views updated


I. SOCIAL RESEARCHRobert L. Kahn and Charles F. Cannell







The interview has been defined as a conversation with a purpose, and the purposes for which interviews are conducted are many and various. They include the purpose of therapeutic change, as in the psychiatric interview; the purpose of instruction and appraisal, as in the interviews initiated by a supervisor with a subordinate; and the purposes of selection and assessment, as in the interviews conducted with applicants for jobs or with students applying for admission to universities. In all these situations, there is the transaction of giving and getting information, and the understanding of this transaction as the immediate task of interviewer and respondent. This immediate task, however, is embedded in a larger cycle of purposive activities that define the roles of interviewer and respondent more exactly, reflect the motives of both for undertaking the interview, and stipulate the consequences of the interview for other aspects of their lives.

To label an interview “psychiatric” or “therapeutic,” for example, implies that it probably has been initiated by the respondent (or patient) and that his motivation in doing so is to obtain relief from certain symptoms or strains of a mental or emotional sort. Moreover, the interviewer is seen not only as an information getter but also as a direct and powerful source of help; the interview is seen not only as an informational transaction but also as part of the therapeutic experience.

By contrast, the research interview, to which this article is addressed, may be defined as a two-person conversation that is initiated by the interviewer for the specific purpose of obtaining information that is relevant to research. Such an interview is focused on content specified by the usual research objectives of systematic description, prediction, or explanation. Other characteristics of the research interview are more variable. Typically, however, the differentiation of roles between interviewer and respondent is pronounced. The interviewer has not only initiated the conversation; he presents each topic by means of specific questions, and he decides when the conversation on a topic has satisfied the research objectives (or the specific criteria which represent them) and when another topic shall be introduced. In the research interview the respondent is led to restrict his discussion to the questions posed.

The consequences of the research interview for an individual respondent are often minimal and almost always removed in time, space, and person from the interview experience itself. The respondent is asked to provide information about himself, his experiences, or his attitudes to an interviewer who has no direct power or intention to provide therapy, instruction, a job, or any other major tangible reward. If the research interview does contribute to such a reward, it does so through a sequence of events that involves the aggregation of responses from numerous interviews, some process of data reduction and inference, and some additions to the description or explanation of social facts. From this enlarged base of knowledge may come applications or decisions of policy that have great importance for the respondent, but the sequence is complex and often uncertain. Nevertheless, the prospect of ultimate benefit, public or personal, from the accumulation of knowledge is one major basis for respondent agreement to participate in a research interview.

Perhaps the prototypical example of research interviews is provided by the national census. Most countries of the world conduct some kind of population count, and in many countries the census has been expanded to provide with regularity an inventory of social resources and problems. Census interviews usually make only modest demands on interviewer and respondent. They are brief; they ask for demographic data well within the respondent’s knowledge and not of a kind that he is likely to regard as confidential. Moreover, the information is requested under circumstances familiar to, or expected by, most respondents, and the request is backed by the legitimate power of the national government.

Similar to the census in most of these respects is a whole class of brief, officially sponsored, information-getting interview surveys. In the United States alone, hundreds of thousands of such interviews are conducted by agencies of government in randomly selected homes each year to provide continuing data on family income, employment and unemployment, health and illness, and other aspects of economic and social welfare.

Almost as widely known as census taking and other government-sponsored research that involves interviewing are the activities of those private agencies which conduct recurrent interview studies of public opinion on national and international affairs, family life, and other subjects of public interest. The Gallup Poll is typical of private organizations that conduct such surveys and, in the United States and much of Europe, its name has become a general term for describing them. The interviews conducted by such polls resemble those of the census in brevity, simplicity, and the avoidance of very private material. However, public opinion interviewing differs from most government-sponsored surveys in dealing with matters of attitude rather than of fact and in depending on interviewer persuasiveness rather than on legal authority and prestige to obtain respondent cooperation. Market research and studies of readership are usually of like simplicity and brevity, although more elaborate and indirect techniques of interviewing have often been used in such studies.

It is likely that the most ambitious and demanding use of the interview as a research technique has been made by social scientists in the course of psychological, sociological, political, and economic investigations. Such studies often involve interviews of an hour or more, on subjects that may raise difficult problems of recall, potential embarrassment, and self-awareness. Consider as examples the recurring studies of consumer behavior and family income (Katona 1960), the studies of fertility and family planning (Freedman et al. 1959), the studies of sexual behavior (Kinsey et al. 1948), the studies of mental health and illness (Gurin et al. 1960), of political behavior (Michigan . . . 1960), and the many studies of supervisor–subordinate relations and worker attitudes (Argyris 1964; Herzberg et al. 1959; Organizational Stress 1964; Likert 1961).

These examples suggest a conclusion that can hardly be questioned: much of the data of social science is generated by means of the interview. Sociologists, psychologists, anthropologists, political scientists, and economists depend on interviews to obtain data for describing the phenomena of interest to them and for testing their theories and hypotheses about those phenomena.

Moreover, the use of the interview in these disciplines is not limited to surveys and other studies done in the field; interviewing is a necessary element in laboratory research as well. Laboratory experiments in psychology and other social sciences typically involve a situation contrived by the experimenter in order to introduce some factor into the experience of the people who are his experimental subjects, and to do so under conditions in which their reactions can be closely studied. That study often requires interviewing as well as observation, physiological measures, and the like. The experimenter depends on the subject to report anxiety or elation, increased confidence or reduced self-esteem, and feelings of acceptance or rejection (see, for example, Asch 1952; Mil gram 1965). The interview helps the experimenter to learn whether the intended manipulation of a variable really “took,” and if so, whether it had the predicted effects.

In short, the social scientist, from the first nineteenth-century British surveys of poverty (Booth et al. 1889-1891) and the early psychophysical experiments in the laboratory (Boring 1929), has been, willy-nilly, an interviewer. Whether he has been too much an interviewer (Webb et al. 1966) is a question of prime importance for the strategy of social science but not for the present discussion. Like other scientists, the social researcher has attempted to measure rather than merely to describe in qualitative terms, and for him the interview has been the most useful instrument in the measurement process.

The interview as measurement

We have defined the interview as a conversation with a purpose and have further specified that the purpose with which we are concerned is information getting. The research interview, however, is not after mere information; it has to do with that particular quantitative form of information getting called measurement. The interview is one part, and a crucial one, in the measurement process as it is conducted in much of social research. Thus, the use of the interview is subject to the laws of measurement; it can be properly judged by the standards of measurement, and it suffers from the limitations of all measurement processes in degrees peculiar to itself.

The key concept for thinking about the adequacy of measurement is validity, defined as the extent to which an instrument and the rules for its use in fact measure what they purport to measure (Kaplan 1964; Research Methods in Social Relations 1959). Inferences about validity, however, particularly about interview validity, are too often made on the basis of face validity, that is, whether the questions asked look as if they are measuring what they purport to measure. A preferable way of thinking about validity, and a basis for developing tests of validity, is the question of what a given measurement will do. Does the measure do the things that theory and experiment have convinced us it should do (Campbell 1957; Cronbach 1946; Coombs 1964)? For example, does a test that purports to measure intelligence enable us to predict scholastic achievement to some significant degree?

A similar approach to the validation of interview measures involves the comparison of the interview measure with some other measure that has already met the test of validity. This kind of comparison has been called convergent validity, in part to distinguish it from other approaches to the validation of constructs. If the two measures in fact agree, there is a presumption of validity for the measure being tested at least as great as that of the measure taken as standard. Thus hospitalization data obtained from interviews might be validated against hospital admission records. When the standard measure has already been validated, the method of convergent validity is powerful. For example, when the results of political surveys correspond to election statistics, or when the results of consumer surveys agree with the volume of actual purchases, we have confidence in the survey measures. The problem of validation becomes more difficult, however, in the case of interview measures of attitudes, for which no independent and objective measures exist in quite the same terms.

Questions of validity and invalidity are only part of the problem of measurement adequacy. A measure is invalid to the extent that it measures something more than or less than it purports to measure. Put another way, the mark of invalidity is bias, which is a systematic or persistent tendency to make errors in the same direction, that is, to overstate or understate the “true value” of an attribute. Scarcely less important than validity is reliability, which has to do with the stability and equivalence of a measure (Cronbach 1949). The reliability of an interview measure is defined by such questions as these: If the measure is used repeatedly in the same circumstances, will it yield the same results? If it is used by different interviewers to measure the same attribute, will it produce the same results? Methods for determining the reliability of a measure are to arrange for repetitions of it in identical circumstances (“test-retest reliability”) or, if the measure involves numerous items, to compare the results obtained on the basis of one half of the items, randomly selected, with the results obtained by using the other half of the items. The latter method is called the “split-half reliability” test.

The relationship between validity and reliability is complex. For the interview method, however, it is important to remind ourselves that a measure may be valid without being reliable; that is, it may measure what it purports to, but do so badly. On the average, such a measure obtains the “true value,” but its variance is large. Repeated use of the measure in the same circumstances produces values which are random about the mean but vary from it by large amounts.

