Personality and Social Interaction

We begin with a reminder of two fundamental facts of human nature:

  • Humans are intelligent creatures. We do not operate merely by reflex, taxis, instinct, and conditioned response. Rather, our behavior reflects active cognitive processes of perceiving, learning, remembering, thinking, and communicating through language.
  • But humans are also social creatures. Our experiences, thoughts, and actions take place in an explicitly social context of cooperation, competition, and exchange, family and group memberships, institutional, social, and cultural structures.

For that last reason, psychologists need to understand the relations between psychological processes within the individual and social processes that take place in the world outside. While much of cognitive psychology is concerned with how the individual acquires, represents, and uses knowledge in perception, memory, thought, and language, social psychology is concerned with the role that cognitive (as well as emotional and motivational) processes play in social interactions -- between individuals, and between individuals and groups.

Analyzing Social Interaction

              Social InteractionThe classic framework for the analysis of social behavior was provided by Kurt Lewin, a refugee from Hitler's Europe who settled in the United States in the 1930s. Lewin sought to apply the principles of Gestalt psychology to the study of social behavior. The Gestalt school is known for its assertion that "the whole is greater than the sum of its parts". Applied to perception, this means that perception encompasses the entire stimulus field. Individual stimulus elements form a coherent, integrated whole, and cannot be isolated from each other. Similarly, Lewin argued that social behavior is responsive to the entire field of social stimuli -- not just the other person immediately present, but also the wider social context in which the interaction occurs. Lewin went even farther to assert that the social situation includes the person him- or herself: the person is part of the stimulus field to which he or she responds.

Employing the language of mathematics, Lewin asserted that:

B = (P, E)


B = the individual's overt behavior: behaviors that are publicly observable;

In other words... P = personal determinants: mental (cognitive, emotional, and motivational) states and dispositions residing within the individual's mind, such as beliefs, feelings, motives, traits, and attitudes.

is a function of both

personal and environmental


E = environmental determinants: factors impinging on the individual from outside, including aspects of the physical ecology (temperature, humidity, altitude, etc.) and aspects of the sociocultural ecology (the presence and behavior of other people, constraints imposed by social structures, social roles, situational demands and expectations, social incentives, etc.

This bit of pseudo-mathematics represents the idea that personal and environmental determinants combine somehow to cause individuals to do what they do. The comma in the equation indicated that Lewin was open as to precisely how these factors combine. Lewin was heavily influenced by Gestalt psychology, and believed that the person and the environment constituted an organized and integrated field in which behavior takes place.

The Americanization of Kurt (and Muzafer)

There were social psychologists before him, but Kurt Lewin, with his formula B=f(P,E), is universally regarded as the intellectual godfather of social psychology. Accordingly, successive generations of graduate students have struggled with the pronunciation of his name. The proper German pronunciation is, of course, "Luh-veen'. But when Lewin emigrated to the United States he Americanized it to "Loo-win". Later in life, his increasing interest in his German-Jewish background led him to attempt to revert to the original German, but the Americanization stuck. The adjectival form, though is the more Germanic "Luh-vin-ian".

Something similar happened to Muzafer Sherif, discussed in the Introduction as the person who completed the development of psychology as an experimental science by studying social influences on perception. Sherif was of Turkish heritage, and so the proper pronunciation is "Sher-riff", like the law-enforcement officer, except with the "r" rolled (a little like Scottish brogue). And those who knew him said that, after he emigrated, he kind of liked that image of him, as a sheriff in the Wild West. But most Americans pronounce his name "Sher-reef", like the actor Omar Sherif (who was, in any event, Egyptian).

Defining the Subdisciplines

As subfields within psychology, personality and social psychology have historically emphasized different aspects of Lewin's formula.

Personality Psychology

Traditional personality psychology assumes that behavior is primarily determined by features of the person such as his or her beliefs, attitudes, values, emotions, motives, and traits, and that situational factors are largely irrelevant.

B =f(P).

This viewpoint is exemplified by traditional research on personality traits such as friendliness or aggressiveness.

So, for example individual differences in friendliness, assessed by means of a self-report questionnaire, would be used to predict whether a person would smile in some situation.In this research, which often uses the technique of multiple regression analysis (a variant on the correlation coefficient), the trait measure (e.g., friendliness) serves as the predictor variable, and the behavioral measure (e.g., smiling) serves as the criterion variable.

The canonical method of traditional personality psychology is to construct a "psychological test" to measure some personality trait, and then to use this information to predict individual behavior in some specific situation. The test might take the form of a self-report questionnaire, a rating scale (completed by the subjects themselves or by others who know them well), or even a sample of actual behavior.

The Doctrine of Traits

The Doctrine of TraitsThe canonical method of personality psychology exemplifies the doctrine of traits, derived to a great extent from the work of Gordon Allport (1937):

Social behavior varies as a function of internal dispositions that render it coherent, stable, consistent, and predictable.

These dispositions are commonly studied in the form of traits and attitudes. However, other dispositions are also relevant to behavior, such as
 moods, motives, values, and beliefs.

Conscientiousness and PunctualityHere's an actual research example: as part of an ongoing study at UCB, Ware and John (1995) measured MBA students' level of conscientiousness with a questionnaire known as the NEO Personality Inventory (to be discussed below), and then tracked their arrival times at two appointments, one in the morning and one in the afternoon. Collapsing across the two appointments, they found that highly conscientious individuals tended to arrive a little early, while those who scored low on conscientiousness tended to arrive about five minutes late. In this case, conscientiousness significantly predicted a criterion of punctuality. (We'll return to this study later.)

A similar method is used to study the relationship between attitudes and behavior. For historical reasons, having largely to do with an interest in attitude change in response to persuasive communications, attitudes have primarily been studied by social psychologists. But the logic is the same: attitudes, which are internal dispositions to evaluate certain objects or ideas positively or negatively, are held to cause the individuals who hold these attitudes to behave in particular ways. Thus, in the 2000 election, a registered Democrat was (probably) more likely to vote for Al Gore than for George Bush -- though not enough of them did, from Gore's point of view.

Personality Theory

The following links connect to drafts of chapters prepared for an introductory survey to the psychology of personality. The chapters were drafted in approximately 1990. The project was eventually aborted -- but because these chapters were intended to survey the "traditional" psychology of personality, which in turn was based largely on the Doctrine of Traits, they still provide a useful summary of "classic" personality theory and research up to that time.

Link to a 50-year retrospective of personality psychology since Allport's seminal 1937 text, published as Kihlstrom, J.F. (1988). Personality. In E.R. Hilgard (Ed.), Fifty years of psychology: Essays in honor of Floyd Ruch (pp. 139-152). Glenview, Illinois: Scott, Foresman.

Link to a commentary on the relationship between classical psychoanalysis and modern personality and social psychology, published as Kihlstrom, J.F. (1994). Psychodynamics and social cognition: Notes on the fusion of psychoanalysis and psychology. Journal of Personality,62, 681-696.

Social Psychology

Traditional social psychology, by contrast, assumes that behavior is primarily determined by features of the environment, and especially features of the sociocultural ecology, such as interpersonal, organizational, and cultural factors, and that individual differences in personality are largely irrelevant.

B =f(E).

This viewpoint is exemplified in social psychology by traditional research on social influence, or the effects on behavior of the presence or behavior of other people. It is very congenial to the behaviorism espoused by John B. Watson and B.F. Skinner -- that latter of whom, famously, wrote (in Beyond Freedom and Dignity, 1971) that "a person does not act upon the world, the world acts upon him".

              Social Psychology So, for example, we might arrange an encounter between a subject and an acquaintance or a stranger, and see if smiling occurs more often in one situation than the other. In this research, which often uses the technique of analysis of variance (a variation on the t-test), the manipulated variable (e.g., presence of acquaintances or strangers) serves as the independent variable, and the observed variable (e.g., smiling) serves as the dependent variable.

The canonical method of traditional social psychology is to manipulate some aspect of the social environment (such as whether behavior is private or public, or whether the subject receives information about other people's attitudes and opinions), and observe the effects of this manipulation on behavior in some specific situation. All subjects might be exposed to all conditions (this is known as a within-subjects design), or different groups of subjects might be randomly assigned to each condition (this is known as a between-groups design).

              Time and PunctualityReturning to the Ware & John (1995) study, recall that punctuality was measured in two different situations -- an appointment in the morning and another in the afternoon. When they examined punctuality as a function of the situation, rather than as a function of the person, they found that the students tended to arrive late for morning appointments, but a little early for appointments in the afternoon. In this case, punctuality depends on a feature of the situation -- the time of day at which the appointment is scheduled.

Just to confuse things, sometimes in regression analyses, the predictors are labeled as independent variables, and the criteria are labeled as dependent variables. This is because, mathematically, multiple regression is formally equivalent to the analysis of variance.

Link to an interview with Elliot Aronson.

The Doctrine of Situationism

The Doctrine
                of SituationismThe canonical method of experimental social psychology exemplifies the doctrine of situationism, sometimes attributed to Lewin himself, but more closely related to the behaviorism espoused by John B. Watson and B.F. Skinner:

Social behavior varies as a function of features of the external environment, particularly the social situation, that elicit behavior directly, or that communicate social expectations, demands, and incentives.

As Skinner himself put it (in Beyond Freedom and Dignity, 1971), "a person does not act upon the world, the world acts upon him".

The relevant features of the situation may be found in the external physical environment. More likely, though, they are to be found in the external social environment, such as the presence and activities of other people, social demands, and social rewards.

The Leading Edges of Social Psychology

Link to the report of a working group of the National Research Council, projecting the future of social psychology as the 20th century was drawing to a close, published as Working Group on Social Interaction [Berscheid, E., Darley, J., Hastie, R., Jones, E.E., Kelley, H., Kihlstrom, J.F. (chair), & Stryker, S.]. (1989). Social interaction. In R.D. Luce, N.J. Smelser, & D.R. Gerstein (Eds.), Leading Edges in Social and Behavioral Science (pp. 137-159). New York: Russell Sage Foundation.

Putting the Person and the Environment Together

However, these traditional formulations are largely misleading. Nobody believes that one factor is exclusively responsible for behavior, and the other is wholly irrelevant. Dispositional and situational factors probably combine somehow to cause behavior to occur.

              and Environmentl Factors Are Independent One possibility is that P and E are independent -- that is, that each set of factors exerts its own separate influence on behavior, without affecting the other in any way. In such a situation, behavior is partly predicted by personality traits, and partly affected by situational manipulations. In mathematical terms, personal and environmental factors are additive:

B =f(P +E).

If P and E are independent:

  • the effect of the personality variable is the same, regardless of the situation the person is in;
  • and the effect of the situation is the same, regardless of the kind of person in that situation.

In statistics, these two effects would be characterized as the main effect of personality and the situation, respectively.

Thus, friendly people may smile more than unfriendly people, and people may smile more at acquaintances than at strangers, but the difference between friendly and unfriendly subjects is constant across the two situations (subtract the means), and the difference between acquaintance and stranger targets is constant across levels of friendliness (again, subtract the means). Statistically speaking, there is no interaction between these main effects.

Conscientiousness, Appointment Time, and PunctualityReturning to our study of punctuality, when Ware and John (1995) examined the data as a function of both the person (level of conscientiousness) and the situation (time of day), they found that the effects of the two variables were completely independent of each other -- as indicated by the almost perfectly parallel lines. But another way, punctuality varied as an additive function of conscientiousness and time of day,

For most of the 20th century, personality and social psychology proceeded largely independently of each other, but in the 1960s and 1970s there arose a trait-situation controversy over which factors were more powerful predictors of behavior -- internal traits or external situations. This debate devolved into a contest over whose "effect size" was bigger -- and, like most characteristically masculine competitions, it was essentially a pointless exercise, and generated much more heat than light. It's over now, and in most psychology departments personality and social psychologists work side by side -- though they still keep their hands on their swords.

The Doctrine of Interactionism

Personal and
              Environmental Factors InteractThe trait-situation controversy faded partly due to exhaustion of the participants, but also because psychologists began to consider a more interesting possibility -- that personal and environmental determinants interacted with each other in a variety of ways.

The Doctrine of
              Interactionism The Doctrine of Interactionism proposed by K.S. Bowers (1973), holds that people influence the situations that, in turn, influence their behavior. As he put it:

An interactionist... view denies the primacy of either traits or situations in the determination of behavior....

More specifically, interactionism argues that situations are as much a function of the person as the person's behavior is a function of the situation.

The Doctrine of Interactionism

The Doctrine of Interactionism was originally intended to counter the Doctrine of Situationism:

Environmental factors are not independent of personal factors. Rather, people can influence their environments, just as their environments influence their behavior.

Interactionism agrees that people's behavior is influenced by the situations in which they find themselves. But because it views people as part of the environment, it holds that personal factors of the sort envisioned in the doctrine of traits can still play an important role in behavior.

From an interactionist perspective, different kinds of people show different patterns of response across different situations. In mathematical terms, personal and situational factors are multiplicative:

B =f(P x E).

If P and E interact:

  • The effect of the personality variable depends on the situation the person is in; and
  • The effect of the situation depends on the kind of person who is in it.

Thus, for example, friendly people might smile more than unfriendly people, but this difference would be bigger when they encounter a stranger than when they encounter a friend. Or, put another way, friendly people might discriminate less between the two situations that unfriendly people would. Such a situation is known statistically as the person-by-situation interaction.

The person-by-situation interaction takes a number of forms.

  • In the crossover interaction, there is no main effect of either person or situation (again, compute the means), but the difference between the two groups (or situations) reverses from one situation (or group) to the other.
  • In the fan effect, there is no difference between the two groups in one situation, but a big difference between them in the other.

There are other forms of interaction, all variants on these.

Here's an actual example of interaction, a study of behavior problems in toddlers. Elisabeth Conradt and her colleagues (2013) studied the children of mothers who were at risk for parenting problems of various sorts.  When the children were infants, just 5 months old, the researchers measured their respiratory sinus arrhythmia (RSA), a physiological index of activity in the parasympathetic nervous system.  Children with high levels of RSA tend to be more sensitive to, and engaged with, their environments.  One year later, at 17 months, the researchers assessed the environment in which the child was raised, as "secure" or "disorganized".  At the same time, they assessed the toddlers' levels of social and emotional problems. Children with low levels of RSA showed the same amount of problem behavior, regardless of the environment in which they were being raised.  But children with high levels of RSA were very responsive to their environments: those raised in disorganized environments showed a much higher frequency of behavioral problems than did those with low levels of RSA.  So the high-RSA children really are more responsive to their environments: those exposed to a disorganized environment showed it in relatively high levels of behavioral difficulties; those exposed to a more benign environment responded with relatively few behavior problems.

Reciprocal Determinism

Unidirectional Casuation in the Person-Environment
              InteractionTraditionally, the person-by-situation interaction has been characterized as unidirectional. That is to say, in Bowers' preferred formulation, people somehow affect the situations they are in, and this interaction, along with the main effects of the person and of the situation, causes behavior to occur.

              Relations in the Person - Environment Interaction (1) But of course, it may also be that, in addition to the person influencing the situation to which he or she responds, the environment also shapes the internal states and dispositions of the person in it.If people can influence their environments, why should their environments influence them?

The Doctrine of Reciprocal DeterminismAlbert Bandura has called this state of affairs reciprocal determinism. In reciprocal determinism, causality is bidirectional:

  • features of the person affect the environment; and
  • features of the environment affect the person.

The Doctrine of Reciprocal Determinism

The doctrine of reciprocal determinism is essentially a more dynamic extension of the doctrine of interactionism:

Just as people are part of their environment, so the environment is part of the person;

Because the person's behavior takes place in the environment, that behavior is also part of the environment, and part of the person.

Where interactionism asserts that people are a part of their own environment, reciprocal determinism asserts that people, their environments, and the behavior that takes place within those environments form a complex, dynamic, interlocking system characterized by nonlinear, bidirectional, causal relations.

In a very real sense, reciprocal determinism is the state of affairs envisioned by complexity theory (also sometimes known as chaos or catastrophe theory.

But if the causal relations between P and E can be bidirectional, why can't the causal relations between P and B, and between E and B, be bidirectional as well?

              Relations in the Person - Environment Interaction (2)In addition to personal factors such a traits and attitudes causing behavior to occur, behavior might feed back to change the person who emitted it.

Interactionism and Reciprocal DeterminismAnd, to complete the circle, in addition to environmental factors such as social pressure and rewards causing behavior to occur, behavior might feed back to change the environment that elicited it.

              Decomposition of Reciprocal DeterminismIn other words, each element in Lewin's formula -- not just P and E but B as well, is both cause and effect. Because the reciprocal causal relations involve three elements, Bandura also has labeled this expanded notion of reciprocal determinism triadic reciprocality.

Reciprocal determinism in general, and triadic reciprocality in particular, entails a very interesting situation in which everything is simultaneously both the cause and the effect of everything else.

  • Note, however, that reciprocal determinism does not necessarily mean symmetry of causal strength. Bidirectional causal influences are not necessarily coequal in strength. The causal effect of personal factors on behavior might well be stronger (or weaker) than the reciprocal causal effect of behavior on the person. In general, the relative strength of each of the six causal links (two in each direction between each pair of the three elements) may well vary across particular persons, environments, and behavior.
  • Nor does reciprocality mean simultaneity. When personal factors exert their effects on behavior, the behavior may not be influencing the person's states and dispositions at precisely the same time. More likely, as implied by the notion of feedback, these bidirectional influences unfold over time. The time course of this unfolding may also vary across persons, environments, and behaviors.

025Dialectics.jpg (71504 bytes)Still, when everything in a system affects everything else, things are going to get awfully complicated awfully quickly. For this reason, it is rarely possible to study triadic reciprocality in all of its dynamic glory. Fortunately, the lack of simultaneity works to our advantage, permitting an analytic decomposition in which we break triadic reciprocality down into its bidirectional segments, or dialectics (from the Greek word meaning "dialog").

This yields The Three Dialectics in Social Behavior:

P <--------------->B The dialectic between the person and his or her behavior.
E <---------------> B The dialectic between the environment and the behavior that occurs in it.
P <---------------> E The dialectic between the person and the environment in which his or her behavior takes place.

In what follows, we take up each of these dialectics in turn.

Reciprocal Determinism in Marital Relations

No, not those marital relations -- well, those too, actually, if you think about it, but here I'm really talking about conversations between married couples, as an example of reciprocal determinism in action.  The analysis comes from John Gottman, a psychologist at the University of Washington who specializes in marriage and family therapy (his laboratory is called "The Love Lab").  Gottman sets out his system in a monograph, The Mathematics of Marriage: Dynamic Nonlinear Models (2003). 

For our purposes, a complex dynamic nonlinear system is a system in which each element influences the state (behavior) of the other elements, just as in Bandura's analysis of reciprocal determinism, the elements P, E, and B are all influencing each other.  Gottman's insight is that married couples behave like this: each partner is influencing the other's beliefs, moods, desires, and actions. 

Gottman begins with the simple assumption that spouses engaged in conversation react to the most recent statement made by their partners.  Then, they map out influence functions -- for example, a biting remark by one spouse will induce a negative emotion in the other, and a negative response, leading to a downward spiral in affect.  Also important is each spouse's uninfluenced steady state -- essentially, whether each spouse is disposed to be in a positive or negative mood, independent of what the other person says or does.  Also, they take into account repair functions, in which spouses turn a difficult conversation into a more positive direction, or damping functions, which turn them in a more negative direction. 

From this point, Gottman creates bilinear influence functions tracing the course of the conversation, and each partner's effect on each other in turn.  For example, in validating couples tend to move in the direction of their partner's negative behavior, creating a vicious cycle.  In conflict-avoiding couples, one partner behaves more positively when the other is negative.  In hostile-detached couples, one spouse's positive behavior has little or no impact on the other. (Graphic from "Every Unhappy Family Has Its Own Bilinear Influence Function" by David Glenn, Chronicle of Higher Education, 04/25/2003).

The Dialectic Between the Person and Behavior

With respect to the dialectic between the Person and Behavior, analysis of the personal determinants of behavior focuses on internal states and dispositions:

  • Traits are behavioral dispositions, such as friendliness or aggressiveness.
  • Attitudes are evaluative dispositions, such as liberal or conservative, pro-feminist or antiwar.
  • Emotions are feeling states such as happiness or sadness.
  • Motives are states of desire, such as hunger and thirst, or achievement motivation or the need for affiliation.
  • Values are personal priorities, such as the valuing of happiness over money, art over science, or religion over economics.
  • Beliefs are convictions in the truth of certain statements, especially in the absence of definitive evidence -- such as that God exists or that life is sacred. Religious beliefs are perhaps the best examples of beliefs, precisely because they concern things that, almost by definition, can't be publicly observed and verified. Science consists of a set of beliefs about what can be observed, while religion consists of a set of beliefs about what cannot be observed. But, in the technical language of the philosophy of mind, all cognitive states -- percepts, memories, and thoughts -- are beliefs.

In each case, the idea is that these internal factors dispose a person to behave in particular ways:

  • Friendly people tend to smile,
  • Liberals tend to vote for Democrats,
  • Happy people laugh,
  • Hungry people eat,
  • People who value art over science give money to museums, and
  • People who believe in God go to church, temple, or mosque.

Most research on the personal causes of behavior has focused on personality traits, so they receive special attention here.

