Obedience in modern society: the Utrecht studies. H.J. Wim Meeus; Quinten A.W. Raaijmakers.

The Utrecht Studies on Obedience, a series of 19 experiments, demonstrated that obedience is extremely high when the violence to be exerted is a contemporary form of mediated violence, and remains high even when the subjects receive detailed information about the task in advance. Observation of the subjects and analysis of data from questionnaires and the debriefing show that the subject's attitude to the experimenter's commands is critical to very critical, and that they found it unpleasant and stressful to carry out the task. This stress was not, however, sufficient to make the subjects disobedient. They attempted to hide their stress from the victim and to act as if nothing was wrong, displaying the behavior of an official. The explanation for the high level of obedience should not be sought in the inability of the subjects to resist the scientific authority (see the results of the condition of Legal Liability), but in their attitude to social institutions and their distant relationship with fellow citizens. Active role playing offers an attractive opportunity for ethically acceptable obedience research.

The systematic study of obedience in the laboratory began in 1963 with the publication of a series of experiments by Stanley Milgram. This was not, however, the first observation of extreme obedience of subjects to a scientific authority. As early as 1924, Landis (1924, p. 459) had reported in a study of emotional reactions that no less than 15 of the 21 subjects (= 71%) were prepared to decapitate a rat, for no reason other than the insistence of the experimenter. Landis noted this fact with no further comment. Frank (1944), 20 years later, was the first social psychologist to be confronted in an experimental situation with the phenomenon of obedience to the experimenter, and in an unexpected way. While trying to design an experimental procedure for studying resistance to the experimenter's instructions, Frank was surprised by the almost absolute compliance (14 out of 15 subjects) with an unpleasant and entirely pointless instruction of the experimenter. In this context Frank spoke of an implicit agreement between the subject and the experimenter (1944, p. 25) and also gave an initial explanation for this phenomenon, without actually using the term "obedience." In his further analyses Frank did not explore in greater depth the causes of this unexpected obedience. This probably explains why, almost 20 years later, Stanley Milgram was again surprised to be confronted with the phenomenon of obedience.

The Accidental Discovery of Obedience

Unlike Frank, Milgram did not ignore this surprising result, discovered by chance. He made it the core of a new research topic: obedience to a scientific authority (Milgram, 1974, p. 80).

The Ecological Validity of Obedience Research

The second criticism regarding the ecological validity of the experiments is of a rather different order, and is based on ethical aversion to Milgram's approach (Baumrind, 1964). The extreme obedience is said to be unique to the laboratory situation. Summarized in one sentence, this criticism states that "normal people ( . . . ) do not give painful electric shocks to innocent individuals, even when they are instructed to do so" (Milgram, 1974, p. 195; our italics). This criticism comprises two elements:

These two elements in this criticism of the ecological validity of the Milgram paradigm formed the starting point for our further analysis of the specific position of the subject in the obedience experiment, and the kind of violence against the victim which is expected of the subject in that situation.

The position of the subject: The subject is becoming trapped. The obedience experiment gives a representation of a typical social structure: the legitimate authority (the experimenter) instructs the subject to seriously mistreat the victim (a confederate), against his will (in Milgram's case by giving painful electric shocks). In this social structure the factors that create obedience are much more in evidence than the factors that could create disobedience: the experimenter has a higher status and exercises his influence in a more direct and consistent manner than the victim (see Meeus and Raaijmakers, 1987, for a detailed analysis of these power relationships).

