Of the myriad topics, questions, and issues of greatest moment to social psychologists, social influence has been the cornerstone of the field. From the earliest inception of modern social psychology, it has been the defining issue. For example, the first experiment in social psychology (Triplett 1898) probed a phenomenon referred to as social facilitation —or the tendency for individuals engaging in like behavior to spur on the actions of one another. In this study, Norman Triplett determined that the presence of a competing cyclist upped the level of performance of bicycle riders when compared to their performance levels as they pedaled against the clock. Similarly, Triplett observed school-aged children who were instructed to turn fishing reels as fast as they could. He found that when they worked in pairs, their performance was superior to what it was when they worked alone.
As Aristotle (384–322 BCE) noted in detail, we are indeed social beings. The mere presence of others affects us beyond our awareness of those effects. The study of social influence has branched from simple demonstrations of socially facilitated behavior to research on such topics as attitude formation and change, conformity, group consensus, obedience, groupthink, prestige suggestion, political socialization, stereotype formation, and many other socially embedded consequences of social influence. This entry will focus on a description of social influence phenomena in the “classic” studies of and approaches to such influence. These classic studies will be separated into two categories: laboratory-experimental research and field-nonexperimental research.
If a stationary pinpoint of light is viewed in an otherwise completely dark room, the light will appear to move. This illusory effect, known as the autokinetic effect, was used by Muzafer Sherif (1935) in early studies of social influence. Participants, who were not told that the light was actually stationary, were asked to indicate any movement they perceived by pressing a key that turned off the light. When a group of participants viewed the light at the same time, the fastest-acting participant turned off the light. After the light was turned off, the participants reported orally on the extent of movement. The results indicated that participants who viewed the light in groups showed a gradual convergence toward an agreed-upon extent of movement. Later, when tested alone, individuals continued giving the group-established judgment.
Sherif interpreted his results as indicating that the groups established a norm that influenced judgments. As will be discussed below, later researchers also emphasized the importance of norms as a determinant of social influence.
Asch’s Perceptual-Judgments Research
Because Sherif’s demonstration of social influence involved public communication, it was more dramatic than the social influence that occurred in the social-facilitation paradigm. Still, the judgments were of an illusion. This was not the case, however, in an influential series of studies by Solomon Asch (1952). In Asch’s initial study, eight individuals, all but one of whom were confederates of the experimenter, made eighteen consecutive judgments in which they matched the length of a standard line with one of three comparison lines. On each of the eighteen trials, judgments were publicly announced in the order in which the group members sat, and the single naïve participant always sat next to last. On twelve of the trials, the confederates responded incorrectly. The results revealed marked individual differences, with three-fourths of the participants being influenced on at least one trial and one-third influenced on half or more of the trials.
Asch later reported a series of interesting follow-up experiments. In one such study, he found that the conformity effect was maximized with three confederates, was greatly reduced with two confederates, and all but disappeared with one confederate. In another study, he found that when a single confederate responded incorrectly in a group of participants, the confederate’s incorrect judgments created amusement and laughter. In still another study, he found that when the confederate who sat in the fourth position responded correctly, the social influence of the group on the participant sitting next to last was dramatically reduced. This latter finding is an early demonstration of minority influence —a topic later pursued by Serge Moscovici (1976, 1980) and others.
Milgram’s Obedience-to-Authority Research
Asch’s research on perceptual judgment did produce dramatic results. However, momentary judgments regarding the perceived length of lines may, in the final analysis, be of no great import. Such is not the case, however, for Stanley Milgram’s (1965) research on obedience to authority.
Milgram’s initial idea was to provide a test of the extent to which individuals would harm someone when given direct orders to do so. History is replete with instances in which such cruelty has occurred. One of the more notable examples was the mass murdering of Jews at such German concentration camps as Auschwitz, Buchenwald, and Dachau during World War II (1939–1945). During the Vietnam War (1957–1975), there is the example of the 1968 My Lai massacre by American troops. A still more recent example is the murdering in 1978 of children and infants by parents and “nurses” at Jonestown in Guyana.
