least as much as it is reported. Those who produce the news choose which events and people to report about, and what perspective on the events to provide. These decisions are heavily influenced by concerns with television ratings and newspaper sales, and by the political and social preferences of those who own and sponsor the television programs, radio shows, newspapers, and magazines that report the news. And just as Allport and Postman showed over six decades ago, racial stereotypes can play a role as well. Just consider the two different descriptions of similar actions by people dealing with the aftermath of Hurricane Katrina’s devastation of New Orleans in 2005 (FIGURE 3.7). Both images show people leaving a grocery store and wading through floodwaters with food and supplies; the White people are described as “finding” food, whereas the Black individual is described as “looting.”

98

We now have a sense of how important schemas are in helping us to acquire and organize information about the people, ideas, and events that we encounter. But which of the many schemas stored in a person’s memory will be activated and shape thought and action at any given moment?

The person’s current situation plays a major role in activating particular schemas. If the characteristics of a social gathering across the street—loud music, alcohol in abundance—lead Yana to categorize it as a party, she will access her party schema, which gives her information she can use to figure out how else to think about this event, what inferences to make (“This ongoing noise will likely interfere with my social psych reading”), and what actions to take (“I should go to the library”). Of course, the more fine grained a person’s categories are, the more specific will be the schemas activated. Returning to the example, if Yana has different categories for a game-day keg party and a standard keg party, she can access different schemas in order to fine-tune her understanding of the event and her response to it. (“It’s a game-day keg party, so if our team loses the party will probably be over by 8 o’clock.”)

Accessibility refers to the ease with which people can bring an idea into consciousness and use it in thinking. When a schema is highly accessible, the salience of that schema is increased: it is activated in the person’s mental system, even if she is not consciously aware of it, and it tends to color her perceptions and behavior (Higgins, 1996). We just saw that the characteristics of the person’s current situation can increase the salience of a schema, making it more accessible for thinking and acting.

Priming occurs when something in the environment activates an idea that increases the salience of a schema. This happens because the information that we store in memory is connected in associative networks (FIGURE 3.8). These networks are tools that psychologists use to describe how pieces of information stored in a person’s memory are linked to other bits of information (Anderson, 1996; Collins & Loftus, 1975). These links result from semantic associations and experiential associations. Semantic associations result when two concepts are similar in meaning or belong to the same category. The words nice and kind, for example, have similar meanings; the words dog and cat both refer to household pets. Thus, we might expect them to be linked in a person’s associative network. Experiential associations occur when one concept has been experienced close in time or space to another concept. For example, for many people who consume their fair share of television or live in high-crime areas, guns are experientially associated with violence. Through these two kinds of mental links, priming, or “turning on,” one idea will bring to mind other ideas that are closely linked in a person’s associative network, but will be less likely to bring to mind ideas that are not strongly linked.

99

In addition to the immediate environment and priming, the person’s personality determines how accessible certain schemas are. Chronically accessible schemas are those that represent information that is important to an individual, relevant to how they think of themselves, or used frequently (Higgins, 2012; Markus, 1977). Such schemas are very easily brought to mind by even the most subtle reminder. For example, Mary is really interested in environmental issues, whereas Wanda is attuned to contemporary fashion. Mary is more likely to notice a hybrid car in the parking lot or express disdain over the plethora of plastic cups lying around at a party. Meanwhile, Wanda has her fashion radar working and her associated constructs chronically accessible, so she may dislike the tacky cups and be more likely than Mary to notice that Cynthia arrived in last season’s designer shoes. Even though they are in the same situation, the differences in what schemas are chronically accessible for Mary and Wanda lead to very different perceptions and judgments of the scene.

People are also likely to interpret others’ behavior in terms of their own chronically accessible schemas (Higgins et al., 1982). If you read a biography of Herman Melville and if honesty is a chronically accessible trait for you, you would be likely to have a good memory for incidents in Melville’s life that pertain to honesty, and your overall impression of Melville will be largely colored by how honest he appears to have been. On the other hand, if kindness is chronically accessible for you, his incidents of kindness or unkindness would be particularly memorable and influence your attitudes.

Situational and chronic influences on schema accessibility also can work together to influence our perceptions of the world. For example, after witnessing a fellow student smile as a professor praises her class paper, students in one study were more likely to rate her as conceited if they had very recently been primed with words related to the schema arrogance, but this effect was most pronounced for students who showed high chronic accessibility for the schema conceitedness (Higgins & Brendl, 1995). In other words, a certain situation or stimulus may prime particular schemas for one person but not for another, depending on which ideas are chronically accessible to each (Bargh et al., 1986).