In this situation, with strong pressure to show obedience, the subject also finds himself in a very specific position, comparable to the position of someone who is suddenly trapped. When the subject starts the experiment there is absolutely no question of conflict. When the victim - after just ten minutes! - begins to protest loudly, the subject immediately faces an unexpected and serious dilemma, with a choice between two actions: stop giving the shocks and ignore the experimenter's instructions, or carry on giving the shocks and thus ignore the protests of the victim. This is essentially a moral dilemma: can it be justified to cause pain to another innocent individual in the name of science? Under normal circumstances the subject is not prepared to cause pain to the victim. This is shown by the results of the control condition: practically all the subjects stop giving shocks as soon as they suspect that the victim is feeling pain. Apparently, the subject has an everyday moral code that does not permit inflicting pain on innocent victims. In the baseline experiment, however, the subject is confronted with an entirely different situation: the experimenter commands him to carry on giving the shocks, in the name of science. The subject actually lacks the necessary competence to resolve this unexpected new dilemma, in which the interests of scientific progress conflict with those of the victim's well-being. His cognitive orientation on this conflict is inadequate: the subject has not been given the opportunity in advance to reflect on the conflict or to anticipate the consequences of his behavior. He is in social and moral isolation: the subject can only orientate upon the experimenter and is unable to discuss the situation with other individuals or fellow subjects, in order to arrive at an agreed standard of behavior (for the effects of social isolation on obedience, see also Gamson, Fireman, & Rytina, 1982). Furthermore, the subject has no fixed moral standards, established in advance, which would enable the choice to be made between the interests of science and the well-being of the victim. Finally, the subject is committed to carrying out the experimenter's instructions. Before the victim starts to protest and there is any hint of a conflict, the subject has already given several electric shocks. To stop giving the shocks would involve breaking the agreement with the experimenter: both Milgram (1974, p. 39) and Back (1979, p. 291) point out the importance of this initial group formation. Furthermore, once the subject has ignored the first - still mild - protests of the victim, it becomes even more difficult to be disobedient later: it is a foot-in-the-door situation (Eckman, 1977; Gilbert, 1981; Shanab & O'Neill, 1979).

In summary, in a situation with strong pressure to be obedient, the subject rapidly and unexpectedly finds himself in a serious moral conflict. He does not have the necessary cognitive and social-moral competence to resolve the new conflict in favor of the victim. The subject's isolated position could thus provide an explanation for the high level of obedience.

Earlier research permits the assumption that the instruction to carry out mediated violence will lead to greater obedience. The effects of distance from the victim are examined in Milgram's experiments (Nos. 3 and 4). The greater the distance from the victim, the greater the obedience. The obedience actually becomes absolute when the subject can neither see nor hear the victim (Milgram, 1965). The results are also unequivocal in relation to the distance from the direct exertion of the violence: the greater the distance, the less the subjects feel responsible and the greater the obedience. In one of Milgram's experimental variations (No. 18, 1974, p. 121) the subject has the role of teacher, but does not have to give the shocks himself: this is done by a confederate, instructed by the subject. The obedience increases considerably (from 65% in the baseline design to 92.5%). In a similar condition in an experiment by Kilham and Mann (1974), 54% of the subjects are obedient, compared with 28% when the subject must give the shocks himself and the confederate takes the role of teacher.

In summary, in Milgram's experiments the subject is instructed to personally exert physical violence directly upon the victim. Modern violence, however, is mediated: those who exert the violence only indirectly observe the negative consequences of their acts to the victim. In a situation like this, a considerable increase in obedience can be anticipated.

Having noted these characteristics of Milgram's experiments, we designed a new experimental procedure, in which the subject is instructed to exert a more mediated form of violence. We also wished to use this procedure to investigate whether the obedience to the experimenter was a consequence of the subject's isolated position.

The Utrecht Studies on Obedience

Our obedience studies were conducted in the early 1980s at Utrecht University in the Netherlands. A series of six pilot studies was carried out in 1981 in order to design the baseline experiment on administrative obedience; 82 subjects participated in these pilot studies. The series of 17 experiments on administrative obedience was conducted in 1982-1983, with a total of 352 subjects. Finally, two experiments using the Milgram procedure were carried out in 1985; 60 subjects participated.

The Experiments on Administrative Obedience

The psychological-administrative force the subject had to exert constitutes a typical form of mediated violence. The consequences of this mediated violence are in two ways indirect: (a) the victim only becomes tense due to the stress remarks after a period of time, and (b) it only becomes clear after half of the stress remarks that the victim might fail the test as a consequence of them. This constitutes a major difference with the experiment of Milgram in which the shocks of 150 volts and up cause direct pain to the victim. Since it is easier for subjects to exert more mediated violence, we expected a higher level of obedience in our experiment compared to that of Milgram.

The percentage of obedient subjects was higher in the baseline experiment on administrative obedience than in the baseline experiment of Milgram. This [TABULAR DATA FOR TABLE 1 OMITTED] confirms our hypothesis concerning the different types of violence in the two experiments. The violence that the subject must exert in administrative obedience experiments is more mediated than that in Milgram's experiment, making it easier to be obedient.

The baseline experiment was then repeated with a specific group of subjects: personnel officers. Our hypothesis was that these subjects would employ an explicit ethical code in relation to dealing with applicants, and would not cooperate with the procedure, or at least not right to the end. This expectation was not upheld: the percentage of obedient subjects was just as high for the personnel officers as for the subjects in the two earlier baseline experiments. This experiment was an initial indication that the high level of obedience in the baseline experiments on administrative obedience cannot be attributed to a lack of orientation on the part of the subject regarding the experimental situation: personnel officers are involved with application procedures every day in their work.