Milgram’s participants had responded to a newspaper advertisement to take part in a study of learning and memory at Yale University. They were all male and ranged in age between twenty and fifty years, and were of both white-collar and blue-collar backgrounds. Participants believed they were being tested in pairs, but one member of the pair was actually a confederate of the experimenter. Following a rigged drawing, the participant was assigned the role of “teacher” and the confederate was assigned the role of “learner” in a supposed test of the effect of punishment on memory. The participant-teacher witnessed the confederate-learner being strapped into an “electric chair” in a room immediately adjacent to the room containing the “shock-generator”—an apparatus containing a series of thirty switches that supposedly delivered increasingly intense shocks to the confederate-learner. The switches were accompanied by labels indicating increasing danger, finally ending with the label “XXX.” The participant-teacher’s task was to deliver increasingly intense shocks whenever the confederate-learner made an error in memory for the word that had been previously associated with a paired stimulus word. The confederate had been trained to make a standard series of errors. If the participantteacher hesitated, the experimenter delivered a series of four prods, beginning with “please continue” and ending with “you have no other choice, you must go on.” The experiment was terminated if the participant-teacher refused to continue after the fourth prod.
Milgram asked forty psychiatrists from a “leading medical school” to predict the results of the study. These experts on human behavior expected that only 1.25 percent of participants would continue to the top step. In fact, Milgram found that 62.5 percent proceeded to that level. The sizable error in the psychiatrists’ predictions provides evidence that even among “experts” there is a tendency to underestimate the extent of social influence.
Zimbardo’s Stanford Prison Experiment
One might suppose that no one could “top” Milgram’s research on obedience to authority. However, that was arguably accomplished in Philip Zimbardo’s Stanford Prison Experiment (Zimbardo et al. 1973). The participants in this study were twenty-one males between seventeen and thirty years of age who responded to a newspaper advertisement to take part in a study of prison life. Participants were randomly assigned to be either prison guards or prisoners. Zimbardo and colleagues arranged for the prisoners to be unexpectedly arrested by local police officers, fingerprinted, and jailed. After remaining in a cell for a while, the prisoner’s were blindfolded, taken to a mock jail (actually the basement of the Stanford psychology building), “stripped naked, skin-searched, deloused and issued a uniform, bedding, soap and towel” (Zimbardo et al. 1973, p. 38). On the first day, the guards began the routine of having the prisoners line up, cite the prison rules, and repeat their individual ID numbers. On the second day, the prisoners rebelled but the rebellion was put down harshly by the guards, who first called in other guard shifts for reinforcement. On subsequent days, the cruelty and abusiveness of the guards intensified, and the prisoners became increasingly passive and demoralized. Not all guards acted the same way. Some were cruel and some were not, but the “good” guards did nothing to protect the prisoners from the cruel guards. Likewise, not all prisoners acted, or reacted, in the same way. On progressive days, five of the prisoners had emotional breakdowns requiring that they be released. The experiment was terminated after six of the originally planned fourteen days.
In this experiment, the social influence came not from the experimenter, as in the Milgram experiment, but was a dynamic consequence of the interaction between two groups and the learned roles associated with these groups. Following this most dramatic demonstration of social influence, there have been no further attempts at even more dramatic demonstrations of social influence, and, primarily for ethical reasons, there have been no further experimental studies of a prison-like environment.
One field-nonexperimental study that has, quite deservedly, received wide attention is Theodore M. Newcomb’s (1943, 1958) research on Bennington College students and subsequent follow-ups of those students (Newcomb et al. 1967; Alwin et al. 1991). Newcomb collected the initial data in 1935, the first year in which the then new Bennington College had a senior class, and data collection was continued until 1939. Since the student body was small, approximately 250 women, it was possible to collect data from almost all of the students and to follow the initial freshman in 1935 until they were seniors in 1939. An important circumstance was that the college was physically isolated and relatively self-sufficient. Tuition was high and the majority of the students came from wealthy families with conservative social attitudes. On the other hand, the faculty was relatively liberal and “felt that its educational duties included the familiarizing of an oversheltered student body with the implications of a depression-torn America and a war-threatened world” (Newcomb 1958, p. 421).