The question of how to insure or achieve measurement adequacy by means of interview procedures can be answered in several ways. One solution, of course, is to restrict oneself to measures already developed, for which pedigrees of reliability and validity have been established. Many paper and pencil tests and some interview scales have been developed with enough attention to methodo-logical considerations so that such data are available. Examples include numerous personality scales, tests of intelligence and reading ability, the census procedures for ascertaining labor-force status and occupation, measures of political party identification, and others. Unfortunately, in the social sciences standard, well-validated measures are not available for most concepts. As a result the investigator is commonly faced with the need to develop his own measures.

For the investigator who must create his own interview measures, there exists a considerable accumulation of general principles and specific procedures to guide him in the preparation of questions, scales or sets of questions, and questionnaires (Payne 1951; Kahn & Cannell 1957; Richardson et al. 1965). All of these help to achieve validity in measurement. It is also possible to improve validity by including in the data collection measures of potential sources of bias, so that there can be an after-the-fact assessment of the extent of their intrusion and a correction in the raw data. The “lie scale” of the Minnesota Multiphasic Personality Inventory, the Edwards Social Acceptability Scale, and the Mandler-Sarason Test Anxiety Scale are examples of measures that are used for such ex post facto statistical corrections with some success. The achievement of high reliability depends on the same basic principles but is particularly enhanced by specifying the exact wording of the questions to be asked in the interview, as well as the forms and range of behavior that may be used in the interview to evoke response, and by using multiple questions rather than single questions for the measurement of each concept or variable.

In short, adequacy of measurement by means of interviews requires knowledge of the conditions for a successful interview, and the skill to meet those conditions both in the construction of questionnaires or interview schedules and in the conduct of the interview itself. These issues are the subjects of the following sections of this article.

Conditions for successful interviewing

Interviews are not uniformly successful. Respondents differ in ability and motivation; inter-viewers differ in skill; and interview content differs in feasibility. While many approaches have been taken to these problems, three broad concepts seem to comprise much of the available research and advice. These are accessibility of the required data to the respondent, cognition or understanding by the respondent of his role and the informational transaction required of him, and motivation of the respondent to take the role and fulfill its requirements. These are not independent factors; they can be thought of as a set of interrelated conditions for attaining an adequate interview, and most of the specific techniques of interviewing (to be discussed in the following section) can be thought of as means of meeting these conditions.


The simplest condition for interviewing occurs when the datum which the interviewer requires is completely accessible to the respondent, that is, when he has the information in conscious form, clearly conceptualized in the terms used by the interviewer. This condition is typically met for simple demographic data—age, family size, and the like. To the extent that the required data are accessible to the respondent, the interviewer can turn his attention to problems of cognition and motivation, making sure that the respondent understands what is asked of him and that he is willing to provide the information which he possesses. To the extent that the data are inaccessible to the respondent, this inaccessibility constitutes the first problem in interviewing.

Three major reasons for inaccessibility can be distinguished, each with its own implications for the formulation of questions and the conduct of the interview. First, the material may simply have been forgotten (Bartlett 1932). The respondent may once have been in conscious possession of the required information, but it has receded from conscious recollection. A second kind of inaccessibility has to do with repression; an event is important or recent enough to be remembered, but it involves sufficient emotional stress to have been obliterated from conscious memory. The third category of inaccessibility has to do not with the intrinsic content of the material sought but with the terms or categories in which the interview requires recollection and communication. Problems of language, vocabulary, and understanding are involved, as well as differences in social class, subculture, and region. This aspect of inaccessibility is related to the cognitive conditions for successful interviewing. For example, an interviewer may be so ill-advised as to ask workers to recount recent experiences on the job which were “ego-enhancing” and “ego-threatening/’ A respondent may be quite insightful and observant, quite able to describe in some detail his experience on the job, and yet be unable to respond meaningfully to this question. Moreover, his difficulties are likely to persist even after the words “ego-enhancing” and “ego-threatening” have been defined for him, because he has not thought of his experiences in terms of such categories.

Cognitive conditions

A second requirement for successful interviewing has to do with respondent cognition or understanding. The respondent role is by definition an active, self-conscious one, and the respondent can meet its expectations best when he understands them fully. Specifically, he needs to know what constitutes successful completion of the role requirements and to know the concepts or terms of reference by means of which he is being asked to provide data. Without this understanding, data accessible to the respondent are nevertheless likely to remain unreported because interviewer and respondent lack a common frame of reference, a common conceptual language, or common standards of response adequacy and excellence.

How much understanding of the research enterprise by the respondent is appropriate will vary with the demands being made in the interview, and how much effort the interviewer must devote to the development of such understanding will depend on the extent of the interview demand and the sophistication of the respondent. Typically the respondent need not understand the nature of the measurement being attempted, the construction of scales, or the plans for computer analysis. He should understand the requirements of his own role in the interview—the demands to be made on him, and the criteria for relevance and completeness. To understand these things and to be motivated to accept the role may, in turn, require acquaintance with the over-all aims of the research enterprise, information about its compatibility with the respondent’s own goals, and reassurance about its risks. These issues bring us to the third condition for successful interviewing.

Respondent motivation

There is general agreement among students and practitioners of interviewing that respondent motivation or willingness to report is a prime condition for successful data collection; it could hardly be otherwise. There is little agreement, however, about the theory or model of motivation that is most appropriate to the interview, and about the major sources of re-

spondent motivation. Kahn and Cannell (1957) propose a dual emphasis: intrinsic motivation, be-cause the experience and relationship with the interviewer are valued by the respondent; and instrumental motivation, because the respondent sees that the enterprise of which the interview is a part is congruent with his own goals and values. Kinsey and his colleagues (1948) stress altruism as the initial source of respondent motivation, although many of their interview descriptions seem to rely more on the assumption of legitimate authority and the use of medical–scientific prestige. Richardson, Dohrenwend, and Klein (1965) explain respondent motivation in terms of altruism, emotional satisfaction, and intellectual satisfaction. Such partial agreements and discontinuities are hardly surprising; they reflect the more general diversity of motivational theories. The interview is one form of complex molar behavior; attempts to understand it will inevitably share the contemporary strengths and weaknesses of motivational theory as a whole.

Despite the lack of agreement on any one motivational model, the research evidence on the interviewing process (Hyman et al. 1954; Riesman 1958; Kahn & Cannell 1957; Richardson et al. 1965) strongly urges that respondent motivation be conceptualized in terms that take account of the social situation of interviewer and respondent, the nature of the transaction between them, their perceptions of each other and of their joint task, and the effects of such perceptions. In short, the evidence argues in favor of a motivational model that treats the interview as a social process and regards the interview product as a social outcome. One such model is presented in Figure 1.

This model is compatible with the role-oriented view of the interview. It stipulates that the interview product or outcome is the immediate and joint result of interviewer and respondent behavior; that the behavior of both interviewer and respondent stems from their attitudes, moives, expectations, and perceptions; and that these, in turn, can be understood as reflecting more enduring attributes of demography and personality. The model also emphasizes the interaction of respondent and interviewer. The behavior of the interviewer is perceived by the respondent, and it generates or modifies his attitudes and his motivation to continue the interaction. The respondent is reacting not only to the interviewer’s behavior as such, however; he is reacting to it as a cue that evokes role behavior already familiar in other contexts, as well as attitudes already formed. Thus, the respondent may behave toward the interviewer as a polite stranger, a hospitable host, a dutiful citizen, a fellow research worker and scientist, or even an obedient servant or intimidated inferior. The inter-

viewer’s expectations and behavior in turn are a mixed product of his own personality and experience, combined with his immediate reaction to what the respondent is doing.

Such a general model of the interview requires additional specification. It requires, among other things, some means for representing the moment-by-moment state of the respondent’s motivation to provide complete and accurate data. The various forces tending, at any point in the interview, to increase or decrease respondent motivation to work in this sense can be well represented by the Lewinian model of the “quasi-stationary equilibrium” (Lewin 1947), in which each factor urging compliance or resistance is depicted as an arrow (force), and the level of motivation is depicted as a horizontal line that is the resultant of the opposing forces. In Figure 2 the factors identified as opposing forces have been taken from current re-search on the interview process (Fowler 1965; U.S. Department. . . 1965a).

Techniques of data collection

It is clear that the conceptualizations of the interview and of the conditions for success have implications for the specific techniques and procedures to be advocated. We have emphasized the conditions of accessibility, cognition, and motivation, and the role relationship of the interviewer and respondent. It follows, for example, that if the dominant problem in a particular case is accessibility, one thinks in terms of ways in which the demand for data might be limited, of records or documents which might evoke associated recollections, and the like.

The technique of data collection by interview can be usefully separated into two main and related aspects—question formulation (developing a measurement instrument) and interviewing itself (using the instrument). The close relationship of these two aspects of technique is most apparent when the same person is performing both functions in rapid succession, as anthropologists do habitually and other social research workers do at least in the early phases of investigation. Even when question formulation and data collection are conspicuously separated in time and space (as in survey research and opinion polling), the connection remains. The instrument both limits and assists the interviewer, and the interviewer necessarily modifies the instrument and the accompanying rules in the very act of using them.

To relate each bit of detailed practical advice on question formulation and interviewing technique to the above theoretical material would be desirable, but lengthy and difficult. The following sections are based, however, on the same model of the interviewing situation, with special emphasis on respondent-interviewer interaction, role taking, and the attainment of accessibility, understanding, and motivation.