Priming and Social Perception

Now that we have some understanding of how schemas operate, let’s examine in a bit more detail the consequences that schemas have for social perception and behavior. Imagine that you recently met a fellow named Donald. You learn that Donald is the type of person you might see in an energy-drink commercial with a penchant for extreme activities—skydiving, kayaking, demolition derby—and who is now thinking of mountain climbing without a harness. What do you think your impression of him would be? Would you see him as reckless or adventurous?

100

A study by Higgins and colleagues (1977) suggests that your impression will depend on the traits that are accessible to you before you met him. Participants in this study were told they would be completing two unrelated studies on perception and reading comprehension, but in actuality the tasks were related. In the “first” study, participants performed a task in which they identified colors while reading words (commonly referred to as a Stroop task). In this task, you might see the word bold printed in blue letters, and your job would be to identify the color blue. This task gives the researchers a way to make certain ideas accessible for some participants but not for others. Half of the participants were randomly assigned to read words with negative implications (e.g., reckless). The other half of the participants read words with positive implications (e.g., adventurous). In the “second” study, participants were asked to read information about a person named Donald who takes part in various high-risk activities, and to answer some questions about their impression of Donald. You can see from FIGURE 3.9 that the words participants were primed with during task 1 had a dramatic effect on the impressions they formed of Donald. Those participants who had previously read negative words pertaining to recklessness were likely to form more negative impressions of Donald, whereas those participants who had previously read positive words pertaining to adventurousness tended to view Donald more positively. Their impressions differed despite the fact that they were presented with identical information about Donald! What led to these different impressions, of course, were the different ideas that were primed in the participants before they read about him. This finding suggests that our impressions of others are shaped by salient schemas.

Priming and Behavior

Just like impressions, social behavior can be influenced by recently primed schemas without the person being consciously aware of their influence. Consider this scenario. You show up to participate in a psychology study, thinking that it concerns language proficiency. You are asked to complete a task in which you try to unscramble words to make sentences. You’re told to use four of the five words presented. You start on the task and are presented with they/her/bother/see/usually. So you start scribbling something like “they usually bother her” and then proceed to the next set of words. Unknown to you, you have been randomly assigned to be in a condition in which words related to the schema rudeness have been primed (notice the word bother). Other participants were presented with neutral words or words related to the schema politeness (e.g., respect).

Study results suggest that they would (Bargh et al., 1996). When no category was primed, 38% of participants interrupted within a 10-minute time frame. But among those primed with rudeness, 64% were too impatient to wait that long, whereas only 17% of those primed with politeness-related words interrupted. Schemas that are primed in one context can shape behavior in a different context.

101

The Role of the Unconscious

Why are priming studies useful? They illustrate how our experiential system can operate behind the scenes, influencing our everyday thought and behavior outside of our awareness. Indeed, psychologists from Freud (1923/1961b) to Wilson (2002) have made the point that consciousness is the mere tip of an iceberg: We are continually influenced by features of the environment and mental processes without being aware of them.

Of course, this idea is not completely new in popular culture. Over the years there have been a number of controversial media accounts of subliminal priming. For example, in 1957, a movie theater proprietor claimed to have boosted popcorn and soda sales at concession stands by presenting subliminal messages encouraging patrons to visit the snack bar. This was later discovered to be false because no such messages were actually presented, but it certainly raised the ire of many moviegoers at the time. In 1990, the heavy metal band Judas Priest was sued over purportedly presenting subliminal messages in one of their songs that encouraged a young man to commit suicide.

Although these examples turned out to be groundless, subliminal priming is a reality. It’s just that now we understand how and when subliminal primes are likely to influence thought and behavior. Part of this development is owing to advances in technology, because we now have the means to present information (such as words and pictures) precisely long enough to activate them in the mind without bringing them into conscious attention. (In most experiments, exposures range from 10 to 100 milliseconds.)

We can see the effect of such subliminal priming in a study by Bargh and Pietromonaco (1982). They had participants read about a person named Donald (no relation to the earlier mountain climber) who refused to pay his rent until the landlord painted the apartment. Prior to reading about Donald, participants completed a computer task in which they were asked to identify where on the screen a brief flash appeared. Unknown to participants, directly following the flash a word was presented for 100 milliseconds to the periphery of their visual focus, followed by a string of “XXX”s that served to cover (or mask) the stimulus. Participants were shown 100 flashes. If participants were exposed to a lot of words related to hostility (curse, punch), they judged Donald to be more aggressive than did participants exposed to only a few or no hostility-related words. Yet no participant reported being aware of having seen the words, which suggests the power of subliminal priming.