Milgram-type experimental variations (Experiments 6-7). These experiments were carried out to investigate whether changes in the experimental situation would produce a decrease in obedience comparable to that in Milgram's experimental design.

As expected, the level of obedience in both experimental variations was significantly lower than in baseline experiment 1 (see Table 1). The decline in obedience was comparable to that in the same experimental variations in the Milgram experiment. In both procedures the decline in obedience was greater in "Two peers rebel" than in "Experimenter absent."

Our expectation was that in this condition a lower percentage of subjects would be obedient than in the baseline experiment. This expectation was not upheld: the obedience was just as high as in Baseline I. There was, however, a difference in the perception of the experiment between the subjects in "Advance information" and Baseline I. In the questionnaire that they filled in after the experiment but before the debriefing, the subjects in "Advance information" were much more critical of the way the applicant was treated than the subjects in Baseline I. The advance information thus had an effect on the subjects' evaluation of the experiment, but not on their obedience. Their more critical attitude toward the experiment was not translated into behavior.

Legal liability (Experiments 9-11). The unexpected result of "Advance information" led us to question whether individual subjects were capable of any form of disobedience in the baseline experiment on administrative obedience, i.e., in the presence of the experimenter and without the support of others for rebellion. An indication that this is actually possible was found in studies in which the subjects had to carry out unpleasant tasks or tasks that entailed personal danger. Obedience levels were high in the studies with the unpleasant task, such as eating dirty crackers (Kudirka, 1965), receiving painful stimuli (Scott, 1980), or answering intrusive questions about one's personal life (Feldman & Scheibe, 1972). In the experiments with tasks that caused serious personal danger, such as picking up a poisonous snake (Rowland, 1939), taking a coin from a dish of nitric acid (Orne & Evans, 1965; Young, 1952), and carrying out a break-in at their own risk (West, Gunn, & Chernicky, 1975), a low level of obedience was found, ranging from 0 to 17%. Subjects are clearly able to defy authority when obedience involves a serious risk to themselves.

The percentage of obedient subjects was significantly lower in the "Legal liability" condition than in Baseline I. This result is consistent with the experiments mentioned above in which obedience entailed a personal risk for the subjects. Subjects are also capable as individuals of rebelling against the authority if there is a risk to themselves.

The findings in "Legal liability" prompted two further experiments to ascertain explicitly whether there is a difference in obedience when a subject is legally liable for his actions and when the subject is given legal cover. As in the "Advance information" experiment, the subjects received detailed instructions by post about the experiment and their task a week before the experiment. In the condition "Advance information - legal liability," the subject was informed that he would be legally responsible if the subject should take legal action, while in the condition "Advance information - legal cover," the subject was informed that the psychology department was legally responsible for the consequences of the experiment.

The level of obedience in "Advance information - legal liability" was lower than in Baseline I and not different from "Legal liability." Obedience in "Advance information - legal cover" was not different from Baseline I and was greater than in "Legal liability" and "Advance information - legal liability." These results supported the findings of "Legal liability." When subjects themselves can become the victims of obedience, most of them refuse to carry out the experimenter's instructions.

Role-playing experiments (Experiments 12-17). The obedience research has featured prominently in the discussion about the ethical permissibility of deception in social psychological experiments (Kelman, 1967; Miller, 1972, 1986). A proposal put forward within this discussion is to use role-play as an alternative research strategy, since this does not involve subjects being deceived or exposed to extreme stress or other harmful stimuli.

Quite a number of role-plays of Milgram's experiment have been carried out. All the studies on expected behavior (Milgram, 1974, pp. 27-31) are role-plays of the description - imagination type. In these role-plays (Freedman, 1969; Milgram, 1963, 1974; Miller, Gillen, Schenker, & Radlove, 1974; Mixon, 1972) a systematic and serious underestimation of obedience is found: in practically all cases the anticipated obedience is less than 10%. In the nonactive role-plays (Bierbrauer, 1979; Hess, 1971; Mixon, 1972) the results are much closer to the actual level of obedience: compared with the obedience of 65% found in Milgram's baseline experiment, the average obedience in the nonactive role-plays is 43%. In the active role-plays the results are even closer to the level of obedience found in the experiment: there is an average difference of just 1% between the experimental obedience and that in twelve role-plays (Albrecht, 1973; Bierbrauer, 1979; Geller, 1978; Holland, 1969; Mixon, 1972; O'Leary, Willis, & Tomich, 1970)!