The Structure of Personality

Personality may be defined as the distinctive patterns of experience, thought, and action that characterize the individual's unique response to his or her life situation.

Traditionally, the scientific study of personality has begun with the ways in which people describe each other. In this culture, such descriptions tend to employ a mixture of terms:

  • types -- nouns labeling categories of people (e.g., extraverts, creeps, jocks, preppies);
  • traits -- adjectives labeling specific characteristics of people (e.g., sociable, stupid, sophisticated, conscientious).

The earliest "scientific" theories of personality were based on type categories. For example, the humour theory of temperament slotted people into one of four personality types:

  • choleric -- quickly aroused, hotheaded, active, egocentric, exhibitionistic, and histrionic;
  • sanguine -- sociable, hopeful, contented, easygoing, carefree;
  • phlegmatic -- reasonable, persistent, high-principled, calm, controlled, steadfast;
  • melancholic -- serious, thoughtful, anxious, worried, unhappy, suspicious.

The typological approach seems satisfying at first glance, because the categories seem to capture the gist or essence of many people we know. But from a scientific point of view it is ultimately unsatisfying, simply because most people don't fit easily into these or any other set of pigeonholes. A major problem with the classic type theories of personality is that they are based on the classic view of categories as proper sets defined by singly necessary and jointly sufficient features, with sharp boundaries between categories, homogeneity within categories, and no unclear cases. But such an approach fails for people for the same reason that if fails for other natural objects: some people can't be classified under the scheme at all, others seem to possess characteristics of more than one type, and some are more representative of a type than others.

In principle, these problems of partial and combined expression could be solved by shifting to a probabilistic, fuzzy-set view of personality types, and in fact just such a move has been made for the diagnostic categories in psychopathology. But in normal personality, the problem was solved by shifting emphasis from types to traits -- continuous dimensions (analogous to height and weight) representing psychological features that everyone possesses, but in varying degrees. Because everyone is considered to possess each trait, all cases can be classified; and a multidimensional system can represent all possible collections of traits, thus permitting partial and combined expression.

TraitNames.JPG (115513 bytes)Describing personality in terms of traits, however, raises the thorny question of how many dimensions are needed for the job. One count of dictionary terms, by Allport & Odbert (1936), determined that there were 17,943 trait terms in the English language (actually, sad to say, they missed one, and there were really 17,944). More have been added since then.

This is obviously a lot of traits, and it would be convenient to have a more concise summary. In fact, employing a statistical technique called factor analysis (which is based on the correlation coefficient), it has been determined that only five traits are basic to personality description. Put another way, there are five basic dimensions of personality:

  • extraversion -- e.g., talkative vs. silent, sociable vs. reclusive, adventurous vs. cautious, open vs. secretive;
  • agreeableness -- e.g., good natured vs. irritable, cooperative vs. negativistic, mild and gentle vs. headstrong, not jealous vs. jealous;
  • conscientiousness -- e.g., responsible vs. undependable, scrupulous vs. unscrupulous, not dependable vs. quitting, fussy and tidy vs. careless;
  • neuroticism (or its reverse,emotional stability) -- e.g., calm vs. anxious, composed vs. excitable, not hypochondriacal vs. hypochondriacal, and poised vs. nervous and tense;
  • openness to experience (also known as intellectance or culturedness) -- e.g., intellectual vs. unreflective and narrow, artistically sensitive vs. artistically insensitive, imaginative vs. simple and direct, polished and refined vs. crude and boorish.

BigFiveList.JPG (99881 bytes)These traits have been called "The Big Five", and are commonly considered to represent a universally applicable structure for personality description -- encoded in language, and permitting comprehensive descriptions and comparisons of personality, applicable to people of any age in any culture at any time.

I call the Big Five The Five Blind Date Questions: they represent the most important things you'd like to know in advance about someone you're going to spend some time with:

  • Is he or she crazy?
  • Is he or she outgoing?
  • Is he or she friendly?
  • Is he or she trustworthy?
  • Is he or she interesting?

Although most psychologists are fairly satisfied with the Big Five structure of personality, some theorists have suggested that we can reduce these five dimensions still further (remember that we started out with 17,954!), to just two:

  • Social Evaluation (extraversion, agreeableness, conscientious, and emotional stability -- the opposite of neuroticism, which are all positive interpersonal traits); and
  • Intellectual Evaluation (openness, intellectance, or culturedness, which are all positive intellectual traits).

The Structure of Attitudes

Similar considerations apply to the domain of attitudes. There is a potentially infinite number of attitudes -- as many attitudes as we have objects to have attitudes about, which is a lot -- war, abortion, nuclear power, global climate change, feminism, racial integration, my next-door neighbor, General Motors, the sweater that my mother gave me for my birthday. And methods similar to those used to reduce the corpus of personality traits to a manageable number have been used to uncover the basic structure of attitudes.

LiberalConserva.JPG (46824 bytes)In this case, it appears that the structure of attitudes can be boiled down to a single dimension -- "the Big One?":liberal versus conservative, pretty much as those terms are understood today. In addition, there is some evidence for a second dimension, reflecting individual differences in attitudes concerning traditional morality. Interestingly, the two dimensions appear to be largely independent of each other: one can be politically liberal but very traditional with respect to morality; and one can be a conservative libertine.

LibconBig5.JPG (59431 bytes)If that's too monolithic a structure for you, it can be argued that the structure of attitudes consists of yet another "Big Five": that is, people are either relatively liberal or relatively conservative in political,economic, religious,social, and esthetic domains. You can be a raving liberal Democrat and still not like modern art. You can be an Orthodox Jew but still favor liberal social causes. Maybe, like Spearman's g, a single dimension of liberalism-conservatism is just too abstract, and it's more useful to focus on what might be called "primary attitudinal domains".

"Liberal" and "Conservative" Politics

Studying the structure of political attitudes, not to mention their origins (discussed in the lectures on Psychological Development) is complicated by the fact that psychologists, not to mention ordinary people, use terms like "liberal" and "conservative" in a manner that departs from the technical meaning of these terms in political philosophy and political science.

For most people, "liberal" means something like "Democrat", and "conservative" means something like "Republican". But the Republican Party of the 19th century, the party of Abraham Lincoln and abolition, was decidedly more "liberal" than the Democratic Party of the day (which, for example, thought slavery was just fine). The "liberal" label began to be attached to Democrats only in the early 20th century, as a reaction to Republican policies that led up to the Great Depression -- think of Franklin Delano Roosevelt -- and the stirrings of the civil rights movement after World War II (think Hubert H. Humphrey, Adlai Stevenson, and Lyndon B. Johnson).

So what does it mean, in political science, to be a "liberal" or a "conservative"? Both terms emerged in the aftermath of the French Revolution of 1789, which also gave us the shorthand distinction between the "left" and "right" in politics: in the National Assembly, the revolutionaries sat to the left of the Assembly president, while the royalists sat to his right). Mark Lilla, an intellectual historian at Columbia University, offers a succinct overview in his review essay, "Republicans for Revolution" (New York Review of Books, 01/12/2012).

  • The term conservative was introduced by Edmund Burke, an Irish-English political philosopher and member of the English Parliament who supported the American Revolution, but opposed the French Revolution. Burke argued that, because individuals are born into society, society has, in some sense, a prior existence, and our obligations to society take precedence over individual rights. Thus, Burkean conservatives honor custom and tradition, and believe that change should occur slowly.
  • The term liberal was popularized a little later, in the 19th century, by John Stuart Mill, another British political philosopher, argued that, because societies were made up of individuals, individuals had priority. The features of society are created by human action, both individual and collective, and can be un-made and re-made in the same way. Far from bowing to custom and tradition, social change is constrained only by general principles that transcend the customs of particular societies.

Put bluntly: Liberals place the highest value on the individual and his liberty, while conservatism places the highest value on conserving societal traditions.

In modern American politics, there are three general kinds of conservatives, and they don't always agree with each other. But they are all conservatives in the Burkean sense.

  • Paleoconservatives" are essentially isolationist when it comes to foreign policy.
  • "Neoconservatives" tend to be liberal with respect to domestic policy, but embrace an expansionist and interventionist foreign policy (Irving Kristol famously defined a neoconservative as "a liberal who has been mugged by reality").
  • "Theoconservatives" are more concerned with traditional theological views.

To make things more complicated, Lilla argues that the liberal-conservative polarity doesn't exhaust the relevant political categories.

There are, for example, "revolutionaries" and "reactionaries", a distinction based not on one's view of human nature, but on one's view of history.

  • Revolutionaries see themselves as contributing to inexorable process of human development. In Lilla's words, "there is no going back".
  • Reactionaries want to restore some "golden" past era (whether real or imaginary).

And there are also two types of reactionaries:

  • "Restorative" reactionaries think that, despite revolution, it is possible to restore that "golden age".
  • "Redemptive" reactionaries accept the fact of the revolution, but promote a counterrevolution that would establish a new "golden age" -- as opposed to restoring an old one.

What's the difference between a liberal and a conservative?  Maybe their approach to moral reasoning.  Jonathan Haidt, in The Righteous Mind: Why Good People Are Divided by Politics and Religion (2012), identifies six principles that underly moral values and judgments:

  • Care/Harm, emphasizing kindness, gentleness, and nurturance.
  • Fairness/Cheating, emphasizing rights and justice.
  • Loyalty/Betrayal, emphasizing patriotism and self-sacrifice for the good of the group.
  • Authority/Subversion, emphasizing deference to tradition and authority.
  • Sanctity/Degradation, emphasizing religious ideals and "noble" living.
  • Liberty/Oppression, emphasizing autonomy and resistance to oppression.

Haidt argues that liberals' moral judgments are primarily based on considerations of care/harm, fairness/cheating, and liberty/oppression.  Conservatives, in his view, emphasize all six of these principles.  For this reason, liberals and conservatives have difficulty, not just in reaching agreement with each other, but even in understanding how the other side thinks.

Viewed from this perspective, psychological tests that purport to assess "liberal" and "conservative" social attitudes probably are off the mark. But for now, we're stuck with them (which is, of course, an essentially conservative position).

The Assessment of Personality

"Role Models" (1992), painting by Arthur Tress.The background is the score sheet for the California Psychological Inventory.  Reprinted from Natural History, 07-08/04.

It is clear that we commonly describe each other in terms of traits, but the scientific study of personality wants to go beyond mere description to prediction and explanation. In that effort, a particular conception of personality has evolved that asserts that traits have a physical existence in the nervous system and cause people to behave in characteristic ways in a wide variety of situations. Trait theorists disagree about whether personality traits are inherited or acquired through experience but they agree that traits exist and can be measured quantitatively. It is old doctrine in science that whatever exists, exists in measurable form.

The problem of measuring personality traits is one of psychometrics, or measuring psychological attributes of people. The most convenient instruments for this purpose are adjective ratings scales, where the items are trait adjectives such as those that comprise The Big Five; or personality questionnaires in which subjects describe their actual or likely behavior in a number of different situations that bear on the trait(s) in question. Most personality testing regards these self-reports as substitutes for actual behavioral data collected by trained observers -- that is, it assumes some degree of isomorphism between what people say about themselves on personality questionnaires, and what they actually experience, think, and do.

Personality tests, like all measuring instruments must have certain psychometric properties.

  • Standardization -- rules for administering and scoring the test that guarantee that each individual responds to the same situation;
  • Reliability -- some degree of precision in measurement, expressed either by
    • Inter rater Reliability (agreement between two observers rating the same person); or
    • Test-Retest Reliability (agreement between measurements taken on two different occasions) or
    • Internal Consistency (a function of the "item-to-total" correlation between performance on each individual item and performance on the test as a whole (with the total test score corrected to eliminate the item in question).
  • Validity -- some sense that the test accurately measures the trait it is supposed to measure, expressed in terms of
    • Face Validity, or the extent to which test items seem valid "on their face", or superficially, as measures of the trait (a test of extraversion shouldn't be composed of items relating to neuroticism);
    • Content Validity, or the extent to which the test surveys the universe of content represented by the trait (a test of extraversion should have items relating to assertiveness, activity levels, excitement-seeking, and positive affect as well as interpersonal warmth and gregariousness);
    • Empirical Validity: (the scale itself, as well as each of its constituent items, should correlate positively with some criterion index, such as the number of friends one has).
      • Items that have both empirical and content validity are sometimes called "obvious" items, because they are obviously related to the trait in question.
      • Items that have empirical validity, but not content validity, are sometimes called "subtle" items, because their relevance isn't obvious to a naive test-taker. Personality questionnaires that have been constructed solely on the basis of empirical validity, without regard to item content, don't do very well.
    • Construct Validity: test scores should fit into a nomological network of relations with other traits, as predicted by the theory of the trait in question (extraversion might be correlated with agreeableness, because both are socially positive traits, but neither extraversion nor agreeableness should be substantially correlated with IQ -- unless, of course, your theory of extraversion holds that extraverts are smarter than introverts).

In addition, one psychometric property is highly desirable, even if it is not strictly necessary:

  • Utility, or Efficiency-- or some sense that the test provides an economic advantage over alternative measures of the same trait, expressed as the cost/benefits ratio (where cost refers to the expense of constructing, administering, and scoring the test, and benefits refers to the validity of the test in question); the most efficient tests have high reliability and validity but low cost of administration.

The Questionnaire Method

Usually, psychologists assess individual differences in personality by means of a self-report questionnaire, in which subjects answer questions about their own behaviors, thoughts, feelings, and emotions.  Questionnaires are a very economical means of personality assessment, which is why they're so commonly used.  But they depend on the assumption that people are reliable and valid reporters of their own behavior.

Observational Methods

An alternative mode of personality assessment is for psychologists, or others, to observe subjects' behavior directly.  Behavioral observation gets around the principal problem of the questionnaire method, which is that self-reports are of uncertain reliability and validity.  On the other hand, observational methods are very expensive, simply because observers have to observe large numbers of subjects behaving in a large number of different situations.   

Projective Methods

While questionnaires and behavioral observations are considered to be objective techniques for assessing personality, there are also projective methods.  The projective methods arose out of Freud's psychoanalytic theory of personality.  Freud assumed that the important features of personality were unconscious, in that people had no awareness of them.  This makes self-report questionnaires pretty useless.  The projective methods, on the other hand, assume that subjects unconsciously project their personalities onto various aspects of test performance.  Even an IQ test, in this view, can reveal important aspects of personality.

Although Freud himself didn't devise any projective measures of personality, other researchers, inspired by psychoanalysis, did. 

  • One of these, the Thematic Apperception Test, was briefly described in the lectures on "Motivation".  Recall that Murray and his associates assumed that subjects would project their own motives, goals, aspirations, and desires, on the characters about whom they made up their stories.

  • Another of these, perhaps the most famous and controversial, is the Rorschach Inkblot Test.  As described in the lectures on "Perception", the Rorschach was initially devised for studies of the Gestalt principles of perceptual organization; but it was soon co-opted into a projective test for personality assessment.  It's not a particularly good instrument for this purpose, as discussed briefly in the lectures on Perception.  For a nice history of the "Rorschach test", see The Inkblots: Hermann Rorschach, His Iconic Test, and the Power of Seeing (2017) by Damion Searls.

Questionnaires are easy to construct but boring to score; behavioral observation is hard to do, but possibly more valid.  Projectives are exciting for everyone (I loved administering, scoring, and interpreting projective tests when I was a clinical intern in graduate school), but of dubious reliability, validity, and utility.  This is especially so for the Rorschach (and less so for the TAT, in its contemporary incarnation as the Picture-Story Exercise).  As a technique for personality assessment, the Rorschach is probably not worth the paper its printed on.  although it can be scored reliably, the scores have low external validity, and even when valid scores can be derived from the Rorschach, there are easier and cheaper ways to obtain the same information.  For a comprehensive assessment of the Rorschach as a psychometric instrument for personality assessment, see "Roots of the Rorschach controversy" by Garb, Wood, Lilienfeld, & Nezworski, Clinical Psychology Review, 2005.

Assessing The Big Five

Over the years, a number of personality inventories have been published, providing assessments of a number of different personality traits. Among these, perhaps the most famous are:

  • Minnesota Multiphasic Personality Inventory (MMPI),
  • California Psychological Inventory (CPI),developed here at UC Berkeley by Harrison Gough;
  • Eysenck Personality Inventory (CPI),
  • Personality Research Form (PRF), and
  • Jackson Personality Inventory (JPI).

Lately, a questionnaire designed to assess subjects' standing on the "Big Five" personality traits has been developed, called the NEO Personality Inventory (NEO-PI; the "NEO" stands for Neuroticism, Extraversion, and Openness, the three major traits assessed in early versions of the scale; Agreeableness and Conscientiousness were added later).

The NEO-PI assesses each of the Big Five personality dimensions in terms of six "facets". After analyzing each Big Five dimension into these facets, the authors (Robert McCrae and Paul Costa) then developed eight self-report items bearing on each of these facets, for a total of 240 items. The Big Five dimensions, subordinate facets, and sample inventory items are presented below, to give you a flavor for the scale. In addition, I show the distribution of Big Five scores in a large sample of college undergraduates, as assessed by an abbreviated version of the NEO-PI called the NEO Five-Factor Inventory.

neuroticism.gif (6066 bytes) extraversion.gif (5574 bytes) agreeableness.gif (5147 bytes) conscientiousness.gif (5512 bytes) openness.gif (4895 bytes)
Neuroticism Extraversion Agreeableness Conscientiousness Openness to Experience
  • Neuroticism
    • Anxiety ("I often feel tense and jittery")
    • Angry Hostility ("I often get angry at the way people treat me")
    • Depression ("Sometimes I feel completely worthless")
    • Self-Consciousness ("In dealing with other people, I always dread making a social blunder")
    • Impulsiveness ("I have trouble resisting my cravings")
    • Vulnerability ("When I'm under a great deal of stress, sometimes I feel like I'm going to pieces")
  • Extraversion
    • Interpersonal Warmth ("I really like most people I meet")
    • Gregariousness ("I like to have a lot of people around me")
    • Assertiveness ("I am dominant, forceful, and assertive")
    • Activity ("When I do things, I do them vigorously")
    • Excitement-Seeking ("I often crave excitement")
    • Positive Emotions ("I have sometimes experienced intense joy or ecstasy")
  • Agreeableness
    • Trust("I think most people I deal with are honest and trustworthy")
    • Straightforwardness ("I would hate to be thought of as a hypocrite")
    • Altruism ("I go out of my way to help others if I can")
    • Compliance ("I would rather cooperate with others than compete with them")
    • Modesty ("I try to be courteous to everyone I meet")
    • Tender-Mindedness ("I believe that most people are basically well-intentioned")
  • Conscientiousness
    • Competence ("I am a productive person who always gets the job done")
    • Order ("I keep my belongings neat and clean")
    • Dutifulness ("I try to perform all the tasks assigned to me conscientiously")
    • Achievement Striving ("I work hard to accomplish my goals")
    • Self-Discipline ("I'm pretty good about pacing myself so as to get things done on time")
    • Deliberation ("I try to do jobs carefully, so they won't have to be done again")
  • Openness to Experience
    • Fantasy ("I have a very active imagination")
    • Esthetics ("I am sometimes completely absorbed in music I am listening to")
    • Feelings ("Without strong emotions, life would be uninteresting to me")
    • Ideas ("I often enjoy playing with theories or abstract ideas")
    • Actions ("I think it's interesting to learn and develop new hobbies")
    • Values ("I believe that laws and social policies should change to reflect the needs of a changing world")

Alternatively, the Big Five can be measured simply by asking subjects to rate themselves on a series of adjectives representing the various facets of the five dimensions. The following list shows selected correlates of each of the Big Five facets, as measured by the NEO-PI, with items of the Adjective Check List (ACL):

  • Neuroticism
    • Anxiety (anxious, fearful, worrying, tense, nervous)
    • Angry Hostility (irritable, impatient, moody, excitable)
    • Depression (worrying, pessimistic, not self-confident)
    • Self-Consciousness (shy, timid, defensive)
    • Impulsiveness (sarcastic, loud, hasty)
    • Vulnerability (not self-confident, careless)
  • Extraversion
    • Interpersonal Warmth (friendly, warm, sociable, cheerful)
    • Gregariousness (outgoing, social, spontaneous)
    • Assertiveness (aggressive, assertive, self-confident, forceful)
    • Activity (energetic, hurried, determined)
    • Excitement-Seeking (pleasure-seeking, daring, adventurous, spunky)
    • Positive Emotions (humorous, optimistic, jolly)
  • Agreeableness
    • Trust(forgiving, trusting, peaceable)
    • Straightforwardness (uncomplicated, not shrewd, not autocratic)
    • Altruism (warm, soft-hearted, generous, kind, tolerant)
    • Compliance (not stubborn, not headstrong, not hard-hearted)
    • Modesty (not a show-off, not assertive, not clever)
    • Tender-Mindedness (friendly, soft-hearted, gentle, kind)
  • Conscientiousness
    • Competence (efficient, thorough, resourceful)
    • Order (organized, precise, methodical, not absent-minded)
    • Dutifulness (not distractible, not careless, not fault-finding)
    • Achievement Striving (ambitious, industrious, enterprising, confident)
    • Self-Discipline (organized, not absent-minded, industrious, energetic)
    • Deliberation (not hasty, not careless, not immature, not impatient)
  • Openness to Experience
    • Fantasy (dreamy, imaginative, artistic, idealistic)
    • Esthetics (artistic, original, inventive, versatile)
    • Feelings (excitable, spontaneous, insightful, imaginative)
    • Actions (idealistic, wide interests, curious, insightful)
    • Ideas (inventive, curious, original, imaginative, insightful)
    • Values (not conservative, unconventional, not cautious)

              (6011 bytes)Returning to the NEO-PI data, note that the distribution of each Big Five trait approximates a "normal" or "bell-shaped" curve -- and that, happily, the average score on Neuroticism is lower than the average scores on Extraversion, Agreeableness, Conscientiousness, and Openness.