We now also have a much better theoretical grasp of how subliminal priming works, and therefore when it will—and will not—be likely to influence thought and behavior. We now know that subliminal primes do not lead people automatically and robotically to do whatever it is they are told to do, such as buy a soda or commit suicide. Rather, the concept of accessibility that we’ve been discussing suggests that subliminal priming makes some ideas more accessible than others, but they still are only one factor that determines thought and behavior. Other factors include the ideas and goals made salient by the environment and those that are chronically accessible for the person (Strahan et al., 2002). This means that a primed schema will be more likely to tip interpretation of an ambiguous event one way or another, rather than reverse long-standing attitudes. With regard to behavior, a primed schema might subtly nudge a person to respond more aggressively to a stressful situation. But it will not transform a normally peaceful individual into a total jerk.

Assimilation and Contrast

The priming effects described so far are known as assimilation effects. This is because the judgment of the person or event is assimilated in, or changes in the direction of, the primed idea. For instance, priming the schema reckless increased the perception of mountain climbing Donald as reckless. But primes sometimes have the opposite consequence, leading to contrast effects. For example, in some studies, priming the schema hostility led people to view a person as less hostile (Lombardi et al., 1987; Martin, 1986). Although there are some complexities to determining when assimilation effects and when contrast effects are likely to occur (Higgins, 1996), contrast effects seem to emerge under a few conditions. The first is when people are very aware of the primed information and that it might affect their subsequent judgments. Accordingly, assimilation effects are consistently found for subliminal and subtle primes, but contrast effects are common when the primes and their relation to the subsequent judgment are very obvious. In these situations, people’s conscious cognitive system often tries to counteract the potential influence of the prime by shifting their judgment or behavior in the direction opposite of that implied by the prime (Wegener & Petty, 1995).

102

Two other conditions in which contrast can occur are when the prime is extreme or when it evokes a specific example of a category (Dijksterhuis et al., 1998; Herr, 1986; Schwarz & Bless, 1992). For example, when rating oneself after being primed with an extreme example, it may be harder to view oneself as consistent with the category. And when primed with a specific person who fits the category, the individual is more likely to compare the self with that specific person. In one study (Dijksterhuis et al., 1998), when students were asked to think about professors in general, they performed better on a test of general knowledge than students asked to think about supermodels in general (an assimilation effect). However, when participants were asked to think about specific exemplars of those categories (Albert Einstein as a professor and Claudia Schiffer as a supermodel), the specific exemplars caused participants to compare themselves with the exemplars, leading to a contrast effect in which those primed with Albert Einstein did worse than those primed with Claudia Schiffer. After all, it’s rather difficult to think of oneself as smart when compared with Einstein!

Schemas and the expectations and interpretations that they produce are generally quite useful. Your party schema tells you what to expect there, how to dress, and so forth. Your mom schema helps you predict and interpret things your mom will say and do. And the schemas that become active in particular situations are usually the ones most relevant to that situation. However, once we have a schema, we tend to view new information in such a way as to confirm what we already believe or feel. This is known as confirmation bias. In chapter 1 (pp. 13-14) we saw how this bias influenced students’ evaluations of an article on capital punishment (Lord et al., 1979). Confirmation bias helps people preserve their worldview by sustaining a stable, consistent set of beliefs and attitudes about the world. In this way, it provides the individual with psychological security. However, confirmation bias also often leads to inaccurate interpretations of new information.

This effect happens for a number of reasons. First, once we have a schema of a person or situation, that schema guides us to look for certain kinds of information and ignore other kinds of information. We see this demonstrated in a study by Snyder and Frankel (1976). Participants watched a silent videotape of a woman being interviewed. They were told that the interview was about either sex or politics. Participants were told to watch the videotape to assess the woman’s emotional state. When participants thought the interview was about sex, they rated her as more anxious than when they thought the interview was about politics. The videotape was the same in both cases, but when participants thought the topic was sex, they expected the woman to be anxious over discussing such a personal topic, and therefore they watched more closely for nonverbal signs of anxiety. You’ve heard the expression “Seeing is believing”; studies such as these suggest that the converse holds true as well: “Believing is seeing”!