These results were replicated in the role-plays of the administrative obedience paradigm. In two role-plays of the description-imagination type, a significant underestimation of the actual obedience was found. The nonactive role-plays of Baseline I and "Legal liability" also showed a significant under-estimation of obedience. In the active role-plays of Baseline I and "Legal liability," the same percentage of obedience was found as in the comparable experiments.

The results of the role-plays therefore show that the prediction of the obedient behavior becomes more accurate as the amount of information about the situation of the obedience experiment increases. In the description-imagination role-plays the subjects only read a description of the experiment and their prediction of obedience was the least accurate. In the nonactive role-plays the subject observes the experiment and this improves the prediction of obedience. However, actually playing the role of the subject in the active role-plays gives the most precise information about the experimental situation and results in the best prediction of the level of obedience.

Experiments with the Milgram procedure (Experiments 18-19). These two experiments, "Advance information - painful shocks" and "Advance information - deadly shocks" were a replication of the experiment "Advance information" with the Milgram procedure. The subjects here also received by post a week in advance a complete description of the experimental procedure and their task; the subjects thus knew in advance that they would have to administer electric shocks. In the condition "Advance information - painful shocks" the shocks were described as painful or extremely painful, but not injurious, and in the condition "Advance information - deadly shocks" as extremely painful and even deadly. The dependent variable in both experiments was initial obedience: the subjects who came to the laboratory were prepared, after receiving the full instruction, including practice with the shock generator, to start the experiment were classified as obedient.

The percentages of obedient subjects in "Advance information - painful shocks" and "Advance information - deadly shocks" were 57 and 43, respectively, which are significantly lower than in "Advance information." Once again the level of obedience here is high and the influence of the type of violence is seen: with physical violence subjects are more inclined to be disobedient than with mediated psychological violence. It was also found that authoritarian subjects obeyed significantly more often (71%) than antiauthoritarian (35%). This was the only time during the entire series of experiments that we were able to demonstrate an effect of personality characteristics (sex, age, political preference, professional situation, and locus of control) on obedience. Our explanation for this is that the degrees of freedom for the subjects in "Advance information - painful shocks" and "Advance information - deadly shocks" are relatively high. In experiments of this kind it is easier for the influence of personality characteristics to emerge. We therefore do not interpret this result as providing support for a more interactionist explanation of obedience (Blass, 1991, 1992) in the standard obedience experiment.

In percentage terms there is a high level of obedience in both the experiments with the administrative obedience paradigm and those using the Milgram procedure. The authority of the scientific expert is considerable and there is no evidence that this has declined over the last thirty years.

The type of violence has an effect. The experiments on administrative obedience, without exception, reveal greater obedience than the comparable conditions with the Milgram procedure. The reason for this is that physical violence is more direct and therefore more difficult to apply than the more indirect psychological-administrative violence. We assume that psychological-administrative violence is a characteristic of normal social circumstances in modern Western societies. This is why the experiments on administrative obedience are so important: they demonstrate that obedience in the use of modern violence is extremely high.

The obedience in the experiments on administrative violence is not caused by any lack of orientation on the part of the subject in relation to the experimental situation: even when this orientation is supplied in advance, the level of obedience is extremely high. The origins of this obedience must therefore be sought elsewhere.

The high level of obedience does not result from the inability of the subjects to resist the authority. At the first sign of danger that they may themselves become the victims, most of them are perfectly able to disobey, as shown by the experiments on "Legal liability." Here, then, we have a reason for the high level of obedience: it is found under one special condition - when the victim is a third party. Subjects are quite capable of rebelling against authority, but they choose not to because in the ultimate analysis they are indifferent toward the victim.

In the active role-play the same level of obedience is found as in the real experiments. We therefore conclude that role-play can be a useful instrument in obedience research. New research designs or experimental conditions could first be tested using an active role-play. If significant results are achieved in this way, the design could be submitted to the ethical committee of the national professional psychology organization, in the same way as the application to conduct the experimental version. An advance test of this kind would make it possible to design new obedience experiments in an ethically responsible manner. In view of the small number of obedience experiments carried out since Milgram (Miller, 1986), this offers an attractive opportunity.