The formulation of questions

Of the many ways of describing decisions that the interviewer or research worker must make about the form of questions, two seem particularly useful and important: questions may be open or closed; and they may be direct or indirect in their relationship to objectives. Openness has to do with the form of the response required by a question; an open question invites the respondent to reply in his own words, and a closed question asks him to select, from a series of alternatives, the answer that best approximates his views. The closed question thus controls the form, length, and content of the possible response. The classic example is the trial lawyer’s instruction to the witness on the stand that he “answer Tes’ or ’No.’”

Directness and indirectness have to do with the relationship between the question and the objective, that is, the interviewer’s purpose in asking the question. For example, when a respondent is shown an ambiguous picture and asked to tell a story about its meaning for him, so that the story can subsequently be used to infer the intensity of the respondent’s need for achievement (Atkinson 1958), we consider the question to be indirect. A direct (but ill-advised) question to meet the same objective would ask the respondent how achievement-oriented he considers himself to be.

Open versus closed questions. The intolerant advocacy of one or another of these question forms has been largely abandoned, as Lazarsfeld (1944) long ago proposed it should be. But there remains the problem of how and when to choose one type of question rather than another. At least five considerations are relevant to the choice between open and closed questions: interview objectives, the respondent’s information level, the strength of the respondent’s opinions on the topic, the respondent’s motivation to communicate on the topic, and the interviewer’s initial knowledge of these characteristics of the respondent.

With respect to the objectives of the interview, the general principle is that closed questions tend to be more appropriate for straightforward categorizing of respondents according to their agreement or disagreement with some stated point of view. If the interviewer’s objectives go beyond such description and include explanatory aims, such as discovering the respondent’s particular frame of reference or the process by which he came to his present views, an open question will almost certainly be superior to a single closed one, and perhaps to a combination of several closed questions.

The choice between open and closed questions should also be guided by the probable degree of structuring of the respondent’s opinion on or experience with the topic. To the extent that the respondent has done his cognitive work in advance, so to speak, and has formulated his ideas in terms close to those of the question, the closed form is appropriate, as well as economical in terms of the interviewer’s time and the respondent’s effort. On the other hand, to the extent that the respondent’s thoughts are less structured on the topic in question, the interviewer must assist the respondent to recall, order, and perhaps evaluate his experience.

Definitive research on the motivational advantages of open versus closed questions is yet to be done. The closed question demands less effort from the respondent, as well as less self-revelation, and may therefore be less threatening. To the extent that the closed question incorporates extreme alternative responses, it may also make these extremes more admissible (Kahn 1952; Metzner ’ Mann 1951). On the other hand, the closed question may be restrictive and may also invite an easy, invalid response instead of the more difficult “don’t know.” Finally, the choice between the closed form and the open form should take into account the interviewer’s (or research worker’s) advance knowledge of the situation. If he knows relatively little about the range or terms of response he will encounter, he is obviously in a poor position to formulate closed questions, which are meaningful and successful only when their limited alternatives match the respondent’s experience, vocabulary, and frame of reference. If these conditions are not met, the open question is likely to be preferable.

Direct and indirect questions. A direct question, as we have said, simply asks for information stipulated by the objectives of the interview; there is a congruence between objective and question that is obvious on inspection. An indirect question is not congruent with its objective in the same sense; it is related by some inference or theory. One of the major reasons for using indirect questions is to obtain information about the respondent that he is incapable of providing directly because it is beyond the limits of his conscious insight. Indirect questions are also used to get around unwillingness or inability to report certain kinds of material directly.

The forms of indirection include the use of the third person, since people are sometimes willing to impute to others feelings and opinions that they will not admit as their own. They also include questions that have the appearance of directness but are interpreted as measuring attributes which the respondent is not aware of as being measured (see, for example, the F-scale for the measurement of authoritarianism, as described by Adorno et al. 1950). Finally, indirection may take the form of a purposefully ambiguous stimulus—an ink blot, a picture of uncertain meaning, or the beginning of a sentence which the respondent is asked to talk about spontaneously, to interpret, or to complete.

The indirect approach clearly makes accessible data that would otherwise be inaccessible. The disadvantages of indirection have mainly to do with the problem of validation. Face validity is always risky, but with indirect measures it quickly becomes meaningless.

Apart from the decisions regarding the open or closed form of questions and the direct or indirect means of questioning, there are a number of considerations in question wording and question sequence which bear on the attainment of cognitive understanding, accessibility, and respondent motivation. These are considered briefly below.

Language. The primary criterion in the choice of language is complete and accurate communication of ideas to the respondent. The language of the question must therefore conform to vocabulary that is available to him. This does not mean that the respondent’s colloquialisms or regionalisms should be imitated; such efforts are more often ludicrous than convincing. However, it does mean that the basis for communication between interviewer and respondent consists of their shared vocabulary and that questions should be formulated in these terms.

Frame of reference. Inevitably each respondent interprets and replies to a question in terms of his own experience and of his own present concerns and interests. Even the casual greeting “How are things?” leads one person to think of health, another of financial matters, a third of family affairs. A common but undesirable practice in question formulation is to assume the respondent’s frame of reference is the same as the interviewer’s. More desirable practice is to ascertain the respondent’s frame of reference by additional questioning (“Why is that?” “What did you have in mind?”) whenever his reply to a major question has not made clear his frame of reference. An alternative procedure is to stipulate the desired frame of reference as part of the question itself. Thus the un-qualified “How are things?” might become “How are things going—financially, I mean?” The respondent’s frame of reference may be so powerful, however, that it becomes necessary to allow him to answer the question in his own terms before attempting to impose a different frame of reference on him (Bancroft & Welch 1946).

Avoiding ambiguity. A common fault in question formulation is the inclusion of two or more propositions in such a way that the meaning of the response is ambiguous. For example, consider the survey question “Do you favor or oppose raising real estate taxes for new schools or highways?” A direct response of “Favorℍ or “Oppose” is ambiguous with respect to the referent; it may be tax increases in general, schools, highways, or all of these. Avoidance of such ambiguity is simple: the researcher need only keep in mind that each question should have a single and unambiguous referent and should test, with respect to that referent, a single proposition or point of view.

Recognition versus recall. Where recall is a problem for respondents, questions can usefully be formulated in terms that require only recognition. Experiments in the psychological laboratory have shown that in studies of memory the process of recognition is an easier task for the subject than the process of recall. Moreover, there is some evidence (Kahn 1952) that the presentation of alternatives covering a wide range also increases the likelihood of respondents’ choosing extreme statements and admitting to socially unacceptable opinions or behaviors. Unfortunately, little research has been done to explore the disadvantages of such questions or to document unequivocally their relative reliability and validity under varying conditions of data accessibility.

Sanctioning. The problem of defensiveness can be dealt with in many ways, primarily in the interpersonal technique of interviewing rather than in the formulation of questions. Nevertheless, the wording of questions can facilitate or inhibit the respondent’s admission to facts or opinions that are in some fashion ego-threatening to him. One means of making such material admissible is to include it in the hypothetical range of alternatives presented to the respondent, as suggested above. Another is to build into the question some phrase of reassurance, some reminder of the purpose of the inquiry, or some factual indication that the “unacceptable” is common and, in this context, acceptable.

Leading questions. Avoiding “biased” or “leading” questions is a standard and oversimplified piece of advice in the formulating of questions. By a “leading question” we mean one so formulated that respondents find it easier or more acceptable to answer in one way than another, or to choose one alternative over another. When such questions are formulated inadvertently and the responses are interpreted without regard to the asymmetrical tendency of the questions themselves, the results are biased. The most common example in survey research is the yes–no type of questions, in which respondents who answer “Yes” are then subjected to a long series of additional questions about time, place, reason, and reaction, while respondents who answer “No” are asked no more on the topic. People learn quickly that it is easy and brief to say “No.” Still more crude is the use of question wording that assumes a particular answer and thus forces the respondent to contradict the interviewer in order to formulate a response of his own. The generally accepted principle regarding such questions is that they should be asked only for the purpose of imposing some additional stress on the respondent, for example, in order to identify people who feel strongly enough to assert that the interviewer has assumed wrongly (Smith et al. 1956; Litwak 1956; Kahn ’ Cannell 1957; Kinsey et al. 1948). The rule of thumb justly favors balanced wording, questions designed to equalize the amount of work and the degree of social acceptability regardless of which alternative a respondent chooses, and avoidance of emotionally loaded words and phrases.

Organization and sequence of questions. Issues of organization and sequence of questions, like so much else in interviewing, cannot be settled by generalization but must be resolved in relation to the interview objectives and the characteristics of the population from whom information is sought. Generally, a battery of questions is preferable to a single question, for reasons of both reliability and validity. The more complex the issue and the less tested the approach, the more important it becomes to use multiple questions. When multiple closed questions are used, and the respondent is asked to indicate agreement or disagreement with a stated proposition, it is important to randomize the form of the statements. The purpose of doing so is to randomize the tendency of some respondents to be chronic “yea-sayers” or “nay-sayers” (Couch ’ Keniston 1960). Similarly, if lists of items are being presented to respondents, the order should be varied, since some respondents show a tendency to select the alternative first presented. [SeeRESPONSE SETS.]