103

Second, a salient schema leads us to interpret ambiguous information in a schema-confirming manner. In one study, trained therapists watched a videotaped interview with a man. Half the therapists were told it was a job interview, and the other half were told it was an interview with a mental patient (Langer & Abelson, 1974). Although everyone listened to the same interview, therapists who thought the man was a mental patient saw more signs of mental illness than those who thought he was a job applicant. When the interviewee described conflicts with his bosses in past jobs, those who thought he was a mental patient tended to interpret his actions as stemming from his defensiveness, repression, and aggressive impulses. Those who thought it was a job interview interpreted the same actions as signs of perceptiveness and a realistic perspective.

The Ironic Biasing Influence of Objective Information

Could this insidious schema-based confirmation bias actually cause objective information to do more harm than good? To find out, Darley and Gross (1983) had participants watch one of two versions of a videotape about a nine-year-old fourth grader named Hannah, showing her playing in a playground, along with scenes of her neighborhood and school. The videotapes made it clear that Hannah had either an upper-class or lower-class background. Darley and Gross reasoned that participants shared the common schema of upper-class kids as academically successful and the common schema of lower-class kids as unsuccessful.

Half the participants (the no-performance group) were then simply asked to rate Hannah’s academic abilities on a scale ranging from kindergarten to sixth-grade level. The other half (the performance group) were shown a second videotape, which was the same whether Hannah was earlier depicted as upper or lower class, before being asked to rate Hannah. This videotape showed Hannah performing on an oral achievement test, answering questions ranging from easy to hard, doing well on some and not well on others.

Which group do you think was especially likely to be influenced in their ratings by Hannah’s socioeconomic status—the no-performance group or the performance group? We might expect participants given only class-based schemas to rate Hannah higher if they thought she was upper rather than lower class. However, one would hope that participants provided with objective evidence of Hannah’s academic abilities would rely on that information and ignore the class-based schemas.

And yet the opposite occurred, as we see in FIGURE 3.10. The objective evidence increased the bias rather than decreasing it. The group that didn’t have the opportunity to see Hannah perform estimated her math abilities to be the same regardless of whether she was upper or lower class. They seemed to realize that they didn’t have much basis for prejudging her abilities after only seeing her on a playground. However, the group that observed Hannah take an oral achievement test rated her much better if she was upper rather than lower class. These participants saw Hannah’s performance and rated her abilities in line with what they expected from a student of her social class. The point is that the participants didn’t interpret the so-called objective evidence objectively; instead, they interpreted it as confirming what they already believed they knew about Hannah’s ability.

104

Biased Information Gathering

People’s schemas, even when tentative, can also lead to biased efforts to gather additional information, efforts that tend to confirm their preexisting schemas. Participants in one study had a brief discussion with a conversation partner who was described to them as being an extravert or an introvert (Snyder & Swann, 1978). Their job was to assess whether this was true, and they were given a set of questions to choose from to guide their conversation. Participants tended to ask the conversation partner questions that already assumed the hypothesis was true and would lead to answers confirming the hypothesis. For example, a participant wanting to determine if the partner was an extravert chose to ask questions such as, “What kinds of situations do you seek out if you want to meet new people?” and “In what situations are you most talkative?” However, if they wanted to determine if the partner was an introvert, they chose questions such as, “What factors make it hard for you to really open up to people?” and “What things do you dislike about loud parties?” What’s important to note here is that these are leading questions: When answering a question about how she livens up a party, for example, a person is very likely to come across as extraverted, even if she is not; likewise, even an extravert will look introverted when talking about what he dislikes about social situations. This study shows that people tend to seek evidence that fits the hypothesis they are testing rather than also searching for evidence that might not fit that hypothesis.

These effects could have important implications for how clinical psychologists diagnose disorders. In one series of studies, participants were shown a set of drawings of a human figure along with a psychological symptom of the person who drew each picture (Chapman & Chapman, 1967). After viewing all of the pictures, participants were asked to draw conclusions about whether people who share the same symptoms have a tendency to draw certain features of a person in a distinctive way. In a sense, they were given the opportunity to play amateur therapists who use drawings to uncover people’s psychological issues. And their responses showed a great deal of convergence: Participants often concluded that people with paranoid tendencies drew unusual eyes in their pictures, and men who were worried about their masculinity drew images with broader shoulders and more muscular physiques. However, unknown to the participants, the researchers had randomly paired each symptom with a picture so that there was no true correlation between these aspects of the drawings and the mental issues they imagined for the artists. Rather, they saw in the pictures what they expected to see given their expectations for paranoid or worried types. Follow-up studies showed that these biases are present even among experienced clinicians (Chapman & Chapman, 1969). When people do not try to actively disconfirm their expectancies for others, they run the risk of seeing only what they already believe.