We have shown that the degree of obedience increases considerably when subjects are instructed to exert a more modern form of violence. It was also found that the level of obedience cannot be attributed to the isolated position of the subject, and that subjects are capable of disobedience when there is a danger of becoming victims themselves. So how can the high level of obedience in the baseline experiment be explained?

We present here two explanations, focusing on the ritual character of the experimental task and the agentic state of the subject. We look at whether these explanations offer insight into the obedience that is found within the administrative obedience paradigm. We first show how these explanations can be applied to the Milgram experiment.

Gilbert (1981) argues that the high level of obedience in the Milgram experiment is the effect of the ritual character of the subject's task. Once the subject has started out by giving a few mild shocks, in time it becomes practically impossible to stop. Each new shock differs only very slightly from the previous shock. The subject is trapped in this ritual of shocks, which increase very gradually in severity.

This feedback effect has also been investigated in more detail within the administrative disobedience paradigm. Using systematic observation of behavior, we analyzed the role behavior of the subjects (in Baseline I and the condition "Experimenter absent"). The typical role behavior associated with obedience is passive-negative: the subject executes his task in a neutral and official manner, thus ignoring the protest of the applicant. This typical role behavior can be broken in a variety of ways: the subject responds actively to the victim, interrupts the procedure, or protests against it and must be brought back to the role by the commands of the experimenter. This role-breaking behavior of the subject was found to be displayed a total of 90 times (with 19 observed subjects) in the baseline experiment. In two thirds of the cases it was an immediate reaction to the introduction of a new quality of protest from the victim; only in one case, however, did it lead to real disobedience. In the condition "Experimenter absent" this role-breaking behavior was observed 58 times in total. Forty-five percent of these instances were a direct response to a new feedback norm. In no less than 12 of the 15 observed subjects, moreover, this behavior resulted in disobedience.

These results lead us to conclude that the ritual of the subject's task can indeed be broken by the victim's protest, but the results show that breaking the ritual in the administrative obedience paradigm results in disobedience to a much lesser extent than in Milgram's experiment. A further analysis is required in order to explain the high level of obedience in our experiment.

The concept of the agentic state is a crucial feature of Milgram's theoretical explanation of obedience. Milgram uses the concept in two ways. On the one hand, it represents the psychological position in which the obedient subject finds himself during the experiment: he feels no responsibility for the victim's suffering. On the other hand, the agentic state is a response that occurs within complex social contexts. Milgram argues that an individual finds himself in an agentic state when he functions within the framework of the social system. He then acts in conformity with the collective Superego, which is managed by a power base. Only when the individual acts outside the social system, acts autonomously, does he follow his own conscience. In the obedience experiment the subject acts within the framework of the social institutional system of science and is in the agentic state.

Why do subjects enter the agentic state? Like Milgram, we are of the opinion that the ultimate answer to this question cannot be found in the analysis of the experimental situation, but that we must look at the social context in which people live and the historic changes that this has undergone. In other words, we need a historical sociological analysis. An analysis of this kind must focus on the double relationship which the subject has in the obedience experiment: on the one hand with the experimenter and on the other with the victim. The experimenter functions as the representative of a social institution, science, while for the subject the victim is a neutral fellow citizen. An explanation for the existence of the agentic state must therefore contain two elements: (a) the attitude of average citizens toward institutions and (b) their attitude toward fellow citizens. We assume that the answer to the question does not necessarily demand a new theory, but can be provided using existing theories. Finally, we must note that such a historical analysis cannot be bound by the same strict criteria as apply in experimental research, for the simple reason that history is not open to experimental manipulation. The criterion here then cannot be experimental control, but rather plausibility (Sennett, 1986, p. 43).

A more recent analysis focusing specifically on obedience is that of Kelman and Hamilton (1989). They distinguish three ways in which the citizen can be attached to the national (political) system: (a) rule orientation: the citizen complies with the societal rules; (b) role orientation: the citizen identifies with the social system; and (c) value orientation: the citizen takes on the values of the social system. Kelman and Hamilton demonstrate empirically that people with these three orientations have different views on obedience. More specifically, those with a value orientation rate disobedience more highly and are more likely to say that they are disobedient. The problem with Kelman and Hamilton's data, however, is that they are survey data and the respondents estimate their own obedience and that of others. Characterized in terms of role-plays, these are estimations obtained via a description-imagination type of role-play. We showed above that these estimations are completely unreliable and quite irrelevant as regards behavior in the obedience experiment. The importance of Kelman and Hamilton's work is therefore not in their empirical data but in their conceptualization of the three orientations. This indicates that the subject in the obedience experiment is functioning with a role orientation. We see rule orientation as a concept that represents a class of behavior orientations, one of which is the agentic state.