The sequence of the topics themselves should be planned to make the total interview experience as meaningful as possible by giving it a beginning, a middle, and an end. More specifically, the early questions should serve to engage the respondent’s interest—without threatening or taxing him before he is really committed to the transaction—and to teach him the kind of task the interview represents. The most demanding of questions might well be placed later in the interview, at a point when the respondent’s commitment can be presumed to have peaked but fatigue has not yet set in.

Interviewing technique

Two major qualifications must be made of any specific set of recommendations respecting interviewing technique. First, the technique of the interviewer is not really separable from the formulation of questions. In almost all situations, the interviewer is in some degree a formulator of questions. In much case study and anthropological work the interviewer formulates and asks questions almost simultaneously; he interacts spontaneously with the respondent. At the other extreme is the injunction to census enumerators to ask each question precisely as worded, to do no improvising of additional questions, and to respond to requests for explanation by repeating the question. As methodological research has shown, there is considerable variation in interviewer technique even under such constrained circumstances (Hansen et al. 1951). The second qualification to recommendations about technique is that the most appropriate advice must differ with the situation—as defined by respondent characteristics, interviewer characteristics, and above all by the task requirements of the interview itself. For example, to develop a close and trusting relationship as a preliminary to a few questions for a school census is ludicrous, whereas to neglect the development of such a relationship in psychiatric diagnosis is equally bad.

It follows that recommendations about technique must be stated in relation to such situational factors. The following recommendations accordingly are based on the situational factors most common in social research (including but not limited to survey research). In such research the respondent is free to give information or refuse it. Moreover, the power of an interviewer over a respondent is limited. He can neither impose formal penalties for nonresponse nor offer a prize for response. The demands on respondent time and effort are significant but not overwhelming, perhaps as little as one-half hour or as much as two hours. The interviewer approaches the respondent as a stranger; the interviewer is identified with some sponsoring agency (university, research institute, or the like) that has at least modest prestige value without possessing legitimate authority or coercive power to demand information. The interviewer must generate and maintain sufficient respondent motivation to meet the interview objectives, and he must direct and control the communication process in the service of those objectives. In general, he does these things by describing the purpose of the interview, treating the respondent with some reasonable show of warmth and interest, indicating directly and approvingly those responses which are relevant and complete, and letting the respondent know also when he is being irrelevant or fragmentary in his answers. These things the interviewer does by building on the specific questions that have been prepared in advance of the interview. He adds supplementary or “probe” questions; he comments on the completeness or inadequacy of a response; he nods, murmurs, and in other ways exerts control over the communications process. Let us consider these behaviors sequentially and specifically.

The introduction of the interview to the respondent is almost wholly dependent upon appeals to extrinsic motivation. An appropriate introduction would include a statement of the purpose of the interview and identification of the sponsoring agency. In some situations this may be introduction enough. If the data requirements and time demands are more than trivial, however, or if the respondent shows curiosity or reluctance, the introduction should also include reassurances to the respondent with respect to the manner in which he came to be selected for interview, the protection and confidentiality which will be accorded his statements, and the specific ways in which his statements will be used. There is a related misunderstanding that the interviewer must be alert for during the early moments of an interview. Many respondents lack experience and knowledge of interviewing for research purposes, but they have no lack of experience at being interrogated for other purposes. To them the appearance of a stranger who wishes to ask questions suggests truant officers, bill collectors, policemen, unwanted salesmen, and a variety of other sources of threat or annoyance. The interviewer will need to look for signs of such misidentification and be prepared to explain and, on occasion, to document his true identity and function.

A remaining problem that arises in the introductory moments of an interview has to do with the ethics of persuasion, particularly in describing the purpose of the interview and the uses to which it will be put. To counsel absolute and complete truthfulness is easy and irreproachable, but there are circumstances in which the effect of such completeness would negate the purpose of the interview. It seems consistent with the ethics of the social sciences in general to resolve such problems by telling the respondent as much as possible without negating the purpose of the interview, by then withholding or generalizing rather than fabricating explanations, and by being uncompromising in letting the respondent know what the interview will require of him.

The development of intrinsic motivation in the interview—emotional and intellectual satisfaction in the process itself—may begin with the introduction, but it matures only as the task and relationship acquire meaning for the respondent. The opportunity to talk to a good listener, to find one’s opinions of serious interest to another person, to see that person making a real and successful effort to understand rather than to evaluate or criticize—these are experiences which are rare for many people and which are intrinsically satisfying. Thus, the interviewer should create and maintain an atmosphere in which the respondent feels that he is fully understood and in which he is safe to communicate fully without fear of being judged, criticized, or subsequently identified and disadvantaged. At the same time, the interviewer must focus attention on the content of the communication, encouraging the respondent to consider each topic as deeply, fully, and frankly as the interview objectives require.

The interviewer’s means for doing these things are not mysterious; they are his elaborations on the primary (prepared) interview questions and in some respects resemble the processes of mutual influence familiar in informal conversation. Richardson and his colleagues (1965) have referred to them as “encouragements, silences, guggles, and interruptions”; Kahn and Cannell (1957) have called them “controlled nondirective probing”; other authors have used still other terms. The specific behaviors proposed for the interviewer by these various authors are more alike than their terminology. The following is a list of some of these behaviors.

Brief expectant pauses.

Brief expressions of understanding and interest:

“I see, um-hm.”

“Yes, I understand.”

Neutral requests for additional information:

“How do you mean?”

“I’d like to know more of your thinking on that.”

“What do you have in mind there?”

“Is there anything else?”

“Can you tell me more about that?”

Echo or near repetition of the respondent’s words:

Respondent: “I’ve taken these treatments for almost six months, and I’m not getting any better.” Interviewer: “You’re not getting better?”

Summarizing or reflecting respondent expressions:

Examples would follow respondent statements, stating the interviewer’s understanding of a key feeling or meaning. Such summaries often begin with phrases like “You feel that” or “You mean that”

Requests for specific kinds of additional information:

“Why do you think that is so?”

“How did that become clear to you?”

“When was that?”

Requests for clarification:

“I’m not clear on that.”

“Could you explain what you mean?”

Repetition of a primary question:

Interviewer: “What kind of work do you do?”

Respondent: “I work at the paper mill.”

Interviewer: “I see. What kind of work do you do there?”

To these specific forms of supplementing the primary questions in an interview must be added explanations, reassurances, and further information about the interview.

The separation of such interviewing techniques from the formulation of the major questions that present the topics of inquiry is in some degree arbitrary, of course. This distinction between primary and supplementary or “probe” questions is perhaps least useful in the two extreme forms of the interview, that is, when the interviewer is formulating all questions on the spur of the moment in a completely unstructured situation, and when the interviewer is absolutely restricted to reading prepared questions from a script or schedule. But most interviews are well within these extremes and involve some mixture of predetermined questions and spontaneous interactive elaborations on them. In most social research of scale, this functional distinction is emphasized by a division of labor; the people who conduct the interviews have usually not developed the basic questions to be asked. In such circumstances, it becomes essential to develop an understanding of the complementary functions of question formulation and elaboration or probing, so that they may in fact complement each other instead of producing a validity-destroying and unintended competition.

Selection and training of interviewers

Evidence for the importance of interviewer selection and training is compelling and of long standing, although much of it is also indirect. It is clear, for example, that a variety of interviewer characteristics—demographic and attitudinal—can affect the interview product. It is less clear under what circumstances and through what chain of events these effects occur. Race and religion have been shown to inhibit responses, particularly when the respondent holds critical views of a minority to which the interviewer apparently belongs (Hyman et al. 1954; Robinson & Rhode 1946). Differences in age and sex between interviewer and respondent have been shown to reduce the communication of some kinds of data (Benney et al. 1956). Moreover, there is evidence that the social class of interviewers is reflected in the data they obtain on political issues (Katz 1942).

The unwanted relationship between interviewers’ attitudes and interview data, which was first described by Rice (1929), has been demonstrated repeatedly (Cahalan et al. 1947; Guest 1947; Ferber & Wales 1952). Stanton and Baker (1942) have demonstrated that similar outcomes may result from an “expectation bias,” that is, the interviewer’s expectations of the respondent. Studies of interviewer personality and its effects are fewer, but one by Richardson (1965), using TAT (Thematic Apperception Test) measures of personality in relation to interviewer performance, suggests that the effective interviewer enjoys people, seeks friendly relations with them, and has insight into the complex of feeling relationships among widely varying types of people.

It is a plausible assumption that most of the characteristics of interviewers that have been shown to create biased data do so via the interviewers’ behavior rather than through the reaction of respondents to some nonbehavioral characteristic. In other words, interviewers make characteristic errors which might be avoided to some extent by appropriate techniques and training (Cannell 1953) and by selection on the basis of intelligence (Guest & Nuckols 1950) and other measurable criteria.[SeeErrors, article onNonsampling Errors.]

As yet there is not enough research on the selection and training of interviewers to warrant making definitive statements. However, it appears that the most important consideration in selection is that the interviewer must be seen by the respondent as being “within range” of communication on the interview topic. This does not imply a matching on education, age, sex, or other characteristics but, rather, suggests that the respondent must perceive the interviewer as having sufficient knowledge and understanding so that effective communication between the two is possible.