The Self-fulfilling Prophecy

Another vivid testament to the power of schemas is evidence that they not only bias our perceptions of social reality, but can also create the social reality that we expect. More specifically, people’s initially false expectations can cause the fulfillment of those expectations, a phenomenon that Robert Merton (1948) labeled the self-fulfilling prophecy. To investigate this idea, Robert Rosenthal and Lenore Jacobson (1968) went to an elementary school in 1964 and administered some tests to the students. After scoring the tests, they gave the teachers the names of some kids in their class who, according to the Harvard Test of Inflected Acquisition, were on the verge of experiencing a substantial leap forward in their general learning abilities. The teachers were told that these kids were “late bloomers” who were about to display an “intellectual growth spurt.”

105

Two years later, the kids labeled as late bloomers actually scored substantially higher than their classmates did on a test of general abilities. However, unknown to the teachers, the list of kids originally labeled late bloomers was a random selection from the class rosters. So the only reason they experienced a dramatic intellectual growth spurt was that the teachers were led to expect they would!

What accounts for this self-fulfilling prophecy? Years of additional research have revealed that although such effects don’t always occur, when they do, it is because teachers’ expectations affect their behavior toward the students in ways that improve the students’ learning (Rosenthal, 2002). For example, kids expected to do well are given more attention and more nods and smiles, are challenged more, and are given more positive reinforcement for their successes (e.g., Harris & Rosenthal, 1985; Jussim, 1986). Students tend to respond to such behavior with more engagement and more effort, and consequently, more learning. One study also showed that these expectations can work in the opposite direction: If students expect a teacher to be excellent, the teacher performs better (Feldman & Prohaska, 1979).

Since that classic study on teachers and students, self-fulfilling prophecies have been demonstrated in many other contexts as well (e.g., Snyder et al., 1977). If you expect someone to be friendly and sociable, you are likely to act in ways that elicit such behavior. If you expect someone to be unpleasant and annoying, you are likely to act in ways that provoke that kind of behavior. One study found that army platoon leaders led to expect their platoon to be made up of exceptional recruits actually produced better soldiers (Eden, 1990). Mere expectations won’t turn a serial killer such as Jeffrey Dahmer into a humanitarian such as Nelson Mandela, but most of us are capable of being pleasant or unpleasant, industrious or indifferent. Within a moderate range of variability, it seems quite clear that perceivers’ expectations about others often shift people’s behavior toward confirming those expectations.

Limits on the Power of Confirmation Biases

We have beaten the drum for confirmation bias very loudly in this section, and the large body of evidence warrants doing so. However, confirmation biases do not always occur. If people’s observations clearly conflict with their initial expectations, they will revise their view of particular people and events. This is especially likely if the gap between what people expect and what they observe is very extreme. For example, if you play chess with a nine-year-old and don’t expect the child to show much skill, and then the kid beats you, you will likely revise your opinion of the child’s chess ability. In fact, because your expectation was so different from the outcome, you might even overrate the child’s ability.

It is interesting, though, that even in such cases, people usually grant the exception but keep the underlying schema. In the chess example, you’d probably think, “This kid’s a genius, but most nine-year-olds stink at chess.” Of course, if enough nine-year-olds whip you in chess, the schema eventually would give way to the data.

In addition, as we noted in the section on priming effects, when people are aware of and concerned about being biased, their cognitive system may kick in to correct the feared bias. Another way to think about this correction is to say that people’s need for accuracy trumps their need for closure, leading them to think more carefully—or at least to respond in a way that is opposite to what they think is a biased judgment. However, the evidence suggests that this correction process tends to be inexact and sometimes leads people to bend over backward in the opposite direction. Finally, in the context of self-fulfilling prophecies, if the target of your expectation knows you think a certain way about him or her, the person may go out of the way to try to disconfirm your expectation (Hilton & Darley, 1985).