Data collected from most of the students, and in-depth interviews with a subsample of them, indicated that there was an increasing trend toward liberalism with each additional year in college. A questionnaire assessing various social and economic issues (e.g., public relief and the right of labor to organize) revealed a consistent trend toward liberalism. A straw vote in the 1936 presidential campaign indicated that 62 percent of Bennington’s freshman favored the Republican candidate (Alfred Landon), while only 15 percent of the juniors and seniors did so. Consistent with this difference, 29 percent of freshman favored the Democrat candidate (Franklin D. Roosevelt), while 54 percent of the juniors and seniors did so. An even greater ratio of difference occurred for the Socialist and Communist candidates; this varied from 9 percent for the freshman to 30 percent for the juniors and seniors. In observing changes from the freshman to the senior year, Newcomb documented changes toward liberalism that were particularly marked among the more popular students. Newcomb interpreted his results as indicating that the Bennington College community became a positive reference group for most of the students.
Newcomb’s initial data are striking, but the profound effect of social influence was made more obvious by the results of two follow-up studies. Interview data from the first follow-up study (reported in Newcomb et al. 1967) were collected from 1959 to 1960 when the women were in their mid-forties. The results indicated that over the twenty to twenty-five year period, most of the women had retained their liberal political attitudes. Newcomb and colleagues pointed out that the persistence of these liberal attitudes might be partially explained by the fact that they had married men with similar liberal attitudes. Data for the second follow-up study were collected in 1984, the year of Newcomb’s death, when the women were sixty-five to seventy years old (Alwin et al. 1991). Interviews were conducted with 335 of the original sample of 527 women. Most of the interviews were conducted by telephone, but twenty-eight of the interviews were conducted in three-hour, face-to-face sessions. The results indicated that 70 percent identified themselves as “liberal” and agreed with the positions of the 1984 Democrat presidential candidate, Walter Mondale, on such issues as the responsibility of government to improve the social positions of African Americans and other minorities. The women furthermore confirmed that their attitudes had grown more liberal while they were at Bennington.
The persistence of acquired liberal attitudes over an approximately fifty-year period provides strong evidence for the potentially profound effect of social influence. The Bennington study and the follow-up interviews make clear that social influence is not restricted to short-term judgments or behaviors (for example, judgments regarding which line is longer), but can be life changing.
Morton Deutsch and Harold Gerard are widely referenced for interpreting social influence as due to two factors: normative social influence and informational social influence. They define normative influence as “influence to conform to the positive expectations of another” (Deutsch and Gerard 1955, p. 629). The idea is that agreement with another person, another group, or even the self (in the case of internalized norms) leads to expectations of positive feelings. For conformity to internalized norms (and resistance to the influence of others), they add another process. This is that such conformity “leads to feelings of self-esteem or self-approval” (p. 630). As speculated below, acceptance of the influence of others may also be related to self-esteem.
Deutsch and Gerard define informational influence as “influence to accept information obtained from another as evidence about reality” (Deutsch and Gerard 1955, p. 629). They rely on Leon Festinger’s (1950) argument that “where no physical reality basis exists for the establishment of the validity of one’s belief, one is dependent upon social reality (i.e., upon the belief of others)” (Deutsch and Gerard 1955, p. 630). Following his 1950 article, Festinger (1954) elaborated his basic argument into a theory— social comparison theory—which became one of the more influential theories in social psychology. The theory assumes that we have a drive to evaluate our beliefs in terms of their correctness or validity. The validity of our beliefs can sometimes be determined through simple observation, but, when this cannot be done—and only when this cannot be done—we evaluate the validity of our beliefs through social comparison with the beliefs of others.
The idea that the beliefs of others are a source of information determining social influence has become very influential. However, one could argue that even social comparison theory underestimates the informational value of social influence. For example, a dramatic study by Bibb Latané and John Darley (1970) demonstrated that the assumed danger indicated by an increasing amount of smoke pouring into a room was influenced by the number of seemingly unconcerned others. The greater the number of unresponsive others—actually confederates of the experimenter—the longer it took for the single participant to leave the room. This experiment provides a clear example of how, contrary to social comparison theory, social influence can alter even the judgments of physical reality.