There have been several studies within social psychology that are related to the obedience experiments and investigate the relationship among citizens in modern societies. In their research on the unresponsive bystander, Latane and Darley demonstrated citizens' low level of willingness to assist their fellows in emergencies. Personal responsibility is also a necessary condition here for intervention on behalf of the victim (Latane & Darley, 1970, p. 121). In one of his studies on deindividuation, Zimbardo (1970) showed that particularly in large cities people will plunder the property of fellow citizens without the slightest reservation. Milgram (1977) also presented a negative image of the relationship among citizens, which he described in terms of the "familiar stranger." Citizens who repeatedly encounter each other remain strangers because they cannot find the energy to initiate interaction.

Parson's analysis explains the relationship between the subject and the victim in the obedience experiment. For the subject, the victim is not a person but someone in a certain role position, whom he judges according to general standards - in a test procedure! - in an affective neutral way. Given these basic characteristics of the relationship between subject and victim, it is only "normal" that the subject feels no involvement with the victim and is guided by the representative of the institution with legitimate authority: science.

Baumrind, D. (1964). Some thoughts on ethics of research: After reading Milgram's "behavioral study of obedience." American Psychologist, 19, 421-423.

Bierbrauer, G. (1979). Why did he do it? Attribution of obedience and the phenomenon of dispositional bias. European Journal of Social Psychology, 9, 67-84.

Blass, T. (1991). Understanding behavior in the obedience experiment: The role of personality, situations, and their interactions. Journal of Personality and Social Personality, 60, 398-413.

Eckman, B. K. (1977). Stanley Milgram's "obedience" studies. Et cetera, 34, 88-99.

Geller, D. M. (1978). Involvement in role-play simulations: A demonstration with studies on obedience. Journal of Personality, and Social Psychology, 36, 219-235.

Gilbert, S. J. (1981). Another look at the Milgram obedience studies: The role of the gradated series of shocks. Personality and Social Psychology Bulletin, 4, 690-695.

Holland, C. N. (1969). Sources of variance in the experimental investigation of behavioral obedience. (Doctoral dissertation, University of Connecticut). Dissertation Abstracts International, 29, 2802A.

Ingram, P. (1979). Deception, obedience and authority. Philosophy, 54, 529-533.

Kelman, H. C., & Hamilton, V. L. (1989). Crimes of obedience. New Haven, CT: Yale University Press.

Kilham, W., & Mann, L. (1974). Level of destructive obedience as a function of transmitter and executant roles in the Milgram obedience paradigm. Journal of Personality and Social Psychology, 29, 696-702.

Meeus, W., & Raaijmakers, Q. (1986). Administrative obedience: Carrying out orders to use psychological-administrative violence. European Journal of Social Psychology, 16, 311-324.

Meeus, W., & Raaijmakers, Q. (1987). Administrative obedience as a social phenomenon. In W. Doise & S. Moscovici (Eds.), Current issues in European social psychology (pp. 183-230). Cambridge: Cambridge University Press.

Milgram, S. (1963). Behavioral study of obedience. Journal of Abnormal and Social Psychology, 67, 371-378.

Milgram, S. (1965). Some conditions of obedience and disobedience to authority. Human Relations, 18, 57-76.

Milgram, S. (1974). Obedience to authority: An experimental view. New York: Harper & Row.

Miller, A. G. (1986). The obedience experiments: A case study of controversy in social science. New York: Praeger.

Miller, A. G., Gillen, B., Schenker, C., & Radlove, S. (1974). The prediction and perception of obedience to authority. Journal of Personality, 42, 23-42.

Mixon, D. (1972). Further conditions of obedience and disobedience to authority. Dissertation Abstracts International, 32, 4848-B. (University Microfilms No. 72-6477.)

WIM MEEUS is Professor of Adolescent Psychology at Utrecht University, the Netherlands. He received his Ph.D. in psychology from Utrecht University in 1984. His areas of specialty are obedience and personal and social development in adolescence.

QUINTEN RAAIJMAKERS is Associate Professor of Adolescent Psychology at Utrecht University, the Netherlands. He received his Ph.D. in psychology from Utrecht University in 1984. His areas of specialty are obedience, and moral development and political attitudes in adolescence.

Article A18320030