106

So far we’ve focused on people’s use of schemas. It makes intuitive sense that people think about a thing by applying their accumulated knowledge about other things that are like it. But do people ordinarily use other cognitive devices to make meaningful sense of the social world? To find out, listen to how people commonly talk about the abstract ideas that matter in their daily lives:

I can see your point (understanding is seeing)
I’ll keep that in mind (the mind is a container)
Christmas is fast approaching (events are moving objects)
That is a heavy thought (thoughts are objects with weight)
I feel down (feelings are vertical locations)
I devoured the book, but I’m still digesting its claims (ideas are food)
Her arguments are strong (arguments are muscle force)
I’m moving forward with the chapter (progress is forward motion)
The economy went from bad to worse (states are locations)

These are metaphoric expressions because they compare things that, on the surface, are quite different. (These comparisons are reflected in the statements in parentheses.) That is why these expressions do not make sense if taken literally. For example, feelings do not have an actual vertical location, and arguments cannot have muscle strength. According to many philosophers and psychologists, such metaphoric expressions are more than merely colorful figures of speech; instead, they offer a powerful window into how people make sense of abstract ideas.

107

From this perspective, metaphors are cognitive tools that people use to understand abstract ideas by applying their knowledge of other types of ideas that are more concrete and easier to understand (Kövecses, 2010; Lakoff & Johnson, 1980). For example, when Lisa says, “Christmas is fast approaching,” she may be using her knowledge about moving objects to conceptualize time. Why? Because Lisa may find it difficult to get a clear image of time in her mind (not surprising, since physicists aren’t sure what time is!). Yet she has a concrete schema for physical objects moving around, and this schema tells her that objects tend to be more relevant as they draw closer. By using her objects schema to think about time, Lisa can make sense of what an “approaching” event means for her (time to buy gifts!), even though there is no such thing as an event moving toward her.

How does this perspective enhance what we know about social cognition? It suggests that people’s everyday efforts to construct meaning draw on metaphors as well as schemas. Whereas schemas organize knowledge about a given idea, metaphors connect an idea to knowledge of a different type of thing. Often, we construct metaphors around things that are connected to our bodily experiences. For instance, people understand morality partly by using a schema that might contain memories of moral and immoral individuals and behaviors. But people also understand morality metaphorically in terms of their bodily experiences with physical cleanliness and contamination (Zhong & House, 2013). This metaphor is reflected in common expressions such as, “Your filthy mind is stuck in the gutter; think pure thoughts with a clean conscience,” and it operates at a conceptual level to shape how we make judgments about morality.

108

Researchers have tried to go beyond analyzing language to learn more directly whether people use metaphor to think about abstract ideas. In one procedure, participants are primed with a bodily experience, such as tasting something, seeing something, or feeling something’s texture. Then, in an apparently unrelated task, they are asked to make judgments or decisions about an abstract idea. The researchers reason that if people in fact use a bodily experience to understand an abstract idea, then the prime should produce parallel changes in those judgments and decisions. To illustrate, if people understand love metaphorically as a journey (“Our relationship is moving forward”), then priming them with the bodily experience of journeying over rocky terrain (versus smooth terrain) should lead them to expect to encounter conflicts as their love relationships progress. Alternatively, if people do not use the metaphor love is a journey, then we wouldn’t expect that priming experiences of physical journeys would influence their judgments and decisions about love.

Williams and Bargh (2008) used this procedure to examine the metaphorical link between physical and interpersonal warmth. They built on prior evidence that people commonly refer to interactions with others by using the concepts warm and cold (Asch, 1946; Fiske et al., 2007), as when one receives a warm welcome or a cold rejection. To determine whether this metaphor influences social perceptions, they had the experimenter—who apparently needed a free hand—ask participants to hold her coffee cup. Depending on condition, the cup was either warm or cold. Afterward, all participants were asked to read a brief description of another person and rate that person’s friendliness and trustworthiness—that is, the person’s interpersonal warmth. As predicted, participants who simply held a warm (versus a cold) beverage perceived a target individual as friendlier and more trustworthy, suggesting that conceptual metaphors can influence social perceptions even when people are not prompted to use metaphoric language.

Similar effects have now been found in dozens of published studies (see Landau et al., 2010; Landau et al., 2013). Subtle primes of bodily experience influence how people perceive, remember, and make judgments and decisions related to a wide range of abstract social concepts. To mention just a few surprising findings: Weight manipulations influence perceived importance; smooth textures promote social coordination; hard textures result in greater strictness in social judgment; priming closeness (vs. distance) increases felt attachment to one’s hometown and families; groups and individuals are viewed as more powerful when they occupy higher regions of vertical space.

These findings highlight an important fact about the way we make sense of the world: We construct an understanding of abstract ideas by drawing on our knowledge of the sensory and motor experiences we have had from the earliest moments of life (Mandler, 2004; Williams et al., 2009).