Despite the fact that Deutsch and Gerard speculated about both informational social influence and normative social influence, their research primarily focused on normative social influence. As in Asch’s research, Deutsch and Gerard had participants and confederates make length-of-line judgments. They made a number of comparisons among their various conditions, but two of these bear most directly on the influence of group norms. One comparison was between a condition in which the groups were told that each member of the five (out of twenty) groups that made the fewest errors would receive a pair of tickets to a Broadway play versus a condition in which there was no mention of competition among the groups. There was more conformity in the former, competitive condition— presumably because the competition increased the anticipated positive feelings resulting from conformity to the group norm.
The second comparison was between a condition in which the participant (and the confederates) responded publicly, and a condition in which the participant (but not the confederates) responded privately on sheets of paper that were subsequently discarded. There was more conformity in the former, public condition—presumably because only with public responding could the participant’s feelings be influenced by the anticipated reactions of the group members.
What about evidence for informational social influence? Deutsch and Gerard provide no evidence directly bearing on informational social influence, but they do point out that such influence is suggested by a further comparison between the private condition and a control condition in which the participants were alone. Participants made fewer errors in the direction of the group in the alone condition.
Chester Insko and colleagues (Insko et al. 1983; Insko et al. 1985) completed research that they believed could show whether individuals would in fact go along with obviously erroneous opinions of others if they had a chance to learn the true answer subsequently. They set forth a related, but alternative, account of Deutsch and Gerard’s (1955) explanations in which they argued that going along with others involves joint concerns of being right (or not being wrong) and being liked (or not being disliked), and that both are consistent with the general tendency of people to think—or want to think—well of themselves. Specifically with regard to informational influence, and following from the social comparison theory assumption that others’ beliefs provide a social reality for validating one’s own beliefs, they argued that such validation is important because being right, or being correct, is consistent with positive self-esteem, or the general tendency of people to think well of themselves. Similarly, with normative social influence, they argued that agreeing with others creates not only the anticipation of being right but also the anticipation of being liked, and that being liked is consistent with positive self-esteem, or, again, the general tendency of people to think well of themselves.
Insko and colleagues (1983, 1985) conducted two studies, both of which followed the Asch procedure of having a series of confederates and a single participant seated in a row of chairs, with the true participant seated next to last. Unlike in Asch’s research, the judgments were not of line lengths but of colors—for example, whether the blue-green in the middle was more like the blue on the left or the green on the right. On twelve of the eighteen trials, the confederates responded incorrectly. Both studies included two cross-cutting independent variables, thereby creating four cells or conditions. One independent variable was whether the participants responded publicly or privately.
The other independent variable related to whether or not the correct answer supposedly could be accurately determined by a spectrometer. In the so-called determined condition, it was explained that a printout of the spectrometer reading would be available for examination at the end of the experiment. In the so-called undetermined condition, no printout was made available and it was explained that spectrometers could not accurately read mixed wavelengths. Assuming that the judgments of the confederates were used as a valid source of information, and assuming that participants were concerned with being right or being correct, it follows that there should have been more conformity to the confederates’ judgment in the determined than the undetermined condition.
The results indicated that there was more conformity to the confederates’ judgments when the judgments were made publicly rather than privately (regardless of whether the correct answers supposedly could or could not be determined), and there was more conformity when the correct answers supposedly could be determined rather than not determined (regardless of whether the participants responded publicly or privately). These results are consistent with expectations based on the assumed joint concerns of being right (or not being wrong) and being liked (or not being disliked). Thus, both the reduction of epistemic uncertainty and the desire to be liked seem to be factors promoting the power of social influence.
While social psychologists have typically studied the darker side of our dependence on social norms and collective information such as slavish conformity, destructive obedience, normative error, interpersonal cruelty, and bystander “apathy,” there are instances of the collective benefits of social influence that are abundantly evident— for example, in the acquiring of knowledge, the prevention of harmful behavior, and the coordination of activities. The fact that research on social influence has tended to focus on negative consequences should not be taken as an indication that social psychologists necessarily subscribe to the bias implicit in Western, individualistic culture that individual action and choice are always preferable to social conformity. It is obviously the case that many norms are adaptive, and that conformity to such norms, and social learning generally, can be beneficial. One of the profound dilemmas of human existence relates to when we should, or should not, conform to the behavior and judgments of others.