              (5229 bytes)Interestingly, the Big Five traits are, to all practical purposes, uncorrelated with each other. The correlations among the traits are much closer to 0.0 than they are to 1.0. This means that, for all practical purposes, we cannot predict a person's standing on one trait from knowledge of his or her standing on another.

However, the correlations are not precisely zero, and so a higher-order resolution of the Big Five structure is possible. An analysis of personality trait adjective ratings by Seymour Rosenberg and Andrea Sedlak has showed that they can all be arranged on two dimensions of evaluation:

  • Intellectual good-bad
  • Social good-bad

rosensedlak.gif (4792 bytes)Moreover, these dimensions are correlated with each other, such that, on average, people who are perceived to have socially desirable traits are also perceived to have intellectually desirable traits. Even though the Big Five traits are essentially uncorrelated with each other, you should note that, with the exception of Neuroticism, each of the Big Five traits represents a socially desirable quality. Neuroticism is something that can be charming in small amounts, but extremely annoying and maladaptive in large amounts. Accordingly, we can propose a higher-order resolution of the Big Five along the following lines:

Trait Undesirable Desirable
Neuroticism High Low
Extraversion Low High
Conscientiousness Low High
Agreeableness Low High
Openness Low High

The Function of Traits

Once the basic traits have been identified and a technology has been developed to measure them, research can proceed on the nature of traits, their origins, and their influence on experience, thought, and action. With respect to the functions of traits, there are two views:

  • The weak view, closely related to what Allport called the biosocial view of traits, holds that traits merely summarize the individual's behavioral tendencies -- for example, that extraverted people frequently engage in behaviors we call extraverted. In this view, traits are merely labels, or perhaps cognitive categories, to which we assign various behaviors.
  • The strong view holds that traits cause behavior to occur --for example, that people frequently engage in behaviors we call extraverted because they are, in fact, extraverted people. It follows from this view that traits are not merely "biosocial" categories for labeling certain behaviors, but instead are "biophysical" entities that exist physically in the nervous system, much like reflexes or memories do, presumably forming part of the individual's brain structure. By virtue of this physical existence, they cause or dispose people to behave in certain ways across a wide variety of situations.

In the weak view, traits have only nominal existence; in the strong view, traits have actual existence. The strong view is exemplified by the doctrine of traits: traits cause the coherence, stability, consistency, and predictability of the individual's behavior.

Obviously, no one expects to observe perfect levels of coherence, stability, consistency, and predictability. But the stronger these features appear in human experience, thought, and action, the more the doctrine of traits seems valid -- the idea that personality traits exist in people and dispose them to particular modes of experience, thought, and action. For the past 50 years, considerable research has been devoted to these issues.


hierarchy.gif (7765 bytes)The coherence of personality is expressed in two ways:

  • coherence among topographically different behaviors (such as smiling and initiating conversation) -- which, in theory, tend to co-occur reliably because they are different manifestations of a particular trait (such as friendliness); and
  • coherence among semantically different traits (such as friendliness and adventuresomeness) -- which, in theory, tend to covary because they are different facets of some higher-order trait (such as extraversion).

Hierarchyc.JPG (58700 bytes) The coherence assumption results in a hierarchical structure of individual differences:

  • At the lowest level are specific actions engaged in some unique episodic context:
    • For example, a person might have taken over the leadership of a particular committee.
  • At the next level are habitual actions, repetitions of the same behavior in different contexts:
    • For example, a person might generally try to assume a leadership role in groups of which he is a member.
  • At the next level are primary traits, summarizing clusters of co-occurring behaviors:
    • For example, a person might be assertive in many different situations, with many different types of people, whether it is appropriate or not.
  • At the next level are secondary traits, summarizing clusters of covarying primary traits:
    • For example, people who are assertive also tend to be interpersonally warm, so that both traits are expressions of extraversion.
  • And at the next level are tertiary traits, summarizing clusters of covarying secondary traits:
    • For example, people who are extraverted also tend to be agreeableness (or at least we tend to think this is so; see below), and both traits are socially desirable.
  • And so on up the hierarchy.

              (77247 bytes)For example, we also tend to think that people who have socially desirable traits also have intellectually desirable traits, forming a "super-superordinate" trait of good-bad. In what is known as the halo effect, we tend to perceive socially desirable traits as correlated with each other. Thus, people who are perceived as extraverted will also be perceived as agreeable, and people who are extraverted and agreeable will also be perceived as intelligent and creative.

Secondary.JPG (50095 bytes) The assumption of coherence lies behind the practice of measuring a trait in terms of many different items, bearing on several different facets, instead of only a single target behavior or a single facet. The NEO-PI represents the Big Five hierarchically in the following manner:

  • as secondary traits (e.g., extraversion and agreeableness),
  • subsuming primary traits (e.g., warmth and assertiveness, trust and altruism,
  • as indicated by habitual behaviors (e.g, liking most people you meet, being dominant, forceful, and assertive; believing most people are honest, and being courteous to everyone).

What the NEO-PI leaves out are:

  • assessments of tertiary traits such as social intellectual desirability; and
  • assessments of specific behaviors in specific situations, such as
    • Liking your roommate when you met him or her for the first time;
    • Being dominant, forceful, and assertive during a particular meeting of your Psych 1 discussion section;
    • Believing that your current girlfriend or boyfriend is honest; and
    • Being courteous to an elderly person who wanted to get ahead of you in the supermarket checkout line last Saturday.

With respect to the Doctrine of Traits, the strongest evidence is for the coherence of personality -- studies in a wide variety of populations seem to converge on The Big Five as a coherent hierarchical structure of personality, uniting habitual behaviors, primary traits, and higher-order traits.

However, it should be noted that most of this literature has concerned the coherence of habitual behaviors, subsumed by primary traits, and the coherence of primary traits, subsumed by secondary traits such as The Big Five. There has been little research on the coherence of actual behavior.

At the same time, it should be noted that the extent of coherence of habitual behaviors and primary traits may be magnified by people's expectations and beliefs about personality structure. For example, The Big Five traits fall out when people make abstract judgments of the semantic similarity or co-occurrence of traits and actions, as well as when they rate specific other people. More critically, The Big Five falls out when people rate the personalities of total strangers -- people for whom they could not possibly have knowledge of how they actually behave. Apparently, people carry around in their heads an implicit theory of personality -- a set of beliefs, not always well articulated, about how various aspects of personality hang together (like "The Big Five Blind Date Questions") that influences their judgments -- in much the same way that perception of the nonsocial world is influenced by the perceiver's beliefs and expectations.


              (53608 bytes)Within each level of the hierarchy, the Doctrine of Traits assumes that behaviors and traits are relatively stable across both long and short intervals of time. People who were extraverts at age 10 should still be relatively extraverted at age 50.

The assumption of stability lies behind the psychometric property of test-retest reliability.

There is less evidence in favor of stability: although personality questionnaire scores and adjective trait ratings show significant degrees of test-retest reliability over even very long periods of time, there is at least as much variability as there is stability. Stability is greatest over relatively short intervals of time.


Consistencyc.JPG (57872 bytes)Again within each level of the hierarchy, the Doctrine of Traits assumes that behaviors and traits are relatively consistently displayed across a wide variety of different situations. The whole point of a trait, given the strong view, is to generate consistent responses in a variety of situations, and to render various situations "functionally equivalent" in terms of the behavior that occurs within them.

There is even less evidence for consistency: human experience, thought, and action is remarkably flexible and finely tuned to the specific situation in which it occurs. Consistency is greatest across situations that are highly similar.


Predictabilityc.JPG (62376 bytes)Finally, it should be possible to estimate a person's standing across different levels of the hierarchy:

  • given knowledge of one aspect of a person's habitual behavior, we should be able to predict one of his or her traits; and
  • given knowledge of a person's traits, we should be able to predict his or her behavior in specific situations.

The assumption of predictability underlies the psychometric property of external validity. Trait measures are considered valid to the extent that they predict what people will actually do. And this returns us to Dialectic Between the Person and Behavior.

Predicting Behavior from Traits and Attitudes

              (44100 bytes) In the context of Lewin's formula,B =f(P,E),the most important implication of the Doctrine of Traits is prediction, because, according to the doctrine, a person's behavior (B) is a function of his or her traits and other internal, personal characteristics (P). So, to what extent can we predict a person's actual behavior in some specific situation from knowledge of his or her generalized personality traits and attitudes?

              (69591 bytes)For example, given that we know that a person is extraverted, and thus tends to be warm, and also to like most people, will he also like Judy when he meets her?

              (74619 bytes)And given that we know that a person is agreeable, and thus tends to be trusting, and also to believe most people are honest, will he also let his co-worker borrow some money?

To be honest, this question has not been studied as much as we would like. A great deal of personality research has focused on identifying the basic structure of personality (the question of coherence), and in addressing questions such as the stability of personality and the consistency of behavior across situations. Relatively few studies have addressed the question of predictability.

Ideology and Voting Behavior

JostSummary.JPG (58954 bytes) Perhaps the strongest evidence for predictability concerns the relation between political attitudes and voting behavior. Jost (2006) analyzed survey data from the American National Election Studies (ANES)covering all US presidential elections from 1972 (McGovern vs. Nixon) to 2004 (Bush vs. Kerry). In the survey, more than 7,500 respondents rated themselves on a 7-point scale of political liberalism-conservatism (where 1 = extremely liberal and 7 = extremely conservative), and then reported how they had voted in the most recent election. In each election, Jost found that voting behavior was almost perfectly correlated with political ideology, with liberals voting much more often for Democrats, and conservatives voting much more often for Republicans. In the 2004 election, for example, the correlation between political liberalism-conservatism and voting for John Kerry vs. George W. Bush was a whopping r = .97!

The following graphs depict individual results for each presidential election, from 1972 to 2004:

Jost2004.JPG (58651 bytes)


Jost2000.JPG (58198 bytes)


Jost1996.JPG (58361 bytes)


Jost1992.JPG (57684 bytes)


Jost1988.JPG (59087 bytes)


Jost1984.JPG (59500 bytes)


Jost1980.JPG (58346 bytes)


Jost1976.JPG (58024 bytes)


Jost1972.JPG (59773 bytes)


Although this is certainly an impressive level of predictability, in most cases our ability to predict specific behavior from measurements of personality traits is extremely limited.

Racial Prejudice

But attitudes are not nearly so strongly predictive of actual behavior in other instances. In a famous study, LaPiere (1934), traveled around the United States with a Chinese couple. At a time when American prejudice against Chinese was particularly strong, LaPiere was surprised to discover that the couple was rarely refused accommodation at hotels and restaurants. After completing the tour, he wrote to the hotels and restaurants that he had visited, and asked whether they accepted "members of the Chinese race as guests".

  • The vast majority of hotels, 43 out of 47, replied in the negative.
    • But in fact, all but 1 of these 43 hotels had actually accommodated the couple!
  • And so did the vast majority of restaurants, 75 out of 81.
    • And every one of those 75 restaurants actually served the couple!

LaPiere concluded:

"In the end I was forced to conclude that those factors which most influenced the behavior of others towards the Chinese had nothing at all to do with race."

In other words, racial attitudes were not predictive of actual behavior.

Delay of Gratification

Consider, for example, a study of delay of gratification in young children (Funder, Block, & Block, 1983), conducted at UCB. Delay of gratification has to do with people's ability to tolerate frustration and control their impulses. It's a pretty important aspect of socialization. Max Weber, a pioneering sociologist, thought that delay of gratification was the basis for the "Protestant ethic" of self-restraint and the negation of pleasure (obviously, he didn't know many Protestants!) which he thought lay at the basis of capitalism. But every culture requires some ability for ego-control: to plan ahead and tolerate delays. This ability is generally acquired early in life, as a result of socialization. Of course, even within a culture individuals will differ in their ability to delay gratification: as traditional personality psychologists might put it, some people have it, some people don't.

In the Funder study, nursery-school teachers administered a test of intelligence, and also rated their pupils' personalities on an instrument (known as the California Q Set) that provided measures of two higher-order dimensions of personality:

  • Ego Control -- the ability of the child to control his or her impulses, including the ability to delay gratification, inhibit aggression, and engage in planful behavior. The scale for this trait includes the item, "Is unable to delay gratification". Ego control is roughly equivalent to the conscientiousness dimension of the Big Five.
  • Ego Resiliency -- the ability of the child to adapt to environmental demands, including feelings of security and of competence. Ego resiliency is roughly equivalent to the neuroticism dimension of the Big Five.

Why didn't Funder et al. assess the children's personalities in terms of The Big Five? At the time the study was done, consensus had not yet developed around The Big Five as the structure of personality. There were other competing systems, including Hans Eysenck's four-factor proposal (neuroticism, extraversion, psychoticism, and intelligence) and the two-factor system (ego-control and ego-resiliency) proposed by Jack and Jean Block, who after all Funder's co-investigators. These two competing systems still undergird much research, and in fact Jack Block has on numerous occasions (e.g., 1995, 2001) pronounced himself a "Big Five Contrarian", and offered trenchant critiques of the research supporting The Big Five as a universally applicable structure for personality description.

Around the same time as the assessments were made, Funder et al. engaged the children in an experimental assessment of their ability to delay gratification employing two different situations

  • In the gift-delay situation, the child was offered a gift-wrapped package, and then told that he must wait a while to open the gift. The measure of delay is how long the child is able to wait, after the experimenter leaves the room, without asking to open the gift.
  • In the resistance to temptation situation, the child is ushered into a room furnished with attractive and unattractive toys, and is forbidden to play with the attractive set because they belong to someone else. The measure of delay is whether the child plays with the forbidden toys after the experimenter leaves the room.
Predictor Variable r
IQ .21
Ego Control .25
Ego Resiliency .23
"Is unable to delay gratification" -.27

All four predictors -- IQ, ego control and ego resiliency, and the specific item "Is unable to delay gratification" correlated in the range of .20 <r < .30 with actual delay behavior (the correlation between delay and the specific item was, of course, negative).

These values are fully representative of the kinds of validity coefficients routinely obtained in these kinds of studies. Personality in general predicts behavior in particular, but the prediction is relatively weak. There's more to predicting behavior than knowing the person's personality traits.

There is a subtle problem in all of these studies, which is that behavior is measured by only a single observation (an episode of the person's behavior, limited to a particular time and place), and single observations are inherently unreliable. Of course, one person's "unreliability" is another person's behavioral flexibility, but even so it may be unfair to ask personality traits to predict specific episodes of behavior with any great power.

Accordingly, some investigators have focused on the prediction of aggregate measures of behavior: instead of predicting a single instance of delay of gratification, for example, they might seek to predict the overall trend in delay behavior, measured across a number of different episodes. Instead of predicting specific instances of behavior, then, these studies seek to predict something closer to habitual behavior.

Predicting Habitual Behavior

For example, in one study, Sampo Paunonen assessed personality with the NEO-Five Factor Inventory (NEO-FFI), which yields scores for each of the Big Five personality traits. Scores from the NEO-FFI were then used to predict subjects' behavior on 13 different variables, such as the number of dates per month (not whether the subjects had a date on a particular Saturday night), and mean number of traffic violations in the past year (not whether they ran a particular stop sign on a particular day). The table below shows the correlations between each of the Big Five traits and selected criterion variables:

Criterion Variable

Big Five Trait

Number of Dates / Month .01 -.03 -.23 .03 .04
Number of Different People Dated / Month .01 .09 -.21 .04 -.06
Smoking Behavior .25 .03 -.16 -.29 .21
Number of Cigarettes Smoked / Month .10 -.01 -.13 -.17 .15
Liberal Arts or Pre-Professional Program .13 -.03 -.20 -.15 .18
Interest in Joining Fraternity/Sorority .26 .06 .05 .16 .19
Traffic Violations / Year .05 .14 -.03 .05 .25

Note that the correlations in question never exceeded .30.

Note, too, that it's generally hard to explain the correlations that Paunonen actually obtained:

  • It makes sense that agreeableness should predict the number of dates that a student has, but why should the correlation be negative? Is it really true that "nice guys finish last"?
  • Why should neuroticism predict interest in joining a fraternity or sorority?

Paunonen12.JPG (87976 bytes)And finally, Paunonen generally failed to confirm these findings in a second study conducted in almost exactly the same manner as the first.

  • In Study 1, dating frequency was negatively correlated with agreeableness; but in Study 2, it was negatively correlated with conscientiousness.
  • In Study 1, traffic violations were positively correlated with openness; but in Study 2, they were negatively correlated with extraversion.

A later study by Paunonen and Ashton (2001) used another version of the NEO Personality Inventory, which yields scores for six "facets" of each of the Big Five personality traits, obtained similar results.The correlations in question rarely if ever exceeded .30.

Note, however, just to reiterate, that the criterion variables used in these studies were actually self-reports of habitual behaviors. So, for example, although scores on Neuroticism, Conscientiousness, and Openness were significantly correlated (some positively, some negatively) with the number of cigarettes smoked per month as reported by the subjects themselves, we do not know how well these scales would predict the number of cigarettes smoked per month as recorded objectively by an independent observer, or whether a person would smoke on any particular day or in any particular situation. It stands to reason that, given the modesty of the predictions obtained in this study, from the level of secondary traits to the level of habitual behavior, the ability of secondary traits to predict behavior in specific situations will be even more limited.

Predicting Rooms

Another perspective on predictability is offered by a study by Samuel Gosling and his associates (2001), which asked whether people's personalities could be inferred from the appearance of their offices and bedrooms. For example, agreeable people might have many pictures of friends, or conscientious people might make their beds. Gosling and his associates obtained both self-ratings and peer-ratings of the Big Five personality traits, and then (with permission, of course) conducted surprise visits to their subjects' offices or bedrooms. For our purposes, we can think of the appearance of these rooms as samples of the subjects' behaviors on the particular day of the visit (my office is generally neat, moderately high level of conscientiousness, but sometimes the desk is quite cluttered; I always make my bed in the morning, too, perhaps for the same reason, but sometimes my nightstand is piled high with books and magazines). Could you use knowledge of my level of conscientiousness to accurately predict what my office or bedroom would look like on any randomly selected day? In my office, I'm often reading a book on some esoteric subject, such as hypnosis or consciousness, while my nightstand usually has artsy or "liberal" magazines on it, such as The New Yorker and The Nation; but sometimes I've got something utterly conventional on my desk, like an introductory psychology textbook, or my nightstand holds a Tom Clancy thriller. Could you use knowledge of my level of openness to experience to predict what would be readily at hand on any randomly selected day?

Big Five Predictor Office Bedroom
Emotional Stability (the reverse of Neuroticism) .24 .22
Extraversion -.04 .20
Openness to Experience .24 .33
Agreeableness .19 .36
Conscientiousness .46 .65

The answer is yes, to some degree. Still, most of the correlations don't exceed Mischel's personality coefficient of .30. The relationship between personality and behavior is stronger for bedroom appearance than office appearance -- probably because there are more institutional and societal constraints on what you can keep in your office as opposed to what you can put in your bedroom. The relationship is especially strong between the appearance of people's bedrooms and their level of openness to experience.

Predicting Smiling

Another recent study also illustrates the use of personality traits to predict behavior. LeeAnne Harker and Dacher Keltner (2001) analyzed data from a longitudinal study of women who graduated from Mills College, a private women's college in Oakland, California, in 1958 and 1960. As seniors, these women had completed a battery of personality assessments; they also had their photographs taken for their yearbook. The question is: Can personality predict how the women posed before the camera? The question is not a trivial one. We might expect happy people to smile and sad people to frown, but almost everyone smiles in their yearbook photo -- because that is what you're supposed to do, and the photographer won't snap the picture until you do so. But when a photographer asks you to "smile for the camera", the resulting posed facial expression (sometimes known as a "Duchenne smile"), often differs significantly from the smile spontaneously displayed by genuinely happy people. As it turns out, women who scored high on negative emotionality, a dimension related to neuroticism, were much less likely (r = -.37) to display an intense, genuinely positive facial expression in their yearbook picture. Somewhat puzzlingly, the correlation between smiling and positive emotionality, a trait related to extraversion, was also negative, but it was not statistically significant. But women who were high on affiliation, a trait related to agreeableness, were more likely to show a genuine smile (r = .33).

The Personality Coefficient

There's a pattern emerging here. For the most part, the correlations between general traits and specific behaviors seem to reach a plateau in the .20s and .30. Walter Mischel (1968), reviewing the literature up to that time, suggested that there was a ceiling, or upper limit, on the extent to which an individual's behavior in some specific situation could be predicted from knowledge of his or her generalized personality traits. Mischel called this ceiling, corresponding to a modest correlation of r = .30, the personality coefficient.

A correlation of .30 means that the two variables have about 10% of their variance in common -- actually, .302 = .09 or 9%.That means that only about 10% of the variance in actual behavior can be accounted for by the personality trait in question. The vast bulk of behavioral variation is left unaccounted for by personality traits.