Evidence for the effects of interviewer training under varying circumstances is sketchy; opinions are strong—and variable. They range from reliance on brief written instructions (characteristic of much commercial opinion polling and market research) to the proposals of Nadel (1951) and Kluckhohn (1945) that interviewers in anthropological research should be psychoanalyzed as well as intensively trained in more specific ways. There is evidence that training makes a difference, some of it indirect and some direct. Cannell (1953) found that carefully trained interviewers produced results that were relatively free of the class and attitudinal biases reported by other investigators; his research, however, did not include experimental comparisons among interviewers differing in training. Richardson (1965) found that intensive training in field methods produced significant increases in measures of interviewing performance and no significant changes in personality measures (TAT). Individual differences in training effects were large, and the performance of individual interviewers before training was a poor predictor of posttraining performance.

To be effective, a training program should place heavy emphasis on interviewing practice. Role playing, observation by experienced persons, and tape recordings are useful, since they provide the opportunity for immediate feedback, which is a most important aspect of training.

Any article on interviewing in social research written today should be tentative in tone. For a number of reasons, the field is in flux. For one thing, some new ideas have been proposed after a long and relatively static period. Webb, Campbell, and their colleagues (Webb et al. 1966) have argued persuasively for less addiction to interviewing among social research workers and more attention to “unobtrusive measures.” Richardson, Dohrenwend, and Klein (1965) have presented the opposing notions of “stressed” versus “unstressed” interviews in terms that invite research on the appropriate use of differing techniques. Cannell (see U.S. Department ... 1965a; 1965b), Fowler (1965), and others have questioned the earlier emphasis on avoidance of interviewer influence and have begun research on the interviewer’s functions as teacher and reinforcer of appropriate respondent behavior, as well as permissive encourager of conversation.

A second factor that makes for change in interviewing theory and practice is cultural and historical; the violation of privacy for trivial or questionable purposes has brought the rights and roles of interviewers and respondents under some discussion in the United States. It is a poor time for predicting the outcome of a trend that is so new and that may show vastly different forms in different cultures. Nevertheless, it seems likely that increasing sophistication about the reasons for interviewing and the consequences of the interview is a part of technological and industrial development in its contemporary forms. Whether the increasing sophistication of respondents will make the collection of data by interviewing easier, more difficult, or virtually impossible depends on the visible uses and abuses of the interview. In all but its most extreme and indirect forms, the interviewing technique is ultimately dependent upon a societal record of individual protection, respect for the confidentiality of personal data, and relevant and benign use of information in research, industry, and government.

Robert L. Kahn and

Charles F. Cannell

[Directly related are the entriesField work ; Observation . Other relevant material may be found inEthics , article onEthical issues in the social sciences ; Mental disorders, Treatment of , article OnClient-centered counseling ; Projective methods ; Reason analysis ; and in the biography ofSullivan ]


Adorno, Theodor W. et al. 1950 The Authoritarian Personality. American Jewish Committee, Social Studies Series, No. 3. New York: Harper.

Argyris, Chris 1964 Integrating the Individual and the Organization. New York: Wiley.

Asch, Solomon E. (1952) 1959 Social Psychology. Englewood Cliffs, N.J.: Prentice-Hall.

Atkinson, John W. (editor) 1958 Motives in Fantasy, Action, and Society: A Method of Assessment and Study. Princeton, N.J.: Van Nostrand.

Bancroft, Gertrude; and WELCH, EMMETT H. 1946 Recent Experience With Problems of Labor Force Measurement. Journal of the American Statistical Association 41:303-312.

Bartlett, Frederic C. (1932) 1950 Remembering: A Study in Experimental and Social Psychology. Cambridge Univ. Press.

Beezer, R. H. 1956 Research Methods of Interviewing Foreign Informants. Technical Report, No. 30. Washington, D.C.: George Washington Univ., Human Resources Research Office.

Benney, Mark; Riesman, David; and STAR, SHIRLEY A. 1956 Age and Sex in the Interview. American Journal of Sociology 62:143-152.

Bingham, Walter; and Moore, Bruce V. (1931) 1959 How to Interview. 4th ed., rev. New York: Harper.

Booth, Charles et al. (1889-1891) 1902-1903 Life and Labour of the People in London. 17 vols. London: Macmillan.

Boring, Edwin G. (1929) 1950 A History of Experimental Psychology. 2d ed. New York: Appleton.

Cahalan, Don; Tamulonis, Valerie; and Verner, Helen W. 1947 Interviewer Bias Involved in Certain Types of Opinion Survey Questions. International Journal of Opinion and Attitude Research 1, no. 1: 63-77.

Campbell, Donald T. 1957 Factors Relevant to Validity of Experiments in Social Settings. Psychological Bulletin 54:297-312.

Cannell, Charles F. 1953 A Study of the Effects of Interviewers’ Expectations Upon Interviewing Results. Ph.D. dissertation, Ohio State Univ.

Coombs, Clyde H. 1964 A Theory of Data. New York: Wiley.

Couch, A.; and Keniston, K. 1960 Yeasayers and Nay-sayers: Agreeing Response Set as a Personality Variable. Journal of Abnormal and Social Psychology 60: 151-174.

Cronbach, Lee J. 1946 Response Sets and Test Validity. Educational and Psychological Measurement 6: 475-494.

Cronbach, Lee J. (1949) 1960 Essentials of Psychological Testing. 2d ed. New York: Harper.

Ferber, Robert; and Wales, Hugh G. 1952 Detection and Correction of Interviewer Bias.Public Opinion Quarterly 16:107-127.

Fowler, F. J. JR. 1965 Education, Interaction, and Interview Performance. Ph.D. dissertation, Univ. of Michigan.

Freedman, Ronald; Whelpton, Pascal K.; and Campbell, Arthur A. 1959 Family Planning, Sterility, and Population Growth. New York: McGraw-Hill. Gorden, Raymond L. 1954 An Interaction Analysis of the Depth-interview. Ph.D. dissertation, Univ. of Chicago.

Guest, Lester L. 1947 A Study of Interviewer Competence. International Journal of Opinion and Attitude Research 1, no. 4:17-30.

Guest, Lester L.; and NUCKOLS, ROBERT 1950 A Laboratory Experiment in Recording in Public Opinion Interviewing. International Journal of Opinion and Attitude Research 4:336-352.

Gurin, Gerald; Veroff, Joseph; and Feld, Sheila 1960 Americans View Their Mental Health: A Nationwide Interview Survey. Joint Commission on Mental Illness and Health, Monograph Series, No. 4. New York: Basic Books.

Hansen, Morris H. et al. 1951 Response Errors in Surveys. Journal of the American Statistical Association 46:147-190.

Herzberg, Frederick; Mausner, Bernard; and Snyder-Man, Barbara B. 1959 The Motivation to Work. New York: Wiley.

Hildum, Donald C.; and BROWN, ROGER W. 1956 Verbal Reinforcement and Interviewer Bias. Journal of Abnormal and Social Psychology 53:108-111.

Hyman, Herbert H. et al. (1954) 1962 Interviewing in Social Research. Univ. of Chicago Press.

Kahn, Robert L. 1952 A Comparison of Two Methods of Collecting Data for Social Research: The Fixed-alternative Questionnaire and the Openended Interview. Ph.D. dissertation, Univ. of Michigan.

Kahn, Robert L.; and Cannell, Charles F. 1957 The Dynamics of Interviewing: Theory, Technique and Cases. New York: Wiley.

Kaplan, Abraham 1964 The Conduct of Inquiry: Methodology for Behavioral Science. San Francisco: Chandler.

Katona, George 1960 The Powerful Consumer: Psychological Studies of the American Economy. New York: McGraw-Hill.

Katz, Daniel 1942 Do Interviewers Bias Poll Results? Public Opinion Quarterly 6:248-268.

Kinsey, Alfred C. et al. 1948 Sexual Behavior in the Human Male. Philadelphia: Saunders.

Kluckhohn, Clyde 1945 The Personal Document in Anthropological Science. Pages 79-173 in Louis R. Gottschalk, Clyde Kluckhohn, and Robert Angell, The Use of Personal Documents in History, Anthropology and Sociology. New York: Social Science Research Council.

Krasner, Leonard 1958 Studies of the Conditioning of Verbal Behavior. Psychological Bulletin 55:148-170.

Lazarsfeld, Paul F. 1944 The Controversy Over Detailed Interviews—An Offer for Negotiation. Public Opinion Quarterly 8:38-60.

Lenski, G. E.; and Leggett, J. C. 1960 Caste, Class and Deference in the Research Interview. American Journal of Sociology 65:463-467.

Lewin, Kurt 1947 Frontiers in Group Dynamics. Parts 1-2. Human Relations 1:5-41,143-153. → Part 1: Concept, Method and Reality in Social Science, Social Equilibria and Social Change. Part 2: Channels of Group Life; Social Planning and Action Research.

Likert, Rensis 1961 New Patterns of Management. New York: McGraw-Hill.

Litwak, Eugene 1956 A Classification of Biased Questions. American Journal of Sociology 62:182-186.

Mandler, George; and Kaplan, Warren 1956 Subjective Evaluation and Reinforcing Effect of a Verbal Stimulus.Science 124:582-583.

Merton, Robert K.; FISKE, MARJORIE; and KENDALL, PATRICIA L. (1944) 1956 The Focused Interview: A Manual of Problems. Glencoe, 111.: Free Press.