SEE ALSO Asch, Solomon; Attitudes; Authority; Cognitive Dissonance; Conformity; Experiments; Experiments, Human; Experiments, Shock; Groupthink; Milgram, Stanley; Norms; Obedience, Destructive; Persuasion; Prisons; Reality; Sherif, Muzafer; Social Comparison; Social Facilitation; Social Psychology; Socialization; Sociology; Zimbardo, Philip
Allport, Floyd H. 1920. The Influence of the Group Upon Association and Thought. Journal of Experimental Psychology 3: 159–182.
Alwin, Duane F., Ronald L. Cohen, and Theodore M. Newcomb. 1991. Political Attitudes over the Life Span: The Bennington Women after Fifty Years. Madison: University of Wisconsin Press.
Asch, Solomon E. 1952. Social Psychology. Englewood Cliffs, NJ: Prentice Hall.
Baumrind, Diana. 1972. Some Thoughts on Ethics of Research: After Reading Milgram’s “Behavioral Study of Obedience.” In The Social Psychology of Psychological Research, ed. Arthur G. Miller, 106–111. Toronto, Ont.: Collier-Macmillan.
Deutsch, Morton, and Harold Gerard. 1955. A Study of Normative and Informational Social Influences upon Individual Judgment. Journal of Abnormal and Social Psychology 51: 629–636.
Festinger, Leon. 1950. Informal Social Communication. Psychological Review 57: 271–282.
Festinger, Leon. 1954. A Theory of Social Comparison Processes. Human Relations 7: 117–140.
Insko, Chester A., Sarah Drenan, Michael Solomon, et al. 1983. Conformity as a Function of the Consistency of Positive Self-evaluation with Being Liked and Being Right. Journal of Experimental Social Psychology 19: 341–358.
Insko, Chester A., Richard H. Smith, Mark D. Alicke, et al. 1985. Conformity and Group Size: The Concern with Being Right and the Concern with Being Liked. Personality and Social Psychology Bulletin 11: 41–50.
Latané, Bibb, and John M. Darley. 1970. The Unresponsive Bystander: Why Doesn’t He Help? New York: Appleton-Century Crofts.
Milgram, Stanley. 1965. Some Conditions of Obedience and Disobedience to Authority. Human Relations 18: 57–76.
Moscovici, Serge. 1976. Social Influence and Social Change. Trans. Carol Sherrard and Greta Heinz. London: Academic Press.
Moscovici, Serge. 1980. Toward a Theory of Conversion Behavior. In Advances in Experimental Social Psychology, ed. Leonard Berkowitz, Vol. 13, 209–239. New York: Academic Press.
Newcomb, Theodore M. 1943. Personality and Social Change: Attitude Formation in a Student Community. New York: Dryden.
Newcomb, Theodore M. 1958. Attitude Development as a Function of Reference Groups: The Bennington Study. In Readings in Social Psychology, ed. Eleanor E. Maccoby, Theodore M. Newcomb, and Eugene L. Hartley. 3rd ed. New York: Holt.
Newcomb, Theodore M., Kathryn E. Koenig, Richard Flacks, and Donald P. Warwick. 1967. Persistence and Change: Bennington College and Its Students after Twenty-five Years. New York: Wiley.
Orne, Martin T. 1962. On the Social Psychology of the Psychology Experiment: With Particular Reference to Demand Characteristics and their Implications. American Psychologist 17 (11): 776–783.
Orne, Martin T., and C. H. Holland. 1972. On the Ecological Validity of Laboratory Deceptions. In The Social Psychology of Psychological Research, ed. Arthur G. Miller, 122–137. Toronto, Ont.: Collier-Macmillan.
Sanna, Lawrence J. 1992. Self-efficacy Theory: Implications for Social Facilitation and Social Loafing. Journal of Personality and Social Psychology 62 (5): 774–786.
Sanna, Lawrence J., and Paul A. Pusecker. 1994. Self-efficacy, Valence of Self-Evaluation, and Performance. Personality and Social Psychology Bulletin 20: 82–92.
Sanna, Lawrence J., and R. Lance Shotland. 1990. Valence of Anticipated Evaluation and Social Facilitation. Journal of Experimental Social Psychology 26: 82–92.