Mischel's claims were controversial, to say the least -- and they were especially challenged by proponents of traditional personality research, based as it was on the Doctrine of Traits. His review was written relatively early in this debate, and drew on a relatively small body of empirical evidence. Still, he seems to have had it right: there is a ceiling, roughly corresponding to a correlation coefficient of r = .30, on our ability to predict specific behaviors on the basis of our knowledge of generalized personality traits.

Interestingly, at about the same time Wicker (1969) came to a similar conclusion about the predictability of behavior from attitudes.

Because the attitude construct was considered to be as central to social psychology as the trait construct was to the psychology of personality, Wicker's review precipitated a major crisis in the field, with large number of theorists wondering if it was a waste of time to study attitudes -- or, to put it another way, almost the entire history of experimental social psychology, from the time of Thurstone and Cave, had been a waste of time. And, interestingly, the solutions to the problem of predictability offered by social psychologists largely paralleled what was going on in personality psychology (e.g., Ajzen & Fishbein, 1977; Fazio & Zanna, 1981). Reflecting these trends, a later "meta-analysis" by Kraus (1995) yielded a an average correlation coefficient of r = .38. Call it, pace Mischel, "the attitude coefficient". But, as this later review indicates, these technical fixes didn't alter the situation very much.

General traits don't predict behavior very well, and neither do general attitudes.

So it's not the case that there is no predictability, just as it's not the case that there is no coherence, stability, or consistency. But levels of coherence, stability, consistency, and predictability of traits are rather modest -- perhaps much more modest than the Doctrine of Traits would have us believe -- because, as Lewin's formula indicates, other factors enter into the determination of behavior besides personal qualities such as traits and attitudes.We will discuss these factors when we discuss the other dialectics that comprise the Doctrine of Reciprocal Determinism.

But What About Freud?

Some students will notice a conspicuous omission from this treatment of the psychology of personality: the work of Sigmund Freud and his followers.

There's no question that Freud was a big influence on psychology, psychiatry, psychotherapy, and culture in general in the 20th century. But we now understand, in a way that we did not understand earlier, just how inadequate Freud's theories of personality and psychopathology are. Put briefly, they are very poorly grounded in empirical observation, being based mostly on Freud's self-analysis and his encounters with his patients, and his readings of ancient mythology (such as Sophocles's Oedipus Rex) -- all of which was filtered through the lens of his theory to begin with. And the theory just doesn't come up to the standards of a modern scientific theory of mind and behavior.

The 20th century was the century of Sigmund Freud, because Freud changed our image of ourselves (Roth, 1998). Copernicus showed that the Earth did not lie at the center of the universe, and Darwin showed that humans were descended from "lower" animals, but Freud claimed to show that human experience, thought, and action was determined not by our conscious rationality, but by irrational forces outside our awareness and control -- forces which could only be understood and controlled by an extensive therapeutic process called psychoanalysis.

Freud also changed the vocabulary with which we understand ourselves and others. Before you ever enrolled in this course you already knew something about the id and the superego, penis envy and phallic symbols, castration anxiety and the Oedipus complex. In popular culture, psychotherapy is virtually identified with psychoanalysis. Freudian theory, with its focus on the interpretation of ambiguous events, lies at the foundation of "postmodern" approaches to literary criticism such as deconstruction. More than anyone else, Freud's influence on modern culture has been profound and long-lasting.

Freud's cultural influence is based, at least implicitly, on the premise that his theory is scientifically valid. But from a scientific point of view, classical Freudian psychoanalysis is dead as both a theory of the mind and a mode of therapy (Crews, 1998; Macmillan, 1996). No empirical evidence supports any specific proposition of psychoanalytic theory, such as the idea that development proceeds through oral, anal, phallic, and genital stages, or that little boys lust after their mothers and hate and fear their fathers. No empirical evidence indicates that psychoanalysis is more effective, or more efficient, than other forms of psychotherapy, such as systematic desensitization or assertiveness training. No empirical evidence indicates that the mechanisms by which psychoanalysis achieves its effects, such as they are, are those specifically predicated on the theory, such as transference and catharsis.

Of course, Freud lived at a particular period of time, and it might be argued that his theories were valid when applied to European culture at that time, even if they are no longer apropos today. However, recent historical analyses show that Freud's construal of his case material was systematically distorted by his theories of unconscious conflict and infantile sexuality, and that he misinterpreted and misrepresented the scientific evidence available to him. Freud's theories were not just a product of his time: They were misleading and incorrect even as he published them.

Drew Westen (1988), a vigorous proponent of psychoanalytic theory and therapy, agreed that Freud's theories are archaic and obsolete, but argued that Freud's legacy lives on in a number of theoretical propositions that are widely accepted by scientists: the existence of unconscious mental processes; the importance of conflict and ambivalence in behavior; the childhood origins of adult personality; mental representations as a mediator of social behavior; and stages of psychological development. However, some of these propositions are debatable. For example, there is no evidence that childrearing practices have any lasting impact on personality. More important, Westen's argument skirts the question of whether Freud's view of these matters was correct. It is one thing to say that unconscious motives play a role in behavior. It is something quite different to say that our every thought and deed is driven by repressed sexual and aggressive urges; that children harbor erotic feelings toward the parent of the opposite sex; and that young boys are hostile toward their fathers, who they regard as rivals for their mothers' affections. This is what Freud believed, and so far as we can tell Freud was wrong in every respect. For example, the unconscious mind revealed in laboratory studies of automaticity and implicit memory bears no resemblance to the unconscious mind of psychoanalytic theory (Kihlstrom, 1998).

Westen also argues that psychoanalytic theory itself has evolved since Freud's time, and that it is therefore unfair to bind psychoanalysis so tightly to the Freudian vision of repressed, infantile, sexual and aggressive urges. This is true, and it is a historical fact that so-called "ego psychology" helped preserve much of what was interesting in psychology during its "Dark Ages" of radical behaviorism (Kihlstrom, 1994). But again, this avoids the issue of whether Freud's theories are correct. Furthermore, it remains an open question whether these "neo-Freudian" theories are any more valid than are the classically Freudian views that preceded them. For example, it is not at all clear that Erik Erikson's stage theory of psychological development is any more valid than Freud's is.

While Freud had an enormous impact on 20th century culture, he has been a dead weight on 20th century psychology. The broad themes that Westen writes about were present in psychology before Freud, or arose more recently independent of his influence. At best, Freud is a figure of only historical interest for psychologists. He is better studied as a writer than as a scientist. Psychologists can get along without him.

The low scientific status of Freudian psychoanalytic theory hasn't prevented it from serving as a basis for a form of psychotherapy that continues to be practiced by some therapists in America and around the world -- a point that will be discussed further in the lectures on Psychopathology and Psychotherapy.

Nor did it prevent Freudian psychoanalytic theory from influencing Western culture. Freud came on the cultural scene at the turn of the century, with his books on , The Interpretation of Dreams (1900), The Psychopathology of Everyday Life (1901), Fragment of an analysis of a Case of Hysteria (1905),and the General Introduction to Psycho-Analysis (1915) -- all of which are beautifully written (Freud didn't in the Goethe Prize for nothing), if scientifically lacking. But already in 1918, Hermann Hesse (author of Siddhartha) noted that artists of the younger generation were discussing Freud. In 1922, T.S. Eliot -- the poet (author of The Hollow Men and The Waste Land) and critic -- argued that there were three major forms of the novel in English:

  1. "the old narrative method" running from Henry Fielding through Jane Austen to Henry James;
  2. the "dangerous" method of Dostoyevsky, depicting characters' internal psychological and spiritual lives, and confronting big philosophical issues;
  3. and, finally, "the psychoanalytic type", depicting conflicts related to unconscious sexuality .

Eliot found the psychoanalytic type "interesting", "but of very short ancestry", and noted that it "rests upon a dubious and contentious branch of science" ("London Letter", The Dial, September 1922).

Over the years, psychoanalysis has become even more dubious and contentious, but it's impossible to understand the arts and culture of the 20th century without it. So you'll find a brief introduction to Freudian psychoanalysis, and to "neo-Freudian" approaches to psychoanalytic theory, elsewhere in these Lecture Supplements:

Neo-Freudian Psychodynamics

The Reciprocal Effect of Behavior on the Person

              (44141 bytes)So much for the effects of the Person on Behavior. What about the reciprocal effect of Behavior on the Person?

Self-Perception Theory

The B==>P relation is illustrated by Daryl Bem's self-perception theory of attitudes, which reverses the traditional account of attitude-behavior relations. Common sense tells us that attitudes come first, and lead us to behave in attitude-consistent ways. If I'm a registered Democrat, then I'll probably vote for the Democratic candidate in an election. My attitudes cause me to behave in particular ways. But Bem argued that the actual direction of causality is reversed: that behavior causes attitudes to form. Specifically, Bem argued that we infer what our attitudes are from observing our own behavior -- in exactly the same way that we infer what other people's attitudes are from observations of their behavior. Put another way, our perception of our own behavior leads us to form attitudes that are consistent with that behavior. In this sense, the behavior comes first, and the attitude comes later.

freedmanfoot.gif (4732 bytes) Self-perception theory is illustrated by the "foot-in-the-door effect" on compliance. In a classic study, Freedman and Fraser (1966) canvassed neighborhoods in the guise of a California "safe driving" campaign. They asked homeowners in a random half of households in particular neighborhoods to sign a petition in support of safe driving. The canvassers did not request any donations for the campaign -- only signatures. Almost everyone complied with this request. Later, the canvassers returned to all the households in the neighborhood, and asked permission to place a large, unattractive poster in their front yards promoting the campaign. Homeowners who had earlier signed the petition were more likely to agree to the poster than those who had not been asked to sign. The trick in the experiment, of course, is that almost everyone who had been asked to sign the petition actually did so. So there could not have been any difference between the two groups in their support of safe driving. Rather, it was the act of signing the petition that made homeowners more likely to permit the poster.This is the "foot in the door" effect -- that granting a small favor makes it more likely that the person will later grant a large favor. The effect, in turn, is explained by Bem's self-perception theory.

cialdini_7AM.gif (4747 bytes)In another study, Cialdini et al. (1976) recruited subjects for an experiment by telephone. Some subjects were informed that the experimental session was scheduled for 7:00 AM before they agreed to participate; others were given this information afterwards. Of course, some subjects in the first condition didn't agree to participate, but we are interested only in those who did. The question is: how many subjects who agreed to participate actually appeared? The result may seem paradoxical: more subjects who agreed after learning the experiment was scheduled for 7:00 AM actually showed up. But if you look at it from the point of view of self-perception theory, the result is not paradoxical at all. Agreeing to participate creates a kind of "foot in the door", and increases the likelihood that the subject will take the more arduous step, and actually show up for the experiment.

chaiken_environmentgif.gif (6014 bytes)A third experiment, by Chaiken and Baldwin (1981), also illustrates self-perception theory at work. In this study, subjects first completed a questionnaire concerned with attitudes toward environmental conservation. Some subjects had strong pro-conservation attitudes, and others were strongly opposed to conservation efforts (these were not necessarily the kind of person who would like to shoot the last elephant hiding behind the last tree in the rainforest, but rather the kind of person who thinks that conservation efforts are fine in their place, but should be tempered by such needs as energy self-sufficiency and economic development), but some subjects had weak, inconsistent pro/con attitudes. The subjects then completed a questionnaire in which they were asked to report on their conservation-relevant behaviors. Some subjects received a version of the questionnaire carefully worded in such a way as to encourage them to reflect on their anti-environment behaviors (e.g., that they did litter on occasion), while others received a version that encouraged them to reflect on their pro-conservation behaviors (e.g., that they frequently did not litter). When these subjects were subsequently retested, their attitudes shifted in the direction of their behavioral self-reports. Those who were encouraged to reflect on their anti-environment behaviors became more strongly opposed to conservation, while those who were encouraged to reflect on their pro-environment behaviors because more strongly favorable. No such shift occurred in those who already had strong attitudes either way. For those who had weak, vague, inconsistent attitudes about environmental conservation before the experiment began, their reflections on their own behavior appear to have created attitudes that did not exist before.

Actually, the Chaiken-Baldwin experiment provides an example of what is known as the person-by-situation interaction. Attitude strength is a feature of the person. Reviewing one's past experiences is a task set by the situation. The task affects attitudes, but only for people who had weak attitudes to begin with. People with initially strong attitudes about the environment, whether pro or con, were unaffected by the situational manipulation.

How to Find Yourself President of Harvard

Drew Gilpin Faust, the first woman to serve as president of Harvard University, tells the story of how she got the job ("Weekend Confidential" Drew Faust" by Alexandra Wolfe, Wall Street Journal, 02-1-2/2014).  There she was, happily teaching history at the University of Pennsylvania, and doing her share of committee service, when Neil Rudenstine, who was then the president of Harvard, called to ask if she were interested in becoming Dean of Radcliffe College (which used to be the women's college attached to Harvard). 

The way he framed it I'll never forget, and I've used it on other people," she explains.  As she recalls, he said, "If you have 5% interest in this job, will you just keep talking to me?" She thought 5% was easy to agree to and has since used the line in hiring, including with her current provost Alan Garber, who left Stanford to take the job.

Rudenstein was replaced by Larry Summers, who resigned in 2006, partly following some ill-considered, if not downright ignorant, remarks he made questioning women's innate talent for math and science.  And so Harvard started searching for a new president, and considering maybe they should have a woman for once, and Dean Faust was right there!

The James-Lange Theory of Emotion

The reciprocal effect of Behavior on the Person is also illustrated by the James-Lange theory of emotion. Traditionally, we think of our behaviors are caused by our internal emotional states: when we see the bear, in William James's famous formulation (1884), we run because we are afraid. But James and Lange reversed the direction of causality: in their view, emotion results from the perception of our own bodily responses to a stimulus. To continue James' famous formulation, when we see the bear, we are afraid because we run. James, as his formula suggests, emphasized the perception of our motor responses; Carl Lange, a Danish psychologist who came up with the theory independently of James (1885), emphasized the autonomic nervous system. When Walter B. Cannon (1935) famously criticized the James-Lange theory on the ground that there were no differential autonomic correlates of various emotional states, he was really talking about Lange's version of the theory. More recently, however, some evidence favoring James's version has come from studies of the facial expressions of emotion.

For many years, UCSF psychologist Paul Ekman and his colleagues have studied facial expressions of emotion as an aspect of nonverbal communication. Following Darwin, who thought that there was evolutionary continuity of emotional expressions between humans and nonhuman animals (e.g., The Expression of the Emotions in Men and Animals, 1872), Ekman has argued that there are a small number of innate "basic emotions" which expressed facially in very similar ways in humans across all cultures. Some of these emotions, and examples of their associated facial expressions, are:

Happiness 053Happy.jpg (51722 bytes) Sadness 054Sad.jpg (52168 bytes) Fear 055Fear.jpg (53440 bytes)
Anger 056Anger.jpg (50253 bytes) Surprise 057Surprise.jpg (50750 bytes) Disgust 058Disgust.jpg (50165 bytes)

There may be other basic emotions, as well: one candidate is contempt. But this list gives the basic idea.

Note that Ekman refers to these as emotional expressions, implying the traditional view that the emotion comes first, and the expressive behavior comes later. Or, in Lewin's terms,P (emotional state) ==>B (facial expression). However, the James-Lange theory asserts that the expressive behavior comes first, and the emotions later:B (facial expression) ==>P (emotional state). Of course, from the point of view of reciprocal determinism, both could be true: there could be a kind of feedback relationship between emotional state and emotional behavior.

FacialFeedback.JPG (76469 bytes)The James-Lange theory was formalized by Sylvan Tomkins in the facial feedback hypothesis of emotion, which states that feedback from the face affects the person's emotional state. The facial feedback hypothesis, in turn, comes in two forms:

  • The weak version, proposed by Darwin (1872) himself, states that emotional expressions do not create the emotion, but can modulate an emotion that is already present. If you're angry, putting on a happy face can make you feel less angry. Similarly, William James suggested that if an emotion is not expressed in behavior, it is not felt, either.
  • The strong version, proposed by Tomkins himself and elaborated on by Laird (1974), proposes that facial expressions are sufficient to create an emotional state, all by themselves.

"Put On a Happy Face"

From Bye, Bye Birdie!"

Gray skies are gonna clear up,
Put on a happy face;
Brush off the clouds and cheer up,
Put on a happy face.
Take off the gloomy mask of tragedy,
It's not your style;
You'll look so good that you'll be glad
Ya' decide to smile!

Pick out a pleasant outlook,
Stick out that noble chin;
Wipe off that "full of doubt" look,
Slap on a happy grin!
And spread sunshine all over the place,
Just put on a happy face!
Put on a happy face
Put on a happy face

And if you're feeling cross and bitterish
Don't sit and whine
Think of banana split and licorice
And you'll feel fine
I knew a girl so glooming
She'd never laugh or sing
She wouldn't listen to me
Now she's a mean old thing
So spread sunshine all over the place
Just put on a happy face
So, put on a happy face

LairdPhotos.JPG (66565 bytes)In any event, is there any evidence for reversing the direction of causality? Yes. Laird et al. (1989) studied this question in a psychophysiological experiment in which they used the electromyogram (EMG) to record the activities of subjects' facial muscles. Under the guise of "calibrating the equipment", Laird induced subjects to move their facial muscles in such a way as to mimic the facial expressions of emotion documented by Ekman. At the same time he collected ratings of the subjects' moods. Lo and behold, when the subjects put on the "fear" face, they actually felt afraid. The same thing happened for anger and sadness. The results were a little more complicated for disgust, but overall it appeared to be the case that putting on the emotional face caused changes in subjects' emotional mood states.

There are some interesting twists here. When Botox is employed cosmetically for the treatment of "frown lines", it is injected into precisely those facial muscles, the corrugator muscles, that are involved in producing frowns. One side-effect of these Botox injections appears to be the alleviation of depression. In a study by Lewis et al. (2009), a group of women receiving Botox for treatment of frown lines scored lower on questionnaire measures of anxiety and depression than a control group. The idea is that, by paralyzing the corrugator muscles, Botox prevents feedback from those muscles from depressing the patient's mood. Finzi claims that Botox is safer, and possibly more effective, than any other treatment for depression, psychological or pharmacological. See the Face of Emotion: How Botox Affects Our Mood and Relationships by Eric Finzi (2013), a dermatologist who frequently treats patients with Botox..

You can get a sense of what happened in Laird's study by mimicking a clever experiment performed by Strack, Martin, and Stepper (1988), in which subjects who held a pen in their mouths experienced different emotions depending on whether they held the pen between their teeth, thereby forcing a smile, or between their lips, thereby inhibiting smiling.

Put on Your Sunday Clothes

From Hello, Dolly! (revived in the Pixar film WALL-E)


Out there
There's a world outside of Yonkers
Way out there beyond this hick town, Barnaby
There's a slick town, Barnaby
Out there
Full of shine and full of sparkle
Close your eyes and see it glisten, Barnaby
Listen, Barnaby...
Put on your Sunday clothes,
There's lots of world out there
Get out the brilliantine and dime cigars
We're gonna find adventure in the evening air
Girls in white in a perfumed night
Where the lights are bright as the stars!
Put on your Sunday clothes, we're gonna ride through town
In one of those new horse-drawn open cars

Cornelius & Barnaby:

We'll see the shows at Delmonicos
And we'll close the town in a whirl
And we won't come home until we've kissed a girl!


Put on your Sunday clothes when you feel down and out
Strut down the street and have your picture took
Dressed like a dream your spirits seem to turn about
That Sunday shine is a certain sign
That you feel as fine as you look!

Dolly & Ambrose:

Beneath your parasol, the world is all a smile
That makes you feel brand new down to your toes

Dolly, Ambrose, Cornelius, & Barnaby:

Get out your feathers, your patent leathers
Your beads and buckles and bows
For there's no blue Monday in your Sunday...
No Monday in your Sunday...
No Monday in your Sunday clothes!

Townspeople, All:

Put on your Sunday clothes when you feel down and out
Strut down the street and have your picture took

Dolly, Townspeople, All:

Dressed like a dream your spirits seem to turn about
That Sunday shine is a certain sign
That you feel as fine as you look!


Beneath your parasol, the world is all a smile


That makes you feel brand new down to your toes
Get out your feathers
Your patent leathers
Your beads and buckles and bows
For there's no blue Monday in your Sunday clothes!

[Dance interlude]


Put on your Sunday clothes when you feel down and out
Strut down the street and have your picture took


Dressed like a dream your spirits seem to turn about

Townspeople, All:

That Sunday shine is a certain sign
That you feel as fine as you look!
Beneath your bowler brim the world's a simple song
A lovely lilt that makes you tilt your nose
Get out your slickers, your flannel knickers
Your red suspenders and hose
For there's no blue Monday in your Sunday clothes!


Ermengarde keep smiling no man wants some little meaning!
Ambrose do a turn, let me see!
Mr.Hackl, Mr.Tucker don't forget Irene and Minnie,
Just forget you ever heard a word from me!
All Aboard!


All Aboard!


All Aboard!


All Aboard!




All Aboard! All Aboard! All Aboard! All Aboard!

Put on your Sunday clothes there's lots of world out there
Put on your silk cravat and patent shoes
We're gonna find adventure in the evening air


To town we'll trot to a smokey spot
Where the girls are hot as a fuse!