Metzner, Helen; and Mann, Floyd C. 1951 A Limited Comparison of Two Methods of Data Collection: The Fixed-alternative Questionnaire and the Open-ended Interview. Unpublished manuscript, Univ. of Michigan, Institute for Social Research.

Michigan, University of, Survey Research Center 1960 The American Voter, by Angus Campbell et al. New York: Wiley.

Milgram, Stanley 1965 Some Conditions of Obedience and Disobedience to Authority. Pages 243-262 in Ivan D. Steiner and Morton Fishbein (editors), Current Studies in Social Psychology. New York: Holt.

Nadel, Siegfried F. 1951 The Foundations of Social Anthropology. London: Cohen & West; Glencoe, 111.: Free Press. → See especially pages 169-176. Organizational Stress: Studies in Role Conflict and Ambiguity, by Robert L. Kahn et al. 1964 New York: Wiley.

Payne, Stanley L. 1951 The Art of Asking Questions. Studies in Public Opinion, No. 3. Princeton Univ. Press.

Quay, Herbert 1959 The Effect of Verbal Reinforcement on the Recall of Early Memories.Journal of Abnormal and Social Psychology 59:254-257.

Research Methods in Social Relations. Rev. ed. By Claire Selltiz et al. (1959) 1964 New York: Holt.

Rice, Stuart A. 1929 Contagious Bias in the Interview: A Methodological Note. American Journal of Sociology 35:420-423.

Richardson, Stephen A. 1965 A Study of Selected Personality Characteristics of Social Science Field Workers. Pages 328-358 in Stephen A. Richardson, Barbara S. Dohrenwend, and David Klein, Interviewing: Its Forms and Functions. New York: Basic Books.

Richardson, Stephen A.; Dohrenwend, Barbara S.; and Klein, David 1965 Interviewing: Its Forms and Functions. New York: Basic Books. → See especially pages 64-65 and 277-278.

Richardson, Stephen A.; Hastorf, Albert H.; and Dorn-Busch, Sanford M. 1964 Effects of Physical Disability on a Child’s Description of Himself. Child Development 35:893-907.

Riesman, David 1958 Some Observations on the Interviewing in the Teacher Apprehension Study. Pages 266-370 in Paul F. Lazarsfeld and Wagner Thielens, Jr., The Academic Mind: Social Scientists in a Time of Crisis. Glencoe, 111.: Free Press.

Robinson, Duane; and Rohde, Sylvia 1946 Two Experiments With an Anti-Semitism Poll.Journal of Abnormal and Social Psychology 41:136-144.

Salzinger, Kurt; and Pisoni, Stephanie 1960 Reinforcement of Verbal Affect Responses of Normal Subjects During an Interview. Journal of Abnormal and Social Psychology 60:127-130.

Saslow, George et al. 1957 Test-Retest Stability of Interaction Patterns During Interviews Conducted One Week Apart. Journal of Abnormal and Social Psychology 54:295-302.

Shapiro, Sam; and Eberhart, John C. 1947 Interviewer Differences in an Intensive Interview Survey. International Journal of Opinion and Attitude Research 1, no. 2:1-17.

Smith, M. Brewster; Bruner, Jerome S.; and White, Robert W. 1956 Opinions and Personality. New York: Wiley.

Stanton, Frank; and Baker, Kenneth H. 1942 Interviewer-bias and the Recall of Incompletely Learned Materials. Sociometry 5:123-134.

U.S. Departmentof Health, Education, AND Welfare, Public Health Service 1965a Reporting of Hospitalization in Health Interviewing Survey: Methodological Study of Several Factors Affecting Reporting of Hospital Episodes. Prepared by C. F. Cannell et al. Washington: Government Printing Office.

U.S. Department of Health, Education, and Welfare, Public Health Service 1965b Comparison of Hospitalization Reporting. Prepared by C. F. Cannell et al. Washington: Government Printing Office.

Webb, Eugene et al. 1966 Unobtrusive Measures. Chicago: Rand McNally.


Despite the developments in twentieth-century psychological measurements, the interview, the most ancient method for personality appraisal, remains both the most powerful and the least affected by psychometric refinements. The primary appeal of the interviewing method lies in its potential. The interview can provide a broader spectrum of information than any standard procedure and can be focused on classes of information for which no standard procedure is available; entirely in contrast with the sense of a standard procedure, it permits a change in emphasis and an adjustment in observational focus as the goals of appraisal may define or redefine themselves during the course of the interview. The secondary appeal of the interview is its convenience; it can be of any duration. In addition, there is no standard procedure which can be applied in a greater variety of circumstances because the interview requires no equipment, no definite setting, and no response modalities other than verbal; with a telephone it does not even require the physical proximity of the interviewer and the interviewee.

In the appraisal of personality, it is possible to distinguish three kinds of orientation. There is the normative orientation, where the manifestation of a quality or an attribute in a given individual is appraised on a comparative basis, using as reference the manner in which the manifestation may be found in a standard population from which the individual in question may have been drawn; this orientation emphasizes individual differences between people. In addition there is the ipsative orientation; this is the orientation which attempts to determine the relative strength of various qualities within the individual. It emphasizes intraindividual differences which indicate the relative dominance of qualities within individuals, without providing a basis for comparison between individuals with respect to the normative strength of these qualities. There is also the idiographic orientation; this is the orientation which seeks within the individual to identify manifestations or qualities which are distinctive to him but may not be of sufficient general interest or frequency to provide a meaningful basis for either ipsative or normative appraisal.

Regardless of the orientation chosen by the appraiser, it is apparent that the interview may be used in a manner which provides a normative assessment comparable with available psychometric procedures for personality appraisals. The interview is uniquely suitable for determining the relative or intraindividual strength of various motivational components and the relative extent to which the ego employs certain mechanisms of defense. The interview is without peer in the broad screening of atypical idiographic qualities pointing to personality weaknesses which could interfere with proper functioning under certain high-demand situations. While the appeal of the interview remains in its potential, there are situations in which its vulnerabilities make it a second choice to the relatively cumbersome and circumscribed standard psychometric devices. [SeePsychometrics.]

Vulnerabilities of the interview

Reliability and validity

The interview is not an objective procedure. Any given result may not always be reproducible, and in this sense it is unreliable. The results of an interview may not be reproducible because the individual himself has changed, perhaps in some way which is pertinent to the appraisal, but more probably in his perception of the appraisal interview and the manner in which he should respond. He may have become more fearful, more confident, or more opportunistic in the situation. The interviewer, too, may have changed in consequence of the first interview; he may have come to see his task in a different way, or he may have developed biasing feelings about the interviewee.

The literature is rich in studies which show that the interviewer is not only a primary stimulus to the interviewee’s response (Cairns & Lewis 1962; Cohen et al. 1954; Greenspoon 1955; Kanfer & McBrearty 1962; Verplanck 1955; Williams 1964; Zigler & Kanzer 1962); unless the interview is recorded, he is also the fallible observer and variable recorder of the interviewee’s response (Cuadra & Albaugh 1956; Hildum & Brown 1956). He is the interacting participant-observer; his effective role is not only subject to change but is also unknown and ungauged. It is possible, of course, to place restraints on the interviewer’s role, to standardize his questions and his responses, and in this way to approach reproducibility. To the extent to which this standardization is obtained, however, the interview loses its unique potential and assumes both the guarantees and the limitations which characterize psychometric standardization.

The issue of reliability as a challenge to interview appraisals is no more disquieting than the issue of validity. The substance of an interview may not be pertinent to its purpose. This paradox is possible because the purposes of the appraisal to which the interview is applied are often poorly defined. Sometimes the interview must both explore and assess so that, in effect, the interview has the confounding task of defining its own purpose.

Sources of distortion

Once the purposes of the appraisal are known, the pertinence of the content of the interview to these purposes remains a matter of considerable uncertainty. The meaning of the content of interview responses is qualified by many factors, and the array of content which emerges in any interview is biased and selected by many influences (Bolster & Springbett 1961; Dohrenwend & Richardson 1963; Gynther 1957; Jones 1954). The amount of any given kind of responsive material may be truncated by the expressive capacity of the interviewee, by the defensive requirements of the interviewee’s personality, or by the tolerance of the interviewer for material which is threatening to his own defensive needs. In addition to such limitations there are distortions which may be deliberate, as the interviewee attempts to direct his performance, or may be unconscious, when one unrecognized requirement of his personality interacts with another.

There are also response distortions which accrue from the context. These may have their origins in the meaning of the interview situation to the interviewee, in the setting of the interview with its facilitating or distracting stimulus properties, and in the general set which accrues from concurrent life circumstances and may characterize the interviewee at that particular time only. Aside from the defensive requirements of the interviewee’s personality and current influences which affect the con-tent of his responses at the time, the manner in which the interviewee perceives both the interviewer and his responses is subject to perceptual biases or predispositions which may or may not be known by the interviewer. Some of these perceptual characteristics reflect the values, the mores, and the experiences that accrue from the interviewee’s subculture and may result in gross distortions or complete failures in communication that may be beyond the control of the interviewer, if not entirely outside his awareness. For similar reasons, the responses of the interviewee may be grossly misconstrued by the interviewer who, because of his ethnic, social, economic, or educational background, may ascribe a mistaken meaning to responses of the interviewee (Douvan 1956; McArthur 1955; Ort 1952; Rankin & Campbell 1955; Rosen 1956; Siller 1957).