Sherif, Muzafer. 1935. A Study of Some Social Factors in Perception. Archives of Psychology 27 (187): 1–60.
Triplett, Norman. 1898. The Dynamogenic Factors in Pacemaking and Competition. American Journal of Psychology 9 (4): 507–533.
Zimbardo, Philip G., W. Curtis Banks, Craig Haney, and David Jaffe. 1973. The Mind Is a Formidable Jailer: A Pirandellian Prison. New York Times Magazine, April 8: 38–60.
Chester A. Insko
The influence of others on an individual's behavior.
Human behavior is influenced by other people in countless ways and on a variety of levels. The mere presence of others—as co-actors or spectators—can stimulate or improve one's performance of a task, a process known as social facilitation (and also observed in non-human species). However, the increased level of arousal responsible for this phenomenon can backfire and create social interference, impairing performance on complex, unfamiliar, and difficult tasks.
Overt, deliberate persuasion by other people can cause us to change our opinions and/or behavior. However, a great deal of social influence operates more subtly in the form of norms—acquired social rules that people are generally unaware of until they are violated. For example, every culture has a norm for "personal space"—the physical distance maintained between adults. Violation of norms generally makes people uncomfortable, while adherence to them provides security and confidence in a variety of social situations. Norms may be classified as one of two types: descriptive and injunctive. Descriptive norms are simply based on what a majority of people do, while injunctive norms involve a value judgment about what is proper and improper behavior.
Both conformity and compliance are attempts to adhere to social norms—conformity occurs in response to unspoken group pressure, as opposed to compliance, which results from a direct request. Research has shown that conformity is influenced by the ambiguity of a situation (people are more apt to go along with the majority when they are uncertain about which course of action to pursue), the size of the majority, and the personal characteristics of the people involved, including their self-esteem and their status within the group. A person may conform by acting in accordance with group norms while privately disagreeing with them (public conformity) or by actually changing his or her opinions to coincide with those of the group (private acceptance).
In contrast to compliance, which characterizes behavior toward those who make direct requests but have no authority over us, obedience is elicited in response to a specific demand by an authority figure. The most famous experiment involving obedience was conducted by Stanley Milgram in the early 1960s at Yale University. Forty men and women were instructed to administer electric shocks to another person, supposedly as part of an experiment in learning. (In fact, there were actually no shocks administered, and responses were faked by the "victim," who was part of the experiment.) When the scientist in charge directed the subjects to administer increasingly severe shocks, most of them, while uncomfortable, did so in spite of the apparent pain and protests of the supposed victim. This experiment—which is often referred to in connection with German obedience to authority during the Nazi era—gained widespread attention as evidence of the extent to which people will forfeit their own judgment, will, and values in order to follow orders by an authority figure (65 percent of the volunteers, when asked to do so, administered the maximum level of shock possible). In variations on this experiment, Milgram found that factors affecting obedience included the reputation of the authority figure and his proximity to the subject (obedience decreased when instructions were issued by phone), as well as the presence of others who disobey (the most powerful factor in reducing the level of obedience).
Another type of social influence that can lead normal people to engage in cruel or antisocial behavior is participation in a crowd or mob. Being part of a crowd can allow a person's identity to become submerged in a group, a process known as deindividuation. Contributing factors include anonymity, which
brings with it a reduction of accountability; a high level of arousal; and a shifting of attention from oneself to external events, resulting in reduced self-awareness. The so-called "herd mentality" that results weakens people's normal restraints against impulsive behavior, increases their sensitivity to environmental stimuli, and reduces their abilities to think rationally and fear censure by others.
The relatively new field of environmental psychology investigates the ways in which human behavior is affected by proximity to others in urban environments, most notably the effects of noise and overcrowding. Living in high-density environments has been associated with feelings of helplessness resulting from lack of control and predictability in one's social interactions.
Freedman, J. L., D. O. Sears, and J. M. Carlsmith. Social Psychology. 4th ed. Englewood Cliffs, NJ: Prentice-Hall, 1981.
Milgram, Stanley. Obedience to Authority: An Experimental View. New York: Harper and Row, 1974.
Paulus, P. B., ed. Psychology of Group Influence. 2nd ed. Hills-dale, NJ: Lawrence Erlbaum Associates, 1989.