Put on your silk high hat and at the turned up cuff
We'll wear a hand made gray suede buttoned glove


We wanna take New York by Storm!


We'll join the Astors
At Tony Pastor's
And this I'm positive of
That we won't come home
No we won't come home
No we won't come home until we fall in love!

A similar experiment, reported more recently, examined the effects of posture on power, status, and leadership. In the guise of a marketing test of ergonomic office chairs, Li, Galinsky, and their colleagues (2011) induced subjects to adopt an "expansive" posture (legs spread wide and arms reaching outward) or a "constricted" posture (legs together, hands under thighs, shoulders hunched). Later on, they engaged in a number of simulated activities in which they had the opportunity to exercise leadership. Subjects who had earlier adopted the expansive "power" posture were more likely than those who had adopted the constricted "subordinate" position to choose to speak first in a debate, to leave the site of a plane crash to find help, or to join a movement to free someone who had been wrongly imprisoned. Apparently, behaving like a leader increased the person's level of power motivation and the tendency to exercise leadership.


Our final example of the influence of behavior on personality comes from perceived self-efficacy, a concept developed by Albert Bandura (1977) as part of his cognitive social-learning approach to personality. Bandura agrees with Lewin that personality factors, such as motives, determine how a person is going to behave, but argues that efficacy expectations mediate between the person and his or her behavior. If you're a basketball player facing a difficult three-point shot, and you don't think you can make it, you might not even try -- thus guaranteeing failure. But if you think you'll can do it, then you're likely to try -- and you might actually succeed.

Link to an interview with Albert Bandura.

Self-efficacy is the person's belief, or expectation, that he or she can act effectively to bring about desired results in some situation.

Bandura argues that self-efficacy expectations have a variety of sources:

  • vicarious experience: you see other people like you attempt that same shot and succeed;
  • verbal persuasion: your coach tells you that you can make it;
  • emotional arousal: you feel good, "pumped up"; and, most critically for our purposes here,
  • performance accomplishments: you've actually made that shot before, perhaps under similar circumstances.

Thus, performance accomplishments -- actual behavior -- feeds back to revise the person's self-efficacy expectations, and thus his or her motivation to engage in some activity. Expectancies and motives don't just determine behavior -- behavior also determines expectancies and motives.

"Whistle a Happy Tune"

                  (83604 bytes)The idea that our behavior can affect our internal states and traits is beautifully expressed in this song from The King and I, the 1951 musical comedy by Rogers and Hammerstein. The musical is based on the true story of Anna Leonowens, a widowed Englishwoman who from 1862-1866 served as tutor to the children of King Mongkut (also known as Rama IV) of Siam (now Thailand). As they arrive at the wharf in Bangkok, In the play, Anna and her son Louis are greeted by the imperious and hostile Kralahome, the Prime Minister. Louis asks Anna if she's afraid of what will happen on her new job in such a strange country. She confesses to a little fear, but then she shares the strategy that she uses to control her feelings:

Whenever I feel afraid, I hold my head erect
And whistle a happy tune, so no one will suspect
I'm afraid!

While shivering in my shoes I strike a careless pose
And whistle a happy tune And no one ever knows
I'm afraid!

The result of this deception is very strange to tell
For when I fool the people I fear I fool myself as well!
I whistle a happy tune and ev'ry single time
The happiness in the tune
Convinces me that I'm
not afraid!
Make believe you're brave and the trick will take you far
You may be as brave as you make believe you are!

Link to a performance of the song by Gertrude Lawrence, who originated the role of Anna on Broadway.

Anna's technique is very closely related to Bandura's concept of self-efficacy, as well the self-fulfilling prophecy, about which more below. For now, it's enough to say that behaving as if something is true sometimes has the effect of bringing it about.

The Dialectic Between the Environment and Behavior

According to Lewin's formula, environmental factors are also relevant to behavior. These include all the features of the situation that exist "outside" the person. As a social psychologist, of course, Lewin was mostly interested in the social situation: the presence of other people, and what they say and do.

Situational Influences on Behavior

From the point of view of the dialectic between the Environment and Behavior, features of the environment, especially the social environment, elicit behavior regardless of the individual's internal states and dispositions. This direction of causality is entailed in the Doctrine of Situationism, and is exemplified by a large body of social-psychological research illustrating social influences on behavior. In each case, the subject's behavior is influenced by the environmental situation in which that behavior occurs.

Delay of Gratification

Research on delay of gratification in children nicely illustrates the role of situational, as well as personal, determinants of behavior. In a study by Mischel & Ebbesen (1970), children were asked which of two rewards, cookies or pretzels, they preferred. Then the children were told that the experimenter would go away for a while (actually, about 15 minutes). If they waited for the experimenter to return, they would receive their preferred reward. But if they could not wait, they would receive their non-preferred reward. Then the experimenter left. In one condition of the experiment, he took the cookies and pretzels with him. In other conditions, he left one or the other behind. In a fourth condition, he left them both behind. How long did the children wait before signaling the experimenter to return?

MischelEbbesen_1970.JPG (56836 bytes)In fact, children who waited in the absence of both rewards were able to wait a fairly long time -- some of them even outlasted the experimenter! If either the preferred or the non-preferred reward was left with the child, waiting time decreased. And if the child was left to wait in the presence of both rewards, waiting time dropped almost to zero. Young children cannot delay gratification long in the presence of a reward. But this experiment shows clearly the influence of the situation -- whether rewards are present or not -- on the child's behavior.


A classic study of the effects of the social situation example is Asch's research on conformity. In Asch's experiment, subjects were recruited for what was presented as an experiment on perception, in which they were asked to indicate which of three comparison lines matched a standard line in length. This is a straightforward perceptual task, on which subjects make few errors -- although of course the task is more difficult on some trials than on others. In Asch's experiment, subjects were either individually or in groups. If they were run in groups, all the other "subjects" were actually confederates of the experimenter, acting in accordance with a prearranged script -- unbeknownst to the real subject, of course.

              (49317 bytes)In one group experiment, all the "subjects" (including the confederates) announced their judgments publicly, and on critical trials the situation was arranged so that the one "real" subject went last. On some trials, the confederates were instructed to agree on the correct response. On other trials, they were instructed to agree on the incorrect response. When subjects were alone, or run with only one or two confederates, their judgments were largely unaffected by the group. But once the group consisted of three or more individuals (Asch ran groups as large as 15), subjects began to conform to the (erroneous) judgments of the group, in about 1/3 of the trials.

066Conformity.jpg (30262 bytes)In the previous experiment, the confederates were unanimous. In a follow-up study, the confederates were unanimous on some trials; but on others, one confederate dissented from the majority. When the subjects were run individually, again, they made few mistakes on the matching task. In the group trials, when the confederates were unanimous, subjects conformed to their erroneous judgment on about 32% of trials -- about the rate observed in the previous study. When there was a single dissenter, however, conformity dropped to about 6%. Thus, the individual was conforming to group norms, and the degree of conformity depended on whether the others were unanimous. A feature of the social situation -- what other people said -- affected the subject's behavior.

In a variant on Asch's experiments, Nemeth and her colleagues (working at UCB) assembled groups consisting of six real subjects, plus 1-4 confederates. Thus, in Nemeth's experiments, the confederates comprised a minority, rather than a majority, of the group.

Nemeth.JPG (45306
              bytes) Even though they were in the minority, the subjects still showed a tendency to conform -- so long as the confederates constituted a "critical mass" of 3 or 4.

Especially when the situation is ambiguous, we look to others for cues as to what to think, feel, want, and do.

"Mere Presence" Effects

Another large literature concern social presence effects on behavior -- the effects of the presence of other people on the behavior of the individual.

  • In social facilitation, the presence of others improves the individual's performance. Social facilitation was first discovered by Triplett (1898), in the very first social-psychological experiment ever published -- although a more recent re-analysis showed that Triplett's effect was not actually statistically significant.
  • In social inhibition, first discovered by Zajonc (1965), the presence of others reduces the individual's performance. In a famous experiment, Zajonc showed that even cockroaches run more slowly in the presence of other cockroaches.

OK, so social facilitation and social inhibition contradict each other. Which is the true effect of the presence of others? Of course, it depends on the task:

  • Social facilitation tends to occur when the task in question is simple, is performed automatically, or is performed by an expert.
  • Social inhibition tends to occur when the task in question is difficult or complicated, is performed by controlled processes, or by novices.

SocialLoafing.JPG (56339 bytes)Social facilitation and social inhibition are effects of the mere presence of other people. But what if the individual is working with other people. Latane and his colleagues have documented a social loafing effect. Subjects were brought into the laboratory and asked to cheer, clap, or shout as loudly as they possibly could. They performed best when they were alone. But when they shouted, etc., along with other people, they shouted less loudly -- and the shouting got progressively less loud as more people were engaged in the task.


Features of the situation can also affect people's tendencies to behave in an aggressive, hostile manner toward each other.

CrimeSeason.JPG (54143 bytes)With respect to the physical environment, Anderson (1989) reviewing crime statistics, found that the rate of violent crimes such as rape and murder was highest during the summer, and lowest during the winter. There seems to be something about high levels of heat and humidity that brings out the worst in us.

The relationship between heat and aggression was shown clearly in a clever field study by Kenrick and MacFarlane (1986).  Kenrick is on the faculty at Arizona State University, near Phoenix, where it gets very hot in the spring and summer.  They arranged for a confederate to stop her car at a green light, and then determine the number of drivers behind her who honked their horns, and for how long. They found a linear relationship between horn-honking and a "discomfort index" based on both temperature and humidity.  the relationship was especially strong when the drivers had their windows rolled down, suggesting that their cars didn't have air conditioning!

One of the most popular situationist theories of aggression has been the frustration-aggression hypothesis originally proposed by John Dollard (Dollard et al., 1939) and Neal Miller (1941), who were colleagues at the Institute for Social Research at Yale. Dollard and Miller argued that aggression is a reflexive response to frustration, where frustration was defined broadly to include any obstacle to goal-attainment. Such obstacles, of course, are typically to be found in the environment.

Research in the ensuing years led Berkowitz (1989, 1993) and others to revise the original frustration-aggression hypothesis somewhat. According to the revised frustration-aggression hypothesis:

  • The definition of frustration is broadened to include any aversive event that is construed as intentionally harmful.
  • Aggression is no longer considered to be an involuntary reflex, but rather a voluntary behavior.
  • The pathway from frustration to aggression is mediated by an emotional state, namely anger.
  • In response to frustration, anger is most likely to lead to aggression if the environment contains cues to aggressive behavior.

Thus, in the revised frustration-aggression hypothesis, the important situational feature is not an obstacle on the way to a goal, but rather the presence of environmental cues for aggressive behavior.

WeaponEffect.JPG (50586 bytes)As an illustration of what Berkowitz has in mind, consider an experiment by Berkowitz and LePage (1967) on anger, weapons, and frustration. In this experiment, two subjects (one of whom, of course, was a confederate of the experimenter) were brought into a room to work together on some problems. At the end of the session, they were asked to evaluate each other's performance on the problem solving task by giving each other from one to ten electrical shocks (1 shock for really good performance, 10 shocks for really bad performance). The confederate went first, and according to prior arrangement gave some subjects 1 shock, and others 7 shocks. Of course, this "feedback" was entirely independent of the subjects' actual performance, which was overall quite good (and they knew it).

The intent of this procedure, of course, was to induce anger in the subjects. When it came the subjects' turn to evaluate the confederate, of course, they had an opportunity to retaliate -- to aggress -- against the confederate. And so they did: subjects in the angered condition delivered more shocks than those in the control condition.

The really important variable in this experiment, however, was the presence of cues for aggression. As amazing as it is to think about it, in some conditions the laboratory room was outfitted with a number of actual guns - -a revolver, a shotgun, and the like. In a control condition, the weapons were replaced by badminton equipment -- racquets, shuttlecocks, and the like. There was also a control condition in which there were no objects of any sort in the room. Within the angry condition, retaliatory aggression reached its highest levels by subjects who were run in the presence of the weapons -- even when the experimenter made it clear that the weapons did not belong to the subject. Apparently, the mere presence of weapons in the room -- a situational cue -- was enough to increase the amount of aggression directed by the subject toward the confederate. This is known as the weapon effect.

In subsequent years, the weapons effect has stimulated considerable research on the effects of medial portrayals of violence on aggressive behavior. It is well known, for example, that aggressiveness is significantly correlated with exposure to violence on TV, in the movies, and on video games. Whether the exposure causes the aggression, or whether they are both effects of a third causal variable, remains a point of somewhat heated controversy. But the correlation between media violence and personal aggressiveness may be a real-world instance of the weapons effect.


Returning to the social situation, a large body of research on social influence was stimulated by the tragic case of Kitty Genovese, a woman who, according to the familiar story, was brutally assaulted and murdered while a large number of her neighbors looked on without doing anything to help her. The Genovese case was first brought to public attention by A.M. Rosenthal, then a Pulitzer Prize-winning reporter for the New York Times (and later its executive editor) in a series of newspaper articles and later a book, Thirty-Eight Witnesses: The Kitty Genovese Case (1964). Rosenthal's reporting, in turn, was initially based on information from the New York City police commissioner.

Actually, though, the popular story is something of an urban legend.
  • Manning, Levine, and Collins investigated the case, and determined that, contrary to legend, 38 people had not witnessed the stabbing, only a few witnesses thought it was an attack (as opposed to a neighborhood domestic dispute), nobody actually saw the murder itself, and some witnesses actually did try to help (American Psychologist, 2007).
  • For further analyses of the case, see also:
    • Kitty Genovese: The Murder, the Bystanders, the Crime that Changed America by Kevin Cook (2014).
    • Kitty Genovese: A True Account of a Public Murder and Its Private Consequences by Catherine Pelonero (2014)

Still, the overriding question was: what causes people to help? We usually refer to altruism as a personality trait, assuming that some people are more altruistic than others, and that altruistic people are more likely to render assistance in an emergency, even at some cost to themselves. That would be an example of B=f(P). But, of course, features of the social situation may be important too. Surprisingly, it seems that the presence of others can actually deter helping behavior.

070Helping.jpg (38566 bytes)In classic research on "bystander intervention", Darley and Latane recruited subjects for an experiment in which they were asked to complete some straightforward personality questionnaires. Some subjects completed the questionnaires alone in a small booth. Others completed the questionnaires in small groups (none of the subjects in these experiments were confederates of the experimenter). In one version of the experiment, the subjects were left alone, but midway through the session (nontoxic) smoke began to filter into the room through a ventilation duct. In another version of the experiment, the subjects were informed that the experimenter would be in the next room; midway through the experiment, they heard a crashing sound, followed by the experimenter's cry that she had fallen and broken her foot. The results were striking: when subjects were alone, more than 70% of them made some move to get help or render assistance; but when subjects worked in groups, the rate of helping behavior fell precipitously. In a later version of the experiment, when all the "subjects" but one were actually confederates of the experimenter who assiduously ignored what was going on, helping was reduced still further. In this case, the features of the social situation -- whether there are others present, and if so what they were doing -- affected helping behavior. 

Although the Darley-Latane experiment has been replicated many times, a "meta-analysis" of the bystander-intervention literature published in 2011 qualified its conclusions somewhat.  In simulations of real-life  dangers(as opposed to laboratory), with naive bystanders (instead of confederates) who could communicate with each other, the presence of others seems to facilitate helping behavior.  This does not contradict the basic conclusions of Darley and Latane, however.  Their basic point was that, under conditions of uncertainty, people look to others for cues as to what to think and how to behave.

Social Impact Theory

There are more such experiments in the social-psychological literature than you can shake a stick at. The vast bulk of social-psychological research, especially in the 1950s through the 1970s, concerned social influence in some way.

Many of these experiments are variants on the Asch conformity paradigm, examining the influence of real or imagined social pressure on the individual subject's expression of belief. Asch himself distinguished between two levels of conformity:

  • Conformity at the level of behavior -- that is, the subject's verbal or nonverbal expression of agreement with the majority.
  • Conformity at the level of judgment -- that is, not just a change in the subjects verbal report or other behavior, but an actual change in the subject's perceptual experience, opinion, belief, etc.

Asch thought that the conformity he observed might go to the level of judgment -- he was, after all, heavily influenced by the Gestalt movement in perception -- but he admitted he only had evidence for conformity at the level of behavior.

Other variants on the Asch experiment have explored other areas of social influence:

  • Obedience, where there is an unequal power relationship between influencer and influencee.
  • Compliance, or responses to explicit requests.

Social influence is a primary topic in what might be called "the 4 As" of social psychology:

  • attitudes (particularly stereotyping and prejudice)
  • aggression (e.g., the work of Leonard Berkowitz)
  • attraction (e.g., the work of Elaine Hatfield and Ellen Berscheid)
  • altruism (e.g., the work of Darley and Latane).

In each case, a large body of research shows the influence of the social situation on people's behavior. There is also a fairly large body of research, much of it also conducted by social psychologists, on the influence of the physical environment -- density of people, noise levels, etc. -- on behavior. Traditional social psychology is, to a very great degree, the study of environmental influence, with an emphasis on the social environment.

In fact, Latane (1981) -- the same researcher who was involved in the bystander intervention studies -- has proposed social impact theory to predict the effects of "the presence and behavior of other people" on a target individual's behavior. First, Latane proposes that:

SI = f(S, I, N),


S = the Salience of the other people, such as their status relative to the target;

I = their Immediacy, or proximity in space and time; and

N = the sheer Number of sources of social influence.

With respect to N, Latane has further proposed a "psychosocial law",of social impact:

SI = sNt,

where t , 1.

Of course, you will recognize this as a variant on Stevens' Law in psychophysics,S =kIn-- and the resemblance is quite intentional. For Latane, social influence is analogous to sensory experience, and social stimulation is analogous to physical stimulation.

  • Social pressure on the individual increases as the number of sources of social pressure increases, but as in Fechner's Law, social influence grows more slowly than social sources -- which is why the exponent, t, is less than 1. As in Asch's conformity research, it's the first few -- the first three or four -- confederates who make a real difference in whether the subject conforms; after that, adding confederates doesn't make that much of a difference.
  • Of course, the individual can also be a group member, part of a minority group subject to social influence by a minority. As the number of individuals allied with the subject increases, the social pressure on the individual decreases, because the total social pressure is diffused among a number of different targets. This diffusion effect also follows the psychosocial law. As in Nemeth's experiments, a few allies make a big difference to whether the individual conforms -- "three is enough", and beyond that adding allies doesn't make that much of a difference.

It should be understood that Latane's social impact theory, and especially the psychosocial law, represents situationism in its strongest form: it's the objective situation, described in terms of the actual number of people in the environment putting pressure on the individual, that determines whether the target will conform.

Social Contagion

SixDegreesPoster.jpg (14771 bytes)For SixDegreesMovie.jpg (37998 bytes) social psychologists, the environment is a social entity, consisting as it does of other people. And there are a lot of people in our social environment.  Robin Dunbar, a British evolutionary biologist, has estimated that an individual's social group consists of about 150 people -- a figure sometimes known as "Dunbar's number"..  Based on studies of nonhuman animals, Dunbar figures that this is the maximum number of people that we can keep track of, given the size of our brains.  According to Dunbar, our core group of social relationships consists of about 5 people, with whom we talk once a week, and about 100 acquaintances to whom we speak about once a year.  Of course, given social media like Facebook, any individual may know a lot more people than that, but Dunbar claims that if people who claim more than about 150 people in their social circle probably don't know much about them, and he doubts that they are genuine friends.

  • In a short story, "Chains", published in 1929, the Hungarian author Frigyes Karinthy suggested that no individual lies more than six people away from any other person.
    • His idea, in turn, was based on a remark by Gugleimo Marconi, the inventor of radio, in his 1909 Nobel Prize acceptance speech.
  • The social psychologist Stanley Milgram studied the "small world" problem by asking subjects in in Omaha, Nebraska and Wichita, Kansas to get a letter to a target living in a suburb of Boston, Massachusetts (Milgram originally published his observations in Psychology Today in 1967; the technical report, written with Jeffrey Travers, appeared in Sociometry in 1969). They were to give a postcard addressed to the target to someone who might know someone who might know someone who might know someone in Boston who knew the target -- you get the idea. On average, delivering the envelope required about 5.2 intermediaries.
    • Try this for yourself. Think of some faraway place, and then think of someone you know, who might know someone who might know someone who lives there.
    • Milgram's experiment was the basis for John Guare's play, "Six Degrees of Separation" (1990), later made into a film starring Will Smith, Stockard Channing, and Donald Sutherland (1993); and also for the popular Hollywood game, "Six Degrees of Kevin Bacon".
  • Milgram's experiment has some methodological flaws (he was a provocative thinker, but not such a great experimentalist). For example, many of the people to whom he gave postcards never mailed them at all, so that the cards that were eventually received by the target represented a biased sample of the whole set. Still, it turns out that Milgram -- or should we say Karinthy -- was onto something.
  • In 2001, Duncan Watts asked about 48,000 internet users to send a message to one of 19 targets, and found an average number of 6 intermediaries between initial sender and final recipient.
  • A 2008 study of Internet chatroom messages by Jure Lescovec and Eric Horvitz found that the average chain linking any two people in a group of 240 million people worldwide was 6.6.
  • A study involving 721 million Facebook users, yielded an estimate of 4.74 acquaintances separating any two people in the world (the average was 4.37 for the US, which hosts more than half of all Facebook users). As a representative of Facebook put it, given any two people, "a friend of your friend probably knows a friend of their friend".