Controlling responses

Although the interview may be reduced to a standard procedure to assure a modicum of reliability and validity, its greatest value and its greatest challenge may be found when it is used as a process for securing and interpreting information. The process is unique in the particular sense that the information which is secured at any one phase may have been selectively sought on the basis of interpretations accruing from information provided by preceding phases. Because of this, most of the current research on the interview is concerned either with the task of selectively controlling the interviewee’s response (Cairns & Lewis 1962; Cohen et al. 1954; Ferguson & Buss 1960; Heller et al. 1963; Hildum & Brown 1956; Kanfer & McBrearty 1962; Sarason & Campbell 1962; Verplanck 1955; Williams 1964; Zigler & Kanzer 1962) or with the task of interpreting the implications of these responses (Hathaway 1956; Shrauger & Altrocchi 1964; Taft 1955). Thus the concern with standardizing the selection and interpretation of interview material is relatively minor. The major emphasis is on controlling and comprehending.

The problem of controlling the course of the interview and comprehending the data it generates is a problem of developing adequate descriptive concepts and applying the relevant available theories of human behavior. The ways in which such concepts and theories must function can be appreciated in part by a review of the kinds of data that are available from the interview and that presumably could be pertinent to the task of appraising the personality of the individual.

Descriptive information

The verbal content alone usually contains a wealth of information which dwarfs the interpretive resources of the interviewer and may greatly surpass the initial purposes of the interview. As reflected in the tradition of oral examinations, the interview is an excellent device for sampling the individual’s knowledge. Beyond this, it may provide an exhibit of his verbal facility and his conceptual resources. The content of the interview can provide information concerning the intentions, the attitudes, and the values of the respondent with respect to specific topics. At a more personal level, however, the interview may be used to provide information about the individual’s personal circumstances, his background experiences, accounts of how he or others have responded to various experiences and situations, and many other kinds of information which are either directly expressive of the individual’s personality or are in some indirect way clues to qualities which are relevant to an understanding of his personality. The verbal content of the interview may also describe the feelings of the individual, including the prevailing moods that characterize his general outlook, the varieties of moods that he can experience, and the intensity and labile sensitivity of these mood changes.

Nonverbal responses

In most situations the verbal content is the object of primary interest, but other response modalities or qualities (Domey et al. 1964; Secord & Muthard 1955) provide data of substantial relevance to an appraisal of the personality (Giedt 1955). There are such obvious features as voice intensity, pitch, and adaptive modulation to accompany content and intent (Kramer 1963). There is a host of accessory motor responses. The mobility of the eyes and face, the involvement of the hands and arms, postural adjustments (Dittman 1962; Ekman 1964; Krout 1935; Luriia 1932; Sainsbury 1955), tentative movements of various parts of the body which appear to be a part of the expressive effort, and yet other movements which obviously detract from the manifest communication—such as tics, grimaces, lisps, stammers, and respiratory irregularities— may tell a story that is not necessarily congruent with the manifest verbal content. To be included in this class of information are the temporal features of the individual’s behavior, the latency of his responses, the duration (Anderson 1960; Phillips et al. 1961), the number of pauses (Witten-born et al. 1962), and response variability (Boomer 1963; Dibner 1956; Eldred & Price 1958; Mahl 1956; Pope & Siegman 1962). Beyond this there are those accessory response qualities which must be described as physiological (Malmo et al 1956; Shagass & Malmo 1954). These would include flushing or blanching of the face, perspiration, temperature changes of the hands, tremors, the pounding pulse of excitement, the lassitude of fatigue or boredom, and the respiratory irregularities of tension, with shallow or irregular breathing, deep sighs, or arrested inspirations.

Situational context

In addition to the responses themselves, the information from the interview can and should include a description of the stimulus context in which the responses occurred. The most obvious features of the stimulus context would include the questions and the responses verbalized by the interviewer. Nevertheless, it is apparent from several inquiries that the mood (Bolster & Springbett 1961) and the values and expectations (Koltuv 1962; Shrauger & Altrocchi 1964) of the interviewer, as well as the way he feels about the interviewee and the content of the interview, also affect the responses of the interviewee and are a part of the stimulus context (Heller et al. 1963).

In addition, the context includes the manner in which the purposes of the interview are presented of the interviewee’s perception are referred to as are expected, that are prudent, or that are permissible for him. Attempts to control these features of the interviewee’s perception are referred to as the preliminary structuring of the interview and are, in effect, an attempt on the part of the interviewer to instruct the interviewee with respect to what he should expect and how he may respond. There are also the obvious facts concerning the circumstances of the interview, including privacy, assurance of confidentiality, associative distractions within the setting, and incidental distractions during the course of the interview. These all may have a bearing on the interviewee’s responses and are, therefore, a part of the total body of information which may be supplied by the interview and may be relevant to the appraisal of the personality.

Response contexts

The interview also provides information of a more subtle nature. Most important of these subtler considerations is the sequential context in which a given response occurs. As anticipated in prior statements, where the interview lies in the course of the individual’s antecedent and anticipated subsequent experiences is of obvious relevance; where a response lies in the preceding and succeeding verbal content within the interview may be even more informative. The association may be explicit and unambiguous; more often, however, it is symbolic. It is not sufficient to note the sequential context of verbal responses. The antecedents and the consequences of an autonomic response may reveal more about the personality of the individual than many pages of verbal transcription. A similar implication may be found in the temporal variations in response, e.g., delays, pauses, or accelerations.

The relevant sequential context is not to be found in the responses of the interviewee alone, however; as research reports indicate, and as may be well revealed in almost any recorded interview, where the interviewee’s response lies in the sequential context of the interviewer’s responses, verbal or nonverbal, overt or covert, has a basic, almost invariable qualifying significance (Greenspoon 1955; Kanfer et al. 1960; Moos 1963; Sarason 1962; Verplanck 1955). For the purposes of appraising personality these sequential features cannot safely be disregarded.

The interpretive use of information

Frames of reference

Assuming that it is possible to guide, if not direct, the genesis of information during the course of the interview to secure the kinds of information that are desired, there remains the question of how this information may be interpreted in order to serve the purposes of personality appraisal. Such a discussion requires some attempt to define and distinguish what is meant by personality, and it should be observed that personality may be viewed from either of two standpoints.

From the more popular standpoint, personality refers to the social stimulus characteristics of the individual, i.e., how people perceive and respond to him. This obviously includes his appearance, the quality of his performance in various situations, and the manner in which he responds to the presence of others. Since these qualities change rather slowly, their appraisal has some value because it may be used to anticipate how the individual will perform in the future and how people may be expected to respond to him in future situations. The principle that appears to govern the accuracy of such predictions assumes that the more similar a future situation is to a present one, the more the individual’s future personality qualities will resemble his presently perceived qualities.

There is a second, more sophisticated view of personality, however; in this view, personality refers to the dynamic economy of the individual’s motivational adjustment and to both the relative strengths of his basic drives and the skills he has acquired in expressing these drives in the society in which he lives. In response to the demands and opportunities of his society there appears to be a gradual shift in the relative strength of his motives and a somewhat more fluid, although tardy, change in the skills or defensive mechanisms whereby the individual seeks to serve his needs as he negotiates in his society. This lifelong process of acquisition and change tends to be prompt and orderly during years of physical development, but like most other changes it diminishes with advancing age. These changes and acquisitions are sometimes referred to as the socialization of the personality.

There are important advantages in an appraisal of the relative strengths of the individual’s motives and an identification of the preferred modes of defense (Weintraub&Aronson 1964) which he employs in the service of these motives. The advantage lies in the general premise that the concepts involved are applicable to virtually all situations. The dynamic theory of personality is sufficiently well developed that once the strengths of motives are assessed and the preferred coping mechanisms are identified, a transcendent understanding of the individual emerges, making it possible to anticipate qualities of the individual’s behavior in situations that are quite different from those in which these features of the personality were first known and appraised. On this basis, attempts may be made to anticipate the manner in which the individual would respond in new employment situations, to social threats that he has never known, and to various hypothetical future situations, including life with a prospective spouse. Although such predictions can be made on the basis of a personality appraisal that is built upon an understanding of the dynamic economies of the individual’s adjustment, the accuracy of such predictions seems to be highly variable. Some interviewers provide much more accurate predictions than others (Bieri 1955; Campbell et al. 1964; Cronbach 1955; Gage & Cronbach 1955; Koltuv 1962; Landfield 1955; Lundy 1956; Shrauger & Altrocchi 1964; Wittich 1955; Yonge 1956). The various studies of such accuracy serve primarily to confirm that it is a variable which rests upon the individual appraiser; the studies do not clearly illuminate the characteristics of background, temperament, or training that identify the appraisers who will be most accurate. It appears, however, that experience may be an important factor.

The interpretive use of interview information may be seen as guided by several different kinds of considerations. Much of the verbal content of the interview has an obvious relevance to the purposes of appraisal, giving it a quality of face validity which permits its direct interpretive application at a practical, common-sense level.