All of these subsequent experiments have methodological problems, too -- it's not easy to do this kind of research. But it's pretty clear that we're all petty interconnected -- and every reason to think, thanks to social media, that we're getting more so.

And these social networks have effects on us, as they form the structural basis for social contagion effects, in which some piece of social behavior is transmitted from one individual to another -- a little like a virus.

Nicholas Christakis, a physician and sociologist) and James Fowler, a politicla scientist,have pioneered the study of social contagion effects through large social networks.  One of their early studies involved participants in the Framingham Heart Study, a large-scale longitudinal study which, since 1948, has followed about 15,000 residents of Framingham, Massachusetts, and their descendants, given them a comprehensive physical examination every four years.  They found that health behaviors related to smoking and obesity were contagious, with friends and neighbors seeming to "infect" each other for better (quitting smoking) or worse (gaining weight).  such "contagion" effects had long been discussed in the sociological literature, but Christakis and Fowler were the first to demonstrate them conclusively. 

In another study, the Christakis-Fowler group observed the social contagion of positive and negative emotion through networks of "friends" on Facebook.  For more than three years (1180 days, to be exact), Coviello and his colleagues (2014), including Christakis and Fowler, collected "status updates" from more than 100 million Facebook users in 100 large US cities, and coded them for positive and negative language.

  • To begin with, the researchers found a number of situational effects on happiness.
    • Happiness  was greater on weekends and holidays than on weekdays.
    • There was also geographic variation: people living in the Midwest, and on the West Coast, were in general happier than those living on the Eastern Seaboard.
  • Most critically, from the point of view of the study, the weather influenced expressions of happiness.
    • Not surprisingly, the number of negative posts increased, and the number of positive posts decreased, on rainy days.
  • They then examined subsequent posts made by the subjects' Facebook "friends". 
    • Every negative post by a subject in City A generated an average of 1.29 negative posts by his friends in City B -- even if it hadn't been raining in City B.
    • And similarly, every positive post by a subject in City A generated an average of 1.75 positive posts by his friends in City B -- even if it was raining in City B.

So, bad weather negatively affected A's mood, which elicited a negative behavior from A which elicited expressions of negative mood by his friend B.  Although the researchers didn't examine this, it seems very likely that social contagion generated a feedback cycle:

  • A vicious cycle, in which B's negativity reinforced A's. 
  • And there was probably another virtuous cycle, in which B's positivity reinforced A's

All of which began with A's positive or negative posts, which were elicited by the weather. 

In another study, Poncela-Casanovas et al. (2015) studied a group of 5,000+ registrants on weight-loss website.  For those who stuck with the program for six months or more, the best predictors of weight loss ere (1) initial body-mass index (BMI); (2) frequency of weigh-ins (the more the better); and (3) the number of friends who were also in the network.  Clients with no friends on the network averaged about a 4% weight loss; those with two or more friends averaged more than 8% weight loss.  The more friends a client had who were also on the website, the more weight she (they were mostly women) lost -- about 0.7% per friend. 

C&F's research on social contagion is controversial.  For example, some critics have argued that their findings are contaminated by homophily -- the tendency of individuals to associate with others who are like them.  If you're obese, you tend to hang around other obese people.  And other critics think that their findings are a product of shared environment: if friends and acquaintances tend to eat at the same greasy spoon diner, they're all going to put on weight.  But frankly, I think they've got the goods.  The important thing to understand, though is that this influence is bidirectional.  Each person constitutes part of the environment for some other person.  So while our behavior is influencing other people in our social network, their behavior is influencing us.

The Notorious Facebook Study

The study just referred to, by Coviello et al. (2014), is not that Facebook study, the one where Facebook subscribers' emotional states were unwittingly affected by manipulations of their newsfeeds.  But let me talk about that study for a bit, to make a couple of points.

The study, conducted by Adam Kramer, an in-house researcher at Facebook, Jamie Guillory (a postdoctoral fellow at UCSF), and Jeffrey Hancock (a computer scientist at Cornell), made use of Facebook's practice of collecting extensive information on the posts of its subscribers, mostly for use in targeted advertising (a source of revenue), but also to create a database for social-science  research.

In their study, Kramer et al. experimentally manipulated the "New Feeds" that almost 700,000 Facebook subscribers received from their "friends".  In one condition they received an edited selection of News Feeds that reduced the number of positive or negative words in their friends' posts.  Then, using computer software developed by James Pennebaker, a social psychologist at UT Austin, they counted the number of positive and negative words in the subscribers' own subsequent posts over a period of one week in 2012.  The researchers' didn't change the subscribers' "Wall" or "timeline" -- they just edited the postings that came across the "News Feed". 

Here's what they found.  When affective negativity was reduced in the News Feeds, the subjects' own posts contained more positive, and fewer negative, words.  When affective positivity was reduced, the subjects own posts contained fewer positive, and more negative words.  Apparently, the emotion expressed in the news feeds affected the recipients' own emotional states -- just as Coviello et al. had found earlier.  There was also an interaction, such that the effect of omitting positive words was stronger than that of omitting negative words.  (Yes, there was also an effect of emotional content on the sheer number of words written, but the researchers took that into account as well.)  The contagion effect was pretty small, corresponding to d =.001, which is pretty small indeed by Cohen's terms.  But, as the authors note, when this admittedly small effect permeated through almost 700,0000 users over the course of a week, the aggregate effects were not trivial. 

So why did Kramer et al. feel the need to do this all over again?  Recall from the lecture on Methods and Statistics that there are what Lee Cronbach (1957) called "two disciplines of scientific psychology of psychological research: correlational and experimental.  The Coviello study was essentially correlational in nature, and as we learned in that lecture, correlation does not necessarily imply causation.  Coviello showed a correlation between one Facebook user's mood and those of others in his social network, but their demonstration wasn't completely convincing to those who believe that causation can only be established by studies that employ random assignment of subjects to experimentally manipulated conditions. So they did an experiment which had both of these qualities, and confirmed what Coviello found.  

So what made this experiment notorious? Well, to begin with, they neglected to obtain their subjects' informed consent to participate in this research -- which, normally at least, would count as an egregious violation of experimental ethics.  These days, it's customary for research publications to carry a statement that the subjects gave informed consent to participate in the study, and that the research itself had been approved by an Institutional Review Board (IRB) for the protection of human subjects in research.  Both statements are absent from the publication in the Proceedings of the National Academy of Sciences, the house-organ of the most prestigious scientific society in the US (the editor of the article, also a distinguished social psychologist, apparently failed to notice this).

Automaticity in Social Behavior

Automaticity.JPG (51204 bytes)Another, almost-pure expression of situationism in social psychology is the argument that most social behavior is executed automatically. In this line of theorizing, automaticity is defined along the lines originally set out in cognitive psychology. In this formulation, automatic behavior is inevitably evoked by the presence of an adequate stimulus in the environment.

In one demonstration of the automaticity of social behavior, Bargh and his colleagues (1996) recruited subjects for an ostensible experiment on language processing, in which their task was to unscramble randomly arranged words so that they formed meaningful sentences.

  • In one condition the puzzles contained words related to rudeness, such as aggressively, rude, bother, disturb, and intrude.
  • In another condition the puzzles contained words related to politeness, such as respect, honor, considerate, appreciate, and patiently.
  • In a control condition the puzzles contained words unrelated to either rudeness or politeness, such as exercising, flawlessly, occasionally, rapidly, and gleefully.

              (48685 bytes)When the subjects had completed their assigned task, they exited their cubicles to find the experimenter engaged in conversation with someone else (another confederate, of course), and assiduously ignored the subject for a total of 10 minutes. Subjects in the "rude" condition were much more likely to interrupt the experimenter, and subjects in the "polite" condition were much less likely to do so, compared to the subjects in the control condition.

Bargh's interpretation is that reading "rude" or "polite" words automatically activated "rude" or "polite" behavior on the part of the subjects. Actually, it's not clear whether the words automatically elicited the behavior; perhaps the words automatically biased the subjects to perceive the experimenter as rude (or, in the "polite" condition, not to perceive him as rude).

Either way, the argument goes, the subject's behavior was a result of automatic processing of some feature of the stimulus environment. And, in fact, Bargh has argued that pretty much all social behavior is of this sort: executed automatically, without conscious awareness or conscious intent, in response to particular environmental stimuli. It is a pretty pure expression of situationism in social psychology.

The Effect of Behavior on the Situation

              (44910 bytes)Just as the environment influences behavior, so it is that behavior reciprocally influences the environment -- that is, behavior shapes the situation in which it occurs.

Altruism Redux

Interestingly, this point is effectively illustrated by Darley and Latane's research on altruism. They showed clearly that features of the situation influence helping behavior, but the big question is why?

  • Diffusion of Responsibility. Sometimes, individuals don't help because they are afraid, or they don't believe that they have the required skills. But sometimes, they don't help because they believe that someone else has also done so. In some cases, this is quite reasonable. Kitty Genovese's neighbors might well have believed that someone had already called the police (though 911 didn't exist then), and didn't want to clog the phone lines or otherwise complicate the situation. This diffusion of responsibility might account for some of the behavior of Darley and Latane's subjects as well. For all they knew, someone else had already noticed the smoke, or heard the experimenter's cry for help, and done something.
  • Pluralistic Ignorance. Another factor is that many emergency situations are inherently ambiguous -- that is, it's not entirely clear that they are emergencies. Two people fighting in a park might be roughhousing; they might even be rehearsing for a play, or making a movie. When a situation is ambiguous, there is a tendency in all of us to refrain from acting until the ambiguity is resolved -- until we determine whether the situation really is an emergency, and whether our help is really needed. We wait for the situation to be clarified -- and that clarification usually comes from other people. As Asch demonstrated in his experiments on conformity, we tend to look to others for cues as to what to think and do -- especially when the situation is ambiguous.

The problem comes when everybody does this. If I am trying to decide whether to help, and everybody else is doing the same thing, then nobody is helping -- and this lack of action effectively defines the situation as either too dangerous (in which I really shouldn't try to help) or not an emergency (in which my help really isn't needed). This state, in which everybody looks elsewhere for cues as to what to believe, is known as pluralistic ignorance. It is a prime example of how the behavior that takes place in a situation -- in this case, doing nothing -- creates a situation in which nothing can or need be done.

  • Other people's lack of action defines the situation as a non-emergency for the subject.
  • The subject's lack of action defines the situation as a non-emergency for others.

Of course, if the (inaction) of others shapes the situation, so does their action. In fact, when people observe others helping, they themselves are more inclined to help. Evidently, the behavior of other people defines the situation as one in which helping is needed, possible, and safe.

The fact that behavior can change the situation is illustrated by studies of the effects of modeling on helping behavior.

074Modeling.jpg (38636 bytes)In one study by Bryan and Test (1967), a female college student was stationed by a car on the side of a road (actually, I think it was California Highway 1) with an obviously flat tire, trunk open, and the spare leaning against the car. One-quarter mile away they stationed another car with a flat tire, this time an elderly couple. On one condition of the experiment, nobody was helping the couple. In another condition of the experiment, the couple was being helped by another motorist (the couple and the "good Samaritan" were, of course, confederates of the experimenters). More people stopped to help the college student when the couple was being helped, than when they were not.

Another study by the same investigators studied contributions to a Salvation Arm kettle during the Christmas season. On some trials, as a pedestrian approached the kettle, another person (again, a confederate of the experimenter) moved in and deposited some change. The pedestrian was much more likely to make a donation after this modeling episode, than if the modeling had not occurred.

Modeling is just one situation in which behavior changes the situation. The behavior of the model creates situational cues that can influence the behavior of other people. In the Darley and Latane experiment, an unresponsive model leads others not to intervene in an emergency. In the Bryan and Test studies, a helpful model induces others to help.

Learning Redux

If you think about it, instrumental or operant behavior effectively changes the environment in which it occurs. In fact, that's exactly how Skinner defined operant behavior -- as behavior that operates on the environment to change it in some way. The pigeon in a Skinner box can press a key to get food, and by doing so changes the environment from one that didn't have food to one that does. Much of our behavior has this quality -- it changes the environment, whether we intend it to do so or not.

We will say more about the effect of behavior on the environment when we discuss the dialectic between the person and the environment, and in particular the ways that people influence their environments through their behavior.

The Dialectic Between the Person and the Environment

To complete our overview of the reciprocal interactions between the person, his or her behavior, and the environment, we examine the dialectic between the person and the environment -- how environments go beyond eliciting behavior to shape the people in them, and how people reciprocally shape the environments in which their behavior takes place.

The Influence of the Environment on the Person

              (87987 bytes)First we look at the ways in which features of the environment shape the features of the people in them. This is, in fact, a continuation of our discussion of social influence. After all, social influence extends beyond the individual's behavior to his or her thoughts and feelings, beliefs and desires, traits, attitudes, and values -- the internal states and traits that cause the person to behave as he or she does.

"State" of Mind

Data from surveys conducted by the Centers for Disease Control and Prevention, a unit of the US Department of health and Human Services, indicates that people living in the Appalachian and Mississippi Valley regions of the United States experience relatively high levels of mental distress, defined as 2 or more weeks of stress, depression, and poor emotional health.  By (Moriarty et al., "Geographic Patterns of Frequent Mental Distress: Adults, 1993-2001 and 2003-2006", American Journal of Preventive Medicine, 2009).

Persuasive Communications

Along with studying situational influences on behavior, traditional social psychologists have devoted a great deal of effort to tracing the effects of the situation on attitudes. Recall that attitudes are internal, dispositional characteristics of the person, and that they are supposed to dispose people to behave in attitude-consistent ways. Attitudes, like traits, are assumed to be stable across time and consistent across situations, but they can change -- Democrats become Republicans and vice-versa. In particular, social psychologists discovered that attitudes can shift, depending on the situation in which they are assessed.

In a classic study of attitudes (I know, I know, all these studies are "classics" -- but they really are!), Janis and his colleagues (1965) asked subjects to read a set of essays on various controversial subjects (i.e., they were controversial in the 1960s!):

  • whether the federal government should invest money to find a cure for cancer;
  • whether there should be an increase in military spending
  • whether the government should invest heavily in the space program to put a man on the moon;
  • whether the viewing of "3-D" movies should be prohibited on health grounds.

The essays were specifically written to be either "pro" or "con" on each issue. Half the subjects were offered a snack while they read the essays. The other half of the subjects were not.

janis.gif (5872
              bytes) The important finding of the experiment was that subjects who snacked while they read the essays were more likely to agree with the argument in the essays, regardless of the topic -- moonshot or 3-D movies -- and regardless of the position the essay took -- pro or con. Assuming that agreement is a reflection of people's real attitudes, and not just a further example of Aschian conformity, this experiment seems to show that a feature of the situation -- whether food is present -- affects people's internal, attitudinal dispositions.

Interpersonal Attraction

Another example of situational influence on internal mental states comes in the domain of interpersonal attraction? "Liking" someone is a special case of an attitude -- it represents a favorable evaluation of that person. But what leads us to like someone?

Obviously, our attitudes toward other people are determined partly by what are known as target characteristics -- features of the person who is the target of our evaluation. For example, we tend to like other people who are

  • physically attractive (duh!),
  • competent (especially if they make an occasional endearing mistake), and
  • similar to ourselves (it's pretty much a myth that opposites attract).

Similarity and Interpersonal Attraction

Opposites don't attract: the research literature tells us clearly that we are more likely to be attracted to people who are similar to us in terms of personality and demographic characteristics. Internet dating services, such as and, capitalize on this tried-and-true social-science result, and "match" their members based on a "scientifically tested" algorithms that, basically, match people based on similarity.

At the same time, the virtues of online matches shouldn't be overstated. A 2012 review by Eli Finkel et al. identified at least three problems:

  1. The evidence for the superiority of similarity-matching algorithms isn't all that clear. In part, this is because the algorithms, the research that led to them, and subsequent research on their validity is, in many cases proprietary and not available for public inspection. In fact, a massive study published in 2010 by Portia Dyrenforth, involving some 20,000 subjects, showed that couples who were similar in personality were only slightly happier than those who were dissimilar. Similarity matching is a good place to start, but it is apparently neither necessary nor sufficient for happy couplehood.
  2. Instead of matching members based on "objective" assessments, some internet dating sites attempt to find people who match criteria provided by the members themselves. So, if someone is looking for an extravert, they provide a list of them. But Finkel's studies of speed dating indicate that such "self-directed" matches aren't particularly successful either.
  3. Whether online or in a speed-dating situation, people are often offered too many choices. There is a paradox of choice, in which consumers tend to be less happy with choices that they've made from a large number of options, compared to choices they've made from a small number of options. And, apparently, what goes for breakfast cereals in the grocery store goes for dates and mates.

Brangelina.JPG (69737 bytes)Personal characteristics matter, to some degree, but situational characteristics -- that is, features of the situation in which we encounter the target person -- are also powerful determinants of attraction, and can lead us to like a person even if he or she is not particularly attractive, competent, or similar to ourselves.

One of these environmental factors is sheer proximity -- we tend to like people who are physically near us. This is true within towns: Sociologists who explored the development of friendships within a suburban tract, in which everyone moved in at approximately the same time, and everyone was of roughly the same socioeconomic status, found that most people who lived near each other were friends, and that few people who lived far apart were friends. The same holds true within a living unit such as a dormitory, or even within a classroom. Despite the occasional horror story, roommates generally get along remarkably well, considering how they came together. And students often strike up friendships with those who just happen to sit next to them in class. "Workplace romances" occur because the parties have frequent contact with each other.

Propinquity.JPG (84472 bytes)Here's an example, from a study by Segal (1974) on friendship patterns among state police recruits. In classes at the police academy, the recruits were seated in alphabetical order. Thus, all the recruits with names beginning with letters early in the alphabet sat near each other, as did all the recruits with names beginning with late letters. Later, when asked to nominate their friends, there was a high correlation between the last name of the recruit and the last names of his/her friends.

              (92795 bytes)What is really important, it turns out, is not so much sheer physical proximity, but rather functional distance -- how much effort is required to make and maintain contact, and the availability of the other person as a friend. A neighbor with whom you share a yard is more likely to be a friend that one whose yard is separated from yours by a fence. If your dormitory room is near the bathroom, the elevator, or a stairwell, you are likely to have more friends -- just because of the number of people passing by your door, and the frequency with which they do it. Thus, the effects of proximity are partly mediated by familiarity -- we like those who are familiar to us more than we like strangers.

The Mere Exposure Effect

The role of familiarity in evaluation is dramatically illustrated by a series of classic studies performed by Robert Zajonc on what he called the mere exposure effect. In one of these experiments, subjects recruited for an experiment on language learning were taught to pronounce unfamiliar, meaningless material presented to them as either "Turkish words" of "Chinese ideographs" (the words weren't really Turkish, and the ideographs weren't really Chinese, but that's what they looked like). During the experiment, individual items were presented to the subjects from 1 to 25 times in a sort of verbal-learning experiment.

How to Pronounce "Zajonc"

It's "Zy-unce". Prof. William Ickes, who was a graduate student of Prof. Zajonc, once made up this limerick to help people remember:

Repeated exposure," quoth Zajonc,makes liking increase,

In defajonc of maxims that said that contempt comes instead.

Thus witness the triumph of scajonc.

With respect to the last line, according to an apocryphal story, Zajonc also planned to name his daughter Zocial.

zajoncmere1.gif (3969 bytes) Afterwards, the subjects were (falsely) told that the items were adjectives, and were asked to guess whether each item had a "good" or "bad" meaning on a 0-6 rating scale. Compared to control items that had never been presented at all, items zajoncmere2.jpg (22310 bytes) that had been studied previously were more favorably rated -- and the favorable evaluations increased with the number of repetitions. Almost every item employed in the study was rated more favorably if it had been presented many times, as opposed to just a few times (or, in the control condition, not presented at all).

The mere exposure effect occurs when repeated exposure increases judgments of attractiveness, likability, or any other favorable evaluation, even in the absence of any substantive contact with the objects being rated. Because likability is an expression of preference, and preference is an attitude, the mere exposure is a pure situational effect on the person. Apparently, people prefer whatever they encounter frequently in their environment.

              (81934 bytes)A very clever demonstration of the mere exposure effect was performed by Mita and colleagues in an experiment in which subjects were asked to make judgments of faces. Subjects were presented with pairs of near-identical photographs of women, in which one photograph was the mirror-reversal of the other, and asked to indicate which one they preferred. In these photographs of the American actress Marilyn Monroe, Picture A is the actual photo; Picture B has been mirror-reversed.

Some of the subjects were the same women whose faces appeared in the photographs, while other subjects were their friends (in one study) or romantic partners (in another). The idea was that people should prefer the photograph that offers the view with which they are more familiar:

  • Acquaintances should prefer the original photograph, because that is the way the person appears to them. Joe DiMaggio and Arthur Miller, Marilyn's husbands, should have preferred this picture.
  • But the subjects of the photographs themselves should prefer the mirror-reversed images, because that is the way they appear to themselves in a mirror. Marilyn herself should have preferred this picture.

mita2.gif (6588
              bytes)In fact, that was what Mita et al. found: people prefer photographs that represent the view they have encountered most frequently.

As with the effects of the situation on behavior, social psychology is full of studies demonstrating the effects of the situation on the person, and especially of effects on the person's attitudes (traditional social psychology, after all, is virtually defined as the study of social influence. These studies are only samples of this very large body of research.

Proximity, Liking, Contact, and Prejudice

The role of proximity and familiarity in creating and maintaining preferences has led many social psychologists to suggest that ethnic and racial prejudice, and the discrimination that results from it, could be overcome by arranging situations in which members of different ethnic and racial groups have substantive contact with each other.

The principle -- known as the "contact hypothesis" is illustrated by an anecdote from the editor, writer, and publisher Sol Stein, a white Jewish man who co-founded the now-defunct publishing house Stein and Day, and who edited Notes of a Native Son (1955), the first book by James Baldwin, by his high-school classmate (at New York's DeWitt Clinton High, in whose halls at the time also roamed the future Broadway composer Richard Rodgers, the actor Burt Lancaster, the poet Countee Cullen, the playwrights Neill Simon and Paddy Chayefsky, and photographer Richard Avedon) and lifelong friend. In an interview concerning his memoir of their friendship, Native Sons (2004), Stein related the following story from his time as an infantry commander in occupied Germany after World War II (as recounted "A Literary Friendship in Black and White" by Joseph Berger, New York Times, 09/13/04):

His own "conspiratorial" friendship with Baldwin, Mr. Stein said, demonstrates the need for blacks and whites to encounter one another, whatever their attitudes going in. He recalled that when he was commanding an infantry unit in Germany, a white soldier from the South had complained, "They put a negra in the bed next to mine."

"I told him: 'It's in alphabetical order. Goodbye.'"

But the two soldiers quickly grew so close that they were hauled up on charges of cheating together on an exam.

"I think about what is there that's applicable to other people," Mr. Stein said of his friendship with Baldwin, "and the key word is proximity."

The Influence of the Person on the Environment

James Baldwin, the African-American author of such classics as go Tell It On the Mountain (1953), Notes of a Native Son (1955), Nobody Knows My Name (1961), the Fire Next Time (1963), and If Beale Street Could Talk"(1974), was once encouraged by a friend to settle down. He replied: "the place in which I'll fit will not exist until I make it" (quote by Claudia Roth Pierpont in "another Country", New Yorker, 02/09-16/2009). The idea that we make our own environment is the heart of the Doctrine of Interactionism.

              (68089 bytes)As the foregoing experiments show, people are affected by their environments in a variety of ways. But, according to the doctrine of interactionism, people also affect the environments in which their behavior takes place.

How can this happen?

David Buss (1987) has identified three ways in which people affect their own environments:

  • Evocation: The mere presence of a person in an environment alters that environment, independent of his or her traits, attitudes, or behaviors.
  • Selection: People deliberately choose to enter one environment as opposed to another, perhaps out of a desire to match their environments with their individual personalities.
  • Manipulation: People engage in overt behavioral activities that alter the objective environment -- that is, the environment as it is publicly experienced by everyone in it.

In addition to these three modes, a fourth means of affecting one's environment can also be identified (Cantor & Kihlstrom, 1987):

  • Transformation: People engage in covert mental activities that alter their mental representations of their subjective environment -- that is, the environment as they privately experience it.

Now, let us examine each of these four modes in turn.

Interactionism Redux

Because this material is so important, it gets its own section.


By evocation, we mean that the person's mere presence in the environment alters that environment, independent of his or her actual behavior, and in fact even in the absence of any behavior at all. Evocation works, chiefly, by eliciting behavior from others which changes the environment for everyone. It can occur completely unintentionally, and even may be completely unconscious on everyone's part. Thus, the person's effect on the environment is not mediated by his or her behavior, at least in the sense of a deliberate voluntary act. Rather, the effect is produced by the person's mere presence -- most likely, by his or her physical appearance.

Gender-Role Socialization

Evocation plays a major role in gender-role socialization (described in more detail in the Lecture Supplement on Psychological Development). Parents and others treat little boys and girls differently, in order to bring up boys who are appropriately masculine, and girls who are appropriately feminine (as these concepts are defined by the larger culture).

095GenderProg.jpg (83780 bytes)According to an analysis provided by John Money and Anke Ehrhardt (1972), the "program" for gender dimorphism (i.e., sex differences) begins with what they call the phyletic imprimatur, or those (largely biological) aspects of gender dimorphism that are shared by all members of the species. To make a long story short, genetic and hormonal processes during fetal development leave the developing fetus, and newborn child with a set of external genitalia, and an internal reproductive system, that are more or less recognizably male or female. (There may also be some gender differentiation of brain structures, possibly forming the biological basis for gender differences in mental and behavioral processes.)

  • In boys, a penis and a scrotum, testes, vas differens, and seminal vesicles.
  • In girls, a clitoris and vagina, uterus, and fallopian tubes.

At birth the program for gender dimorphism is passed from the genes and prenatal hormones to the environment. The social imprimatur refers to the differential socialization that boys and girls experience, in line with the concepts of masculinity and femininity that prevail in the culture into which they are born. (The program passes back to the hormones at the time of puberty, returns to the autobiographical history, and returns again, for women at least, at menopause.) As Margaret Mead showed, the details of the social imprimatur vary from one culture to another, but in the present context the important point is that every culture distinguishes between the sexes. The distinctions that the culture makes constitute the social imprimatur.

The word imprimatur is derived from the Latin, meaning "it is permitted". In the past, the Roman Catholic Church published a list of books that had been given the imprimatur, meaning that Catholics were permitted to read them. In Money and Ehrhardt's work, the phyletic imprimatur refers to those aspects of gender dimorphism that are permitted by biology, and the social imprimatur refers to those aspects of gender dimorphism that are permitted by culture.

The point of the social imprimatur is that the child's social and cultural environment is structured around masculinity or femininity, literally at birth, in response to the announcement that the child is a boy or girl. The newborn is wrapped in a pink (for girls) or blue (for boys) receiving blanket (before the 20th century, the gender-dimorphism was reversed, with pink for boys and blue for girls), and from that point on the child's parents and others in the social environment raise the child in accordance with prevailing cultural concepts of masculinity and femininity, by

  • demanding and reinforcing gender-typed behaviors;
  • communicating gender-typed expectations; and
  • modeling gender-typed behaviors

But the announcement of the child's sex is based on the perception and categorization of the child's external genitalia. In this sense, the external genitalia structure the child's social environment, by evoking behavior from others that would be different, if the genitalia were different.

The process of evocation is nicely illustrated by what are known as the "Baby X" studies. In these experiments, adults are asked to assist in a study of infants' responses to strangers. They are asked simply to interact with the infant, and are given some toys which they may use for the purpose.

              (44072 bytes)In BabyX75_2.JPG (41951 bytes)the first study (Seavey et al., 1975), the infant was a 3-month-old girl, but the adults (actually university graduate students) were told that the child was either a girl, or a boy, or were given no information. Adults who thought they were interacting with a girl were much more likely to choose a doll as a prop for the interaction, as opposed to those who thought they were interacting with a boy. Moreover, men engaged in more physical contact with the child when it was identified as a boy, while women engaged in more physical contact with the same child when it was identified as a girl.

              (44400 bytes)A follow-up study (Sidorowicz & Lunney, 1980) employed both male and female infants (in order to control for the possibility that the adults' behavior in the first study had been shaped by subtly "feminine" behavior on the part of the infant), and replicated the doll-choice finding (the investigators did not assess physical contact).

In both cases it was the (perceived) gender of the child that evoked differential behavior from the adults who interacted with her.

Stereotyping, Prejudice, and Discrimination

Evocation is also exemplified by stereotyping and prejudice based on race and ethnicity, as well as gender. These are social categories marked by physical attributes -- males look different from females (even with their clothes on -- well, mostly), and whites, African-Americans, and Asian-Americans generally look different from each other as well.Muslims may wear headscarves or caps that make them look different. When a woman is present in a mostly male environment (or vice versa), or a black person is present in a mostly white environment (or vice-versa), it is not necessary for that person to do anything in order to change the situation (mostly by eliciting discriminatory, hostile, or other unpleasant) behavior. His or her mere presence is sufficient to evoke the stereotype and prejudice.

An interesting example of this is portrayed in Out of Africa, a film starring Meryl Streep based on the memoir by Isaak Dinesen (the pen name of Karen Blixen). At one point, Blixen enters a men's club room in Nairobi (or somewhere like that). A moment before, the room had been full of boisterous (at least for the British) hale-fellow-well-met man-talk. But the moment she walks into the room, the entire place falls silent. Her mere presence has changed the character of the situation, by virtue of the behavior it evoked in others.

Many acts of racial prejudice and discrimination have this kind of character to them -- when someone "doesn't know his place", and finds himself "where he doesn't belong". However innocent, by his mere presence he has changed the situation.

Link to an interview with Claude Steele, Vice-Chancellor and provost of UC Berkeley, about stereotype threat as a barrier to achievement.

Link to an interview with Mahzarin Banaji.

Ingroups and Outgroups

Setting aside gender and racial prejudice, this kind of evocation occurs whenever a member of an outgroup enters a situation populated by an ingroup majority -- Yankees and Red Sox fans, 49s and Raiders, Berkeley and Stanford, Sigma Chis and Sigma Nus. The ingroup member(s) will almost inevitably evoke behavior from the outgroup members that change the situation for both of them. Members of the ingroup will behave differently than they would otherwise, if the outgroup member were not present. Gender or racial stereotyping doesn't require any behavior on the part of the person.

The critical feature of evocation is that the person need not do anything to change the situation. His or her mere presence and appearance is sufficient to evoke behavior from others that changes the situation for everyone.


The notion of selection entails the idea that the match between persons and their environments is not random. Rather, individuals choose environments that are congruent with their personalities -- with their traits, attitudes, beliefs, moods, goals, and values. In this way, people place themselves in environments that support and promote their own behavioral tendencies. The selection of the situation is not a trivial effect: each choice pre-empts alternatives, and thus provides little support for the acquisition of new traits, attitudes, and the like. This can result in a kind of positive feedback loop between the person and the environment -- with the person choosing compatible environments, and the environment eliciting behavior that makes it unlikely that the person will chose another environment.

              (70862 bytes)Every act of selection chooses for something, and against something else -- a point nicely illustrated in the film Sliding Doors starring Gwyneth Paltrow.

Consider, for example, a child in the delay-of-gratification situation. If he is asked whether he would like to wait in the presence or absence of the reward, or in the presence of the preferred or non-preferred reward, his choice will affect his behavior by determining whether he will wait in a situation in which long delays are easy or hard to achieve. If he chooses to wait in the presence of both rewards, he will be unable to wait for very long. If he chooses to wait in the absence of both rewards, he may very well outlast the experimenter!

The Person-Environment Fit

A study of environmental selection by college undergraduates illustrates the correlation between persons and environments (Emmons, Diener, & Larsen, 1986). These investigators first assessed a number of personality traits, and then asked the subjects where they spend their time -- distinguishing between freely chosen and imposed situations (e.g., working out at the gym vs. mandatory physical education classes). Finally, they assessed the subjects' self-reported mood in each situation. The general finding was of personality-environment congruence.

  • Subjects' freely chosen environment were significantly correlated with their personalities. For example:
    • Subjects who were high in extraversion spent more time in social than solo situations, and in social than nonsocial recreation;
    • Subjects who were high in achievement motivation spent more time in work than recreational situations, and also more time in social than solo recreation;
    • Subjects who were high in the motive to affiliate spent more time in social situations, and more time in social but not solo recreation.
  • Subjects' mood in these situations was significantly correlated with their personality as well. For example:
    • Subjects who were high in extraversion felt better in social than in solo situations.
.A variant on the person-enviroment fit is the person-culture match.  In a survey of more than 7,000 subjects in 28 societies, individuals whose personalities (extraversion) matched the local cultural norms showed higher levels of self-esteem and subjective well-being than those who were "mismatched".

Personality and Mate Preference

Another example of selection at work is personality and mate preference. Recall from the discussion of interpersonal attraction that similarity attracts: people tend to like others who are like them in personality and attitudes. Not only that, people actively select among potential mates, based on similarity. Walster et al. (1966) proposed a matching hypothesis proposes that, when choosing a partner from the "dating market", assess their own level of social desirability, and then select partners whose (perceived) self-esteem matches their own. This narrowing of the field to those who already match accounts, in part, for the fact that people are more attracted to those who are similar to themselves in various personality and social characteristics.

Actually, most of the research on the matching hypothesis finds that people tend to prefer highly desirable partners -- regardless of their own self-esteem. But what people prefer is not the same thing as what they can actually get -- if for no other reason than that targets are not passive in this process, and are actively engaged in selection of their own. The process of self-other matching is more dynamic than a simple matter of preference. When surveying the dating market, people quite quickly determine that some potential mates are "out of my league", and focus their attention elsewhere.

Surveying the literature on interpersonal attraction, Taylor et al. (2011) found that matching does indeed occur on self-worth, but also extends to physical attractiveness and popularity.

MateSelection.JPG (68964 bytes)As a case in point, Buss (1987) measured subjects' in terms of the "Big Five" personality traits, and then asked the subjects to indicate, via an adjective checklist, the psychological qualities that they would prefer in a mate. There was a high correlation: extraverts preferred mates who were extraverted, and neurotics even preferred mates who were neurotic! Even more than our friends, our mates constitute an important part of our social environment. By selecting mates with personalities that are similar to ours, we help insure that our environments will be compatible with our personalities.

This process can be modeled by a parlor game demonstrated (at a party, of course) by Harold Kelley (he of attribution-theory fame) of UCLA. The host writes numbers (e.g., from 1-10) on sticky labels, and affixes these labels to the foreheads of his guests, in such a manner that they cannot see their own number. He then instructs his guests each to find a partner, with a promise that the couple with the highest total points will win a prize (in California, that usually means a nice bottle of wine). As the guests mill about, you can see them spurning each other, and then pairing off. The result, almost inevitably, is a near-perfect correlation between numbers, as the 10s match with 9s and 10s, and the 1s are stuck with 1s and 2.

Selection of the environment by the person is ubiquitous. Every personal choice changes the environment in which the person resides, and in which his or her behavior takes place. Each of us plays a number of different social roles (professor, friend, husband, brother, son), and each of these imposes different role-demands on the individual. In a sense, each of these roles is a different environment, and choosing one forecloses on others -- at least temporarily. Catholic priests can't be husbands (at least, not yet). Each choice the person makes opens up some opportunities, and imposes some demands, but it also closes off some others. In any event, every act of selection literally changes the person's environment, because it puts the person in one behavior setting rather than another.


But what if you find yourself in a situation you didn't choose, or for some other reason the environment is wrong? People have the capacity to alter their environments through their behavior -- activity that changes the actual character of the environment. We all know someone who can turn any situation into a party -- and most of us also know someone who can turn any party into a funeral.

In manipulation, the person engages in overt, publicly observable behavior that alters the environment -- and alters it in objective terms, for everyone who is also in that environment.

It almost seems superfluous to cite examples of behavioral manipulation of the environment. As noted earlier, every act of instrumental or operant behavior changes the organism's environment in some way: the rat who presses the key in a Skinner box changes the environment from one that doesn't contain food to one that does. In fact, we've already discussed some of this literature, in terms of the influence of behavior on the environment.

The Prisoner's Dilemma

This is a famous problem in game theory, a paradigm often used to study interpersonal behavior (Luce & Raiffa, after a suggestion by A.W. Tucker). Consider two individuals who have been arrested on suspicion of armed robbery. The district attorney does not have enough evidence to convict, so he makes each of them the following offer:

  • If neither suspect confesses, both will get light sentences, because the DA will be able to convict them on lesser charges.
  • If both suspects confess, they will both get reduced sentences, because they will save the DA the costs of a trial (and the risk of acquittal).
  • If one suspect confesses and turns state's evidence, he will do a very short sentence while the DA will throw the book at the one who held out.

Under these circumstances, there really isn't a logical response -- which is why it's called a dilemma:

  • Cooperation, in which both suspects stay silent, maximizes their joint outcome.
    • But cooperation entails trust, that the other suspect will not defect and confess.
  • Competition, in which one suspect rats on the other, maximizes the defector's personal outcome at the expense of the other suspect.
    • But if both of them confess, they will get longer sentences than if they stay silent.

Under these circumstances, there is often an initial tendency to compete. However, as the game is played over multiple trials, gradually the players will evolve into a cooperative strategy. This is often a response to a "tit for tat" strategy in which the victim of competition on one trial will retaliate against the defector on the next trial. Eventually, both players figure out that it's in both their best interest to remain silent.

              Problem of Trait-NamesKelley & Stahelski (1970) used the prisoner's dilemma to look at the relationship between cooperation and competition as the game is played out over time. Their subjects played a standard prisoner's dilemma game, like the one illustrated above, for 3 blocks of 10 trials each. At the outset of the game, the investigators assessed the subjects' initial goals -- to cooperate or to compete -- and then looked at what happens with various combinations of cooperators and competitors are pitted against each other. The results were striking. Examining just the behavior of one player, arbitrarily designated as the actor:

  • Cooperative actors paired with cooperative partners consistently made cooperative choices, and this tendency actually increased across the 3 blocks of 10 trials.
  • Cooperative actors paired with competitive partners actually made fewer cooperative moves: it was as if the competitive behavior of the partners actually made the actors more competitive.
  • Competitive actors paired with cooperative partners made very few cooperative responses at the outset, and actually made even fewer as the game went on. They seemed to take advantage of their partner's cooperative spirits -- or, to put it another way, played their partners for saps.
  • Competitive actors paired with competitive partners seemed to get involved in a "vicious cycle" of competition.

Kelley & Stahelski referred to these effects as behavioral assimilation of cooperators to competitors. In other words, competitors, by their competitive behavior, created an environment that elicited competitive behavior from cooperators as well.

The Evolution of Cooperation

Actually, there is a logical approach to the Prisoner's dilemma game, but it's one that emerges only in iterated versions of the game, in which opponents play each other repeatedly. It's not one that can be used in the classical version of the game, in which players are presented with only one instance of the dilemma.

In the iterated version, there are a number of different strategies that can arise, including:

  • Nice, in which the player always cooperates.
  • Nasty, in which the player always competes.
  • Tit-for-Tat, in which the player cooperates on the first trial and then on subsequent trials behaves exactly as his/her partner did on the immediately preceding trial.
  • Mild Tit-for-Tat, in which the player cooperates after a short series of deceptions.
  • Naive Probing, in which the player competes after a short series of cooperative moves.

A Prisoner's Dilemma "tournament" held by Robert Axelrod (in his book, The Evolution of Cooperation) showed that the Tit-for-Tat strategy is most appropriate in the iterated game. Because each competitive choice is answered by retaliate competition, and each cooperative choice is answered by cooperative reconciliation, cooperation eventually (actually, fairly quickly) evolves. If the game is carried out even further, Nowak and Sigmund found that "Tit for Tat" is replaced by "Generous Tit for Tat", in which one partner replies to a competitive play with a cooperative one.

Strategies for Social Manipulation

One way of manipulating the environment is to manipulate the people in that environment. David Buss and his colleagues (1987) interviewed people about the strategies they use to get people do to certain things (or stop doing them). He classified their responses into six broad strategies:

  • Charm: I'll complement him/her so s/he'll do it"
  • Silent Treatment: "I won't respond to him/her until s/he does/stops it"
  • Coercion: "I'll demand that s/he do/stop it"
  • Reason: "I'll give him/her reasons to do/stop it"
  • Regression: "I'll pout and sulk until s/he does/stops it"
  • Debasement: "I'll lower myself so s/he'll do/stop it"

Buss also found that subjects' preference for these strategies was related to their personality characteristics.

  • Subjects who were high on extraversion preferred charm.
  • Subjects who were high on neuroticism preferred the silent treatment or regression.
  • Subjects who were quarrelsome, or disagreeable, preferred coercion.
  • Subjects who were ambitious preferred reason.
  • Subjects who were lazy preferred debasement.

Apparently, people with different personalities have characteristically different ways of manipulating other people. When these strategies for social manipulation are successful, they change the objective, publicly observable environment, because the other people in that environment are now doing something they weren't doing before.

Delay of Gratification

Examples of behavioral manipulation of the environment also crop up in the literature on delay of gratification. Rather than looking to children's personality traits to understand delay of gratification, Mischel and his colleagues (1989) have asked the question: "What do 'high-delay' children do?"

Consider, for example, children in an experiment in which children must wait in the presence of a promised reward. We know from a previous study that, on average, children do not wait very long in such a situation. But some children do: How do they do it? Traditional personality psychologists would attribute these differences in delay of gratification to differences in personality traits, such as ego-control and ego-resiliency. But while traditional personality psychologists focus their analyses on the traits that children have, interactionists would focus on what the children do.

We know from studies of situational influence that features of the environment affect delay of gratification -- such as whether the rewards are physically present. But even when the rewards are physically present, some children show a high capacity for delay of gratification. Mischel and his colleagues (Mischel et al., 1989) observed high-delay children engaging in a number of interesting behaviors:

  • They avoid deliberately looking at the rewards: They may cover their eyes with their hands, or rest their heads on their arms.
  • They generate their own diversions: talking or singing quietly to themselves, playing games with their hands or feet, or even trying to sleep.