Many of the physiological manifestations in the interview have an implication for stress—that is, tension, fear, or anxiety—which is almost universal in its significance and for interpretive purposes may be assumed to cut across most national, linguistic, and other cultural differences. Various motoric signs of excitement or indifference and indications of lassitude or fatigue may also have a rather general interpretive significance, although there are subcultural differences in the meaning of the temporal aspects of verbal response. For example, the temporal aspects of responses of Scotch descendants in the Appalachian hills cannot be interpreted on the same basis as they might for certain groups of Spanish-Americans.

Normative factors. Most of the information that is generated in the course of the interview cannot be safely interpreted for personality appraisal without some consideration of normative factors. The meaning of the response is greatly qualified by who makes the response, and from this standpoint the most important features of an individual may be defined in terms of the general characteristics of the social or subcultural group or groups with which he may be identified. This is not only true for the meaning of the manifest content of the verbal material but can be true for the symbolic or allusional meaning as well. (This is not to deny that the symbolic significance of some material appears to be almost universal in its implications.)

The normative significance of interview material may present difficulties at several different levels. For example, the way in which language is used by preparatory school students, jazz musicians, or some Negro groups can be seriously misleading if the interviewer is not aware of the peculiarities of denotation and connotation that characterize the group. Perhaps more seriously misleading than the distinctive use of words are the values that are ascribed to certain phrases, gestures, or other communicative expressions. A response that may be viewed as an indication of affectionate acceptance in one social group may be regarded as an expression of rejection or contempt in another.

Almost all of the content of the interview is invested with value implications. The values, attitudes, and affective involvements that may be safely inferred, however, require an application of the appropriate normative reference. For example, in some communities knowledge about the experience and affairs of other members is freely shared (Zborowski 1951), and the requirements for privacy may be moderate. In other communities certain kinds of experiences, sentiments, and personal knowledge are not shared and the requirements for privacy and confidentiality are extreme, so that the interviewee’s loss of privacy is tantamount to a serious loss of face and can be a threatening prospect. There are also important differences in the meaning of an experience. For example, individuals from some groups are tolerant of their blunders and regard themselves as one with the world so that they can safely refer to their own errors as things “that happened.” Persons from other groups, however, must regard a blunder as solely their own responsibility, a deficit in their life balance sheet which can never be fully compensated for. Such distinctions that relate to social norms must be a part of the appraiser’s awareness, because a specific point of view when manifested in one subculture may indicate an appropriate product of socialization, while the same point of view in another subculture may indicate a crippling neurotic aberration.

Crucial interviewer skills. Thus it may be acknowledged that much of the content of the interview has an obvious meaning; that some of the affective expressions, particularly those involving the autonomic nervous system, have an almost universal significance; and that appropriate normative references add substantially to the interviewer’s interpretive acuity and greatly reduce the risk of gross misperceptions. Nevertheless, the great burden of the interpretive use of the information provided by the appraisal interview rests on the inferential and deductive skills of the interviewer. Although the powerful use of inference and deduction in personality appraisal may be described as a skill, it rests upon a fairly definitive body of fact and theory concerning the development of the human personality and its functional economy in the interacting requirements of the living organism in its social and physical environment.

The appraisal of the social stimulus characteristics of the individual makes relatively superficial demands on the inferential and deductive acuity of the interviewer. In contrast, however, the assessment of the strength of the individual’s drives and the identification of the coping mechanisms employed in commerce with society are primarily products of complex inference and deduction and are subject to error of judgment. Nevertheless, it is the complex inferential appraisal of these qualities which offers the greatest promise of providing a defensible basis for anticipating the responses of the individual in new situations.

Psychosexual development

For the most part, the assessment of the relative strength of drive is concerned with attempts to establish the levels of psychosexual development that the individual may have reached in the metamorphosis of his personality and the readiness with which he may regress to earlier, more primitive levels in response to frustration (Fenichel 1945). Level of psychosexual development is primarily a psychoanalytic concept and is based upon the premise that the motivational requirements of the individual evolve through a successive series of stages as the individual’s social and interpersonal role is modified from infancy to maturity (Munroe 1955). Where the individual lies on this hypothetical continuum of motivational maturation can often be safely inferred from the manner in which the individual handles himself in the interview, the way he sees others, the role he plays vis-à-vis others, the significance he ascribes to various personal experiences, and the way in which he responds to both the demands of his social environment and his inclination toward affective expression. Whether psychoanalytic theory can always be relied upon to provide an index to the various strengths of the individual’s motivational requirement has not been established with scientific rigor, but it seems reasonable to suggest at this time that such a theoretical approach does provide the interviewer with powerful and incisive questions that can guide his inferences and sharpen his deductions so that in the course of his appraisal some of the very best answers may be generated.

Other aspects of motivation

Ordinarily, most individuals who are the subject of personality appraisal are part of a middle-class society with its traditional emphasis on achievement (Douvan 1956), and often the aims of appraisal must emphasize achievement striving with its compensatory and competitive features. The dynamic views of personality developed by Alfred Adler (Munroe 1955) and his associates emphasize compensatory striving in response to real or imagined weaknesses relative to the environment. These provide an array of postulates and tenets which may be of great value to the appraiser, particularly as he attempts to formulate useful inferences and provide specific deductions concerning achievement predispositions in the personality of the interviewee. [SeeAchievement motivation; Individual psychology; and the biography ofAdler.]

In addition to these general views concerning the cultural significance of the personality and the dynamic motivational implications of its various manifestations, there is a substantial and rapidly growing body of research-based information which may guide the interviewer as he attempts to assess the motivational requirements of the individual (Feinberg 1953; Handel 1965; Helper 1958; Jourard 1957; Koch 1956; Manis 1958; Payne & Mussen 1956; Peck 1958).

Attempts to infer motivational characteristics of the individual must be made during the course of the interview. The alert and resourceful appraiser will also find or create numerous opportunities to confirm or refute the deductive implications of his inferences. It may be noted that the development of knowledge concerning the individual, like the development of all other integrated bodies of information, requires the reciprocal use of inference and deduction, with the revisions of inference based on attempts at observational verification of the deductions. [SeeMotivation, article onhuman motivation.]

An assessment of the individual’s motivational requirements is not enough, however; it is equally important to identify the devices which the individual characteristically employs in expressing or gratifying these motives. These devices contribute in large measure to the social stimulus value of his personality and are the basis for the description of what is ordinarily known as character (Fenichel 1945) and disposition (Mendelson 1960). It is also relevant to predict the individual’s response in novel situations, particularly those where the probable motivational components may be anticipated. The capacity of the individual to sublimate primitive drives, the extent to which he evades the energy requirements of challenges by resorting to denial, the projective ascription of culpable motives to others, and the extent to which he displaces affect, particularly hostility, from the stimulating source to some innocuous object are all a part of the coping devices or mechanisms of defense (Haan 1964; Weintraub & Aronson 1964) that may characterize the individual’s attempts to handle the insistent requirements of his drives.

As a final comment on the interpretive use of interview information in the appraisal of the personality, it should be noted that in most situations the interviewer is not dependent on the interview information alone. Usually there is supplementary information describing background, education, work history, a statement of current circumstances, the opinions or interests of the referral source, and sometimes the availability of psychological test results. Any such information can have substantial value, either in drawing inferences or in verifying them by providing other sources of data for testing their deductive implications.

Methods of interviewing

The foregoing review of some of the aims of the interview appraisal, its vulnerabilities, the kind of data on which it is based, and the rational basis by which these data may be applied for the purposes of appraisal can lead to an appreciation of the problems of gathering information by an interview procedure. As familiar forms and conventions of the appraisal interview are reviewed in the following paragraphs, it will be noted that there is some rough correspondence between the manner in which the information is gathered and the manner in which it may be interpreted.


There have been numerous attempts to develop a standardized interview procedure. The primary aim of such standardization was to increase reliability and validity. In some instances, however, the standardization reflected a distinctive view of personality as well. One of the most important standardized interview procedures was developed by Chapple (1949), an anthropologist who worked with various collaborators on the premise that the important clues in personality were not to be found in the content of the interview but were revealed in characteristic patterns of interaction. Interaction was examined primarily in terms of its temporal aspects, and an interaction chronograph was developed for this purpose. The procedure is one where the observer presses buttons signaling various events and the machine provides a series of scores reflective of such qualities as initiative, tendency to interrupt, and dominance as they may characterize the verbal behavior of the participants in the interview. Several investigators have examined the properties of these interaction scores. Notable among them are G. Saslow, J. D. Matarazzo, and their collaborators, who have provided reassuring indication of reliability and potential validity (Matarazzo et al. 1956; Phillips et al. 1961).

Wittenborn and his colleagues (1962) developed a standard interview procedure which incorporates the principle of scores based on temporal characteristics but is different in its basic orientation in the sense that it comprises a series of standard questions calculated to involve various motivational qualities selected from Henry A. Murray’s system of needs (see Explorations in Personality 1938). The temporal scoring requires no machine, however, and can be reliably provided by the practiced interviewer counting silently to himself and making marks to indicate periods of speech and silence. The procedure provides excellent interobserver reliability, and indications of validity may be found in the fact that it generates scores that are correlated with symptom ratings and are sensitive to antidepressant medication.

A further approach to the standard interview is provided by Gleser, Gottschalk, and Springer (1961). This procedure requires only that the interviewee talk about a life experience for an uninterrupted period of five minutes. The recorded interviews are transcribed, and the scoring and analysis are provided by counting