In other words, these children were behaving in such a way as to put the rewards out of sight -- to change the environment into one in which the reward, though physically present, is not really there.

The 'Big Five'
              Personality TraitsA formal experiment by Mischel, Ebbesen, and Zeiss (1972, Exp. 1) makes the same point. Children were given the choice of marshmallows or pretzels, and then asked to wait for the experimenter to return before he could receive the preferred reward. The child could also signal for the experimenter to return, in which case he would get the non-preferred reward. Children who were given no distraction could not wait very long. But children who were told to play with a "Slinky" toy were able to wait much longer. Out of sight, out of mind. The rewards can be physically absent, as is the case in situational influence. Or the subject can do something to make them absent, in which case the person's behavior is altering the environment.

The Marshmallow  Test

Mischel's "Marshmallow Test" -- as in, "Would you like to have one marshmallow now or two marshmallows later?" -- has entered popular culture as a metaphor for people's ability to control themselves.  Although many commentators misintererpreted self-control as something people have, in varying degrees, Mischel's own research makes clear that self-control is something people do, and can learn.  In these lectures, I'll return to the MT several times to illustrate various aspects of the person-situation interaction.  Mischel's own research on self-regulation and delay of gratification is summarized in his book, aptly titled the Marshmallow Test: WhySelf-Control is the Engine of Success (2014).  The cartoon, by Paul Noth, appeared in the New Yorker, 01/23/2016, around the time of the inauguration of Donald J. Trump as the 45th President of the United States.


Even when we have no choice over our environment, and are unable to manipulate it through our overt behavior, we can still change the environment through covert, cognitive operations -- by changing its private, subjective meaning. The environment is changed through cognitive activity, not overt behavior.

Manipulation vs. Transformation

By manipulation, we refer to overt behavioral activity that changes the objective environment.

By transformation, we refer to covert mental activity that changes the mental representation of the environment -- how the individual subjectively perceives and categorizes the situation.

Manipulation entails public behavior that changes the environment for everyone. Transformation entails private thoughts that change the environment only for the person who thinks them.

The "n Effect"

A nice comparison of the objective situation with the subjective, mental representation of the situation is the "n Effect" of group size on test performance. Garcia and Tor (2009) examined SAT performance records for 2005 as a function of the average number of test-takers per testing venue in each state. As the average number of test-takers in a venue increased, SAT scores decreased, and this was true for both the verbal and math portions of the test. They followed up with a formal experiment in which they gave college-student subjects a general-knowledge test. All the students took the test alone, but half were told that they were competing against 10 other students for a prize of $5, while the others were told they were competing against 100 students. It was an easy test, so most of the subjects got most of the items right. But subjects who believed that they were competing against 10 people finished the test significantly faster than those who thought they were competing against 100 people.

Garcia and Tor suggest that, when faced with a large number of competitors, people -- especially those who tend to compare themselves to others -- figure that it is not worth trying too hard. Accordingly, they suggest, the number of people in competitive situations should be controlled, so that everyone competes on the same level playing field. Maybe that's what's happening, and maybe that's the implication. But for present purposes the point of the experiment is that it's the subjects' beliefs about the situation -- that they have many rivals, or only just a few -- that affects performance.

Delay of Gratification

Out of sight, out of mind. We have already seen that if the reward is physically absent from the situation, children can delay longer. And we have also seen that even if the reward is physically present, children can delay longer if they behave as if it is absent. But this behavior does not have to entail overt action. It can also entail covert, purely mental, activity.

              Structure of AttitudesFor example, consider the experiment by Mischel, Ebbesen, & Zeiss (1972), discussed earlier, in which children were able to delay for a long time by distracting themselves with a "Slinky" toy. It turns out that the distraction doesn't have to involve overt behavior, like playing with a toy. In another condition of the "Slinky" experiment, children were simply instructed to spend their time thinking of "anything that's fun to think of". These children were able to delay gratification even longer than those who played with the "Slinky" toy. By putting the reward out of sight, or by directing their attention elsewhere (e.g., to a game, a song, or a Slinky toy, or just "fun" thoughts), the have effectively their environment from one in which they cannot wait very long to one in which they can.

An experiment on the role of ideation in delay of gratification by Mischel and Baker (1975) further illustrates the power of cognitive transformation. In this experiment the children were given the usual choice of marshmallows vs. pretzels, and during the waiting period the preferred reward remained in sight.

  • In one condition of the experiment, the children were instructed to spend the waiting period thinking about the consummatory aspects of the reward -- how it would feel and taste when they finally got to eat it.
  • In another condition of the experiment, the children were instructed to transform the reward into something they could not eat.
  • In each condition, half the children were instructed to think about their preferred reward, the others to think about the non-preferred reward.

Ideational Instructions



Look at the marshmallows. They are sweet and chewy and soft. When you look at marshmallows, think about how sweet they are when you eat them.... When you look at marshmallows, think about how soft and sticky they are in your mouth when you eat them.... When you look at marshmallows, think about how white and puffy they are. Clouds are white and puffy too -- when you look at marshmallows, think about clouds.... The moon is round and white. When you look at marshmallows, think about the moon....
Look at the pretzels; they are crunchy and salty. When you look at pretzels, think about how crunchy they are. When you look at pretzels, think about how salty they taste when you lick them or chew them... When you look at pretzels you can think about how long and brown they are. A log is long and brown. When you look at pretzels, think about logs and tree trunks. Or you can think about how round and tall they are. A pole is round and tall....

Hierarchical Structure of AttitudesThe results of the experiment were very striking. Children who focused their thoughts on the consummatory aspects of their preferred reward were not able to delay gratification for very long. Those who transformed their preferred reward into something that did not taste sticky-sweet, or crunchy-salty, were able to delay for a long time. What caused their delay (or lack thereof) was not their personality traits, or the environment, or even their own behavior. The important factor was how they thought about the rewards.

Thus the notion of delay of gratification as a trait that people have, proves to be an oversimplification. Ego-control is not just a personality characteristic, it is also a product of strategic activity. Delay of gratification is accomplished through a combination of selection, manipulation, and transformation, all oriented around the general principle of "out of sight, out of mind".

  • Selection. The child may choose to wait in the absence of reward. Children who select an environment that is free of temptation do so will be able to wait longer than those who do not, because the environment makes delay of gratification easier and more likely.
  • Manipulation. But what if such an environment is not possible -- perhaps because an environment in which the reward is present has been imposed on the child? If the rewards are present, the child can manipulate the environment through behavior -- avoiding looking at the rewards, distracting himself from them -- that changes the character of the environment. By putting the rewards out of sight, these strategies make delay of gratification easier, and more likely.
  • Transformation. But what if such behaviors are not possible -- if, just to be extreme, we tie the child upright in the chair facing the reward, tape his eyes open and his mouth shut? Such a child can transform the environment through purely mental operations, by not thinking about the rewards (admittedly difficult when he's looking right at them), or by thinking about the rewards in a different way. By changing the meaning of the rewards, he changes the meaning of the situation, and behavior changes as well.

Through the process of social learning, children acquire both knowledge of effective behavioral and cognitive strategies for delay of gratification, and the ability to deploy these strategies effectively.

In the cognitive view of social interaction, an individual's behavior in some situation is determined by his or her mental representation of that situation, in the context of that person's history of social learning and goals for the future -- in other words, by the meaning of that situation. This meaning is determined by processes of perception and categorization, memories of past experiences in similar situations, and other cognitive factors, including the self-concept, the perception of other people, and the interpretation of one's own and others' behavior. Cognitive transformations alter the meaning of the environmental situation for the actor, and thus the actor's behavioral response to that situation. We cannot understand a person's behavior in a situation unless we understand the meaning of that situation for that person.

Consider, for example, the behavior of Rosa Parks, Martin Luther King, Jr., and other heroes of the American civil rights movement. King, Medgar Evers, and other black civil rights workers, as well as whites like James Chaney, Andrew Goodman, Michael Schwerner, and others associated with such movements as the Mississippi Summer Project, actively confronted civil authorities in the South, putting themselves in situations, and engaging in behaviors (such as sit-ins, protest marches, and voter-registration drives) that increased the likelihood that they would be jailed, beaten, or killed (in fact, Evers was killed in 1967, King in 1968; the 1967 deaths of Cheyney, Goodman, and Schwerner formed the basis of the film Mississippi Burning). Most of us would avoid these situations and behaviors. The civil rights workers behaved differently not just because they were braver than the rest of us (which, admittedly, they probably were), but because the meanings of the situations were different for them, and they believed that their actions were necessary to obtain civil rights for American blacks. For King, being jailed by Bull Connor was the opportunity to write his famous "Letter from a Birmingham Jail", rallying people all across the country to his cause.

Gandhi's example and teaching were a basic inspiration for the United States civil rights movement led by Martin Luther King Jr. and the Southern Christian Leadership Conference. From A. Philip Randolph's threatened March on Washington in July 1940 to protest exclusionary hiring practices in defense industries to King's successful actions of the 1960s, carefully planned and targeted, nonviolent, civil resistance was the essence of the movement's operations.

Its strategy included inducing opponents to react brutally, thereby inviting sympathetic support from the press and public and thus encouraging the federal government to intervene on the side of law and order. King and the SCLC were masters of this technique. They selected Birmingham, Alabama, for their 1963 campaign, because the commissioner of public safety, "Bull" Connor, was a dependably violent racist hothead who could be relied upon to use dogs, cattle prods, and water cannons against peaceful demonstrators. Connor's brutalities invited TV coverage that made him a national villain and sowed the seeds for President Lyndon Johnson's 1964 Civil Rights Act.

In an ironic tribute, President Kennedy told King apropos of Connor, "in his own way, he has done a good deal for civil rights legislation this year".

Brian Urquhart, "Revolution Without Violence?"

New York Review of Books, 03/10/2011

The example of the American civil rights workers reinforces the implications of the doctrine of mentalism: mental states cause actions. People behave in accordance with their beliefs. If you want to understand someone's actions, you don't look at their genetic heritage or their physiology. You look inside their minds, to determine what they know and believe, feel, and want. That will explain what they do.

Selection, Manipulation, and Transformation in Gender Dimorphism

Early in gender-role socialization, parents and others select the environment for the child. That is, partly, what evocation is all about: based on his or her physical appearance, the parents categorize the child as a boy or girl, and treat her accordingly.

In addition, at about 2 years of age the child notices his or her own genitals and begins to identify him- or herself as a boy or a girl, and attaches a positive affective valence to his or her own gender. And at that point, the child begins to play an active role in his or her own gender-role socialization. By identifying him- or herself as the same as some, but different from other, people, the child effectively creates an (same-sex) ingroup and (opposite-sex) outgroup. By favoring others of his or her own kind (e.g., brothers over sisters or vice-versa), he or she effectively changes his or her environment from one that was relatively balanced (i.e., having roughly equal numbers of males and females) to one that is biased one way or the other.

It's just a matter of fact that, on average, little boys are more active -- maybe even more aggressive -- than are little girls. Although gender-role socialization plays an important role in this sex difference, it is also possible that sex differences in activity level, if not aggressiveness, are also partly based in differences in brain structure that are shaped by prenatal sex hormones. That is to say, there may be some degree of "masculinization" of the brain, as well as of the genitalia.

To the extent this is true, then little boys and little girls will "naturally" behave somewhat different in any situation to which they are exposed. To the extent that they do, these behavioral differences will "naturally" shape the environments in which they live -- especially when little boys and girls get together. On average, the environment for boys will be characterized by higher activity levels than that for girls, including higher levels of roughhousing and maybe even aggressiveness.

To the extent they exist, innate sex differences in activity level or aggressiveness will shape different environments for males and females. However unconsciously or inadvertently it may occur, this shaping may be one way in which males and females alter their own environments -- and, for that matter the environments of those they live with!

Two different aspects of gender dimorphism illustrate the role of cognitive transformations in behavior.

In evocation, the appearance of the child's external genitalia literally structure the environment, giving it a more "masculine" or "feminine" character. But of course, the child's genitalia do not have direct effects on the environment (well, they do, sometimes, but that's not what we're talking about here!). Rather, the effects of evocation are mediated by the parents' (and others')concepts concerning masculinity and femininity -- their beliefs about what constitutes proper behavior for little boys and little girls. The parents identify the child as a boy or a girl, and then go about restructuring the environment in accordance with those beliefs. Parents who define masculinity and femininity one way will structure the environment that way, while those who define masculinity and femininity another way will structure the environment that other way. The structuring begins with the child, but the child's impact on the environment is filtered through the parents' perception of the child, memories of their own childhoods, and thoughts about what is "proper" for little boys and girls.

As soon as the child identifies him- or herself as a boy or a girl, the child becomes a more active participant in the process of gender-role socialization. This identification is a cognitive act of categorization. One the child views him- or herself as a boy or a girl, he or she will start affiliating with others of the same sex, and doing the things they do.

Remember that learning is an active process, and that is as true for little boys and girls as it is for rats and pigeons. Once the child has achieved a stable gender identity (a process that starts about age 2 and is essentially complete by about age 5), he or she actively learns the behaviors that are appropriate to boy- or girlhood. Some of this learning occurs as a result of direct experience (when the parents and others reward gender-appropriate behavior and punish behavior that is gender-inappropriate), but a great deal of it is vicarious social learning by precept and by example:

  • in learning by precept, parents and others instruct children how they should behave;
  • in learning by example, children model their behavior on that of others in their environment who are similarly endowed.

Actually, during gender-role socialization children learn both gender roles, and then adopt the particular role that is appropriate to their individual identities -- or self-perceptions. If a little boy or girl should, for whatever reason, identify himself with the opposite sex, gender role behaviors that seem perfectly appropriate to him or her may seem inappropriate to other people.

Thus, by virtue of cognitive transformation as well as behavioral manipulation and active selection, not to mention passive evocation, the child is a promoter of his or her own psychological development.

The Self-Fulfilling Prophecy

              Self-Fulfilling ProphecyCognitive transformations of the environment are not entirely private, because they lead to overt, public behavior which can alter the objective situation for other people as well. This process is exemplified by what the American sociologist Robert K. Merton (1947) called the "self-fulfilling prophecy".

The Self-Fulfilling Prophecy

The self-fulfilling prophecy was originally described by a sociologist, Robert K. Merton, in 1948:

The self-fulfilling prophecy is, in the beginning, a false definition of the situation evoking a new behavior which makes the originally false conception come true.

The specious validity of the self-fulfilling prophecy creates a reign of error. For the prophet will cite the actual course of events as proof that he was right from the very beginning.

Such are the perversities of social logic.

In the self-fulfilling prophecy, the definition of the situation may be false, but it is the definition that leads to behavior that makes the initially false conception come true.

The self-fulfilling prophecy is illustrated by Rosenthal and Jacobson's study of "Pygmalion in the Classroom" (1968), an experiment involving children (grades 1-6) enrolled in a number of public elementary schools. Early in the fall, in addition to a standard IQ test, the children were administered what was presented to the teachers and (administrators) as a nonverbal IQ test that would identify "intellectual late-bloomers" -- that is, children who might not be doing well now, but who could be expected to catch up and even exceed the achievement of their age-peers. In fact, the test had nothing to do with IQ, or the prediction of late-blooming. It was a fake. Nevertheless, the children's teachers were given false feedback about the children's test scores, identifying a random 20% of pupils in each class as "late-bloomers". Later, in the spring of that same academic year, the children were retested with the standard IQ test. On average, children in all classes and grade levels showed increases -- school really does make you smarter!. But those children who had been falsely identified as intellectual late-bloomers showed greater gains than the other children.

The important point of the study is that the children weren't really "late-bloomers" -- but their teachers believed that they were, and treated them accordingly. Thus, the effect of the cognitive transformation (identifying some students as late-bloomers) was mediated by behavioral manipulations (treating students identified as late bloomers differently than the other students). These behavioral differences boiled down to four factors:

  • Socio-emotional climate: teachers were warmer and more open to the late-bloomers.
  • Contingent Feedback: teachers provided more feedback to the students about when their performance was good or bad.
  • Input: Teachers gave late-bloomers more, and more difficult assignments.
  • Output: Teachers tended to wait for late-bloomers to answer questions, instead of cutting them off, and did more to shape and encourage their responses.

In other words, teachers who believed that a child was a late bloomer behaved as if that child were already doing well -- in short, they treated late-bloomers as if they were already smart. In so doing, they created a different environment for the late bloomers than for the other children -- who after all were in the same physical classroom. And, not surprisingly, the late-bloomers responded by -- well, blooming.

The self-fulfilling prophecy begins with a cognitive transformation -- in this case, by putting children in one category rather than another. The teachers then behaved in accordance with this cognitive categorization, and the children responded by behaving in a manner consistent with how they were treated. In other words:

private belief created public reality.

Note: Rosenthal and Jacobson's study aroused immediate controversy in the teaching profession, and some criticism -- mainly statistical -- from psychologists and other social scientists (particularly Richard Snow). However, the essential point of the Pygmalion experiment -- that expectations can become self-fulfilling prophecies -- has been borne out many times since.


                Pygmalion MythRosenthal and Jacobson's study took its name from the myth of Pygmalion and Galatea, as told by the Roman poet Ovid (47 BC - 7 AD). Pygmalion, the sculptor-king of Cyprus, hated women and resolved never to marry. But he also produced a statue depicting the most beautiful woman in the world. Despite his misogynistic attitudes, he fell in love with his creation, and pretended that it was a real woman. At one point, he asked Venus, the goddess of love, to give him a maiden just like his statue. Venus, impressed by his change of heart, brought the statue to life. Pygmalion named the maiden Galatea: they married and lived happily ever after.

There are many modern variants on the Pygmalion myth, including:

  • George Bernhard Shaw's play Pygmalion (1913), in which Professor Henry Higgins transforms Eliza Doolittle, a Cockney flower girl, into a paragon of high society.
  • Pygmalion, in turn, was the basis for My Fair Lady, a famous Broadway musical by Lerner and Lowe.
  • Bedtime for Bonzo (1951), a movie starring Ronald Reagan (long before he became Governor of California and then President of the United States -- in fact, long before he became a Republican!). In the film Reagan plays a psychology professor who raises a chimpanzee like a child to prove that environment is more important than heredity. He does this not just out of academic interest, but also to overcome the fact that his father was a convicted criminal, so he can marry the Dean's daughter.

Social Intelligence and Personality

The study of Pygmalion in the Classroom reinforces a point from our earlier discussion of learning: the objective stimulus situation rarely determines behavior. With the exception of reflexes, taxes, and instincts, behavior is a response to the subjective situation, as it is perceived by the behaving organism.

Perception is an intelligent process, in which perceivers actively construct mental representations of the world around them, drawing on information extracted from the stimulus, which is then combined with knowledge retrieved from memory. Through this inferential, problem-solving process, the perceptual representation goes beyond the information given by the stimulus.

Because behavior is based on perception, it too reflects intelligent thought. In contrast with reflexive or instinctual behavior, intelligent action shows a number of characteristic features:

  • it is discriminating, not indiscriminate;
  • it is flexible, not rigidly stereotyped; and
  • it is optional, rather than obligatory.

Social behavior, whether mundane (asking a drugstore clerk for toothpaste) or monumental (asking someone to marry you) is also intelligent. Every social interaction represents a problem to be solved -- or, perhaps, a series of problems to be solved in turn.

  • Some social situations are well-defined problems, the social equivalent of simple arithmetic, and may entail highly scripted, mindless interactions.
  • Other social situations are ill-defined problems, and require the thoughtful application of social intelligence as the actor constructs a mental representation of the problem, and achieves an appropriate solution. Many social interactions are of this type.

The person's behavior in ill-defined social interactions may reveal his or personality -- by the manner in which problems are represented, and the means by which their solutions are approached. In other words, the problems presented by social interaction are solved through the individual's application of his or her fund of social intelligence (Cantor & Kihlstrom, 1987, 1989; Kihlstrom & Cantor, 1989), by which we mean the cognitive structures involved in social interaction:

  • the person's fund of declarative social knowledge, by which he categorizes the objects and events encountered in the social world; and
  • the person's repertoire of procedural social knowledge, by which he acts on these categorizations.

The concept of social intelligence has implications for personality traits. These traits, to the extent that they exist at all, are not merely passive attributes of the individual. They also reflect situational demands and personal choices. Consider, for example, a trait like independence:

  • We can think of it as a typical personality trait -- a more-or less permanent attribute of the person, not unlike skin pigmentation or eye color or hair texture, determined by some combination of genetic endowment and childhood experiences (as genotype interacts with the environment to yield a phenotype). Once established, this trait is expressed in behavior that is more or less stable across both short and long intervals of time, and more or less consistent across a wide variety of situations.
  • But independence is also a choice. The individual decides to be independent or dependent, and behaves accordingly, selecting, manipulating, and transforming situations in line with this choice.
  • On the other hand, independence is also an agreement (I thank Jacquelyn Goodnow for this insight during a conversation we had in the summer of 1985, when I was visiting Macquarie University, where she was the head of the School of Psychology). In order for a person to be independent or dependent, the environment must support that decision. Other people have to allow independence, and not impose, or reinforce, dependency.

Personality is not best construed in terms of traits. It is not established once and for all time. Like all other aspects of human intelligence, it is continuously open to change, as people define and redefine themselves, and are defined and redefined by others. Personality emerges out of the interaction between the person and the situation -- an interaction that is cognitively mediated.

This page last modified 06/10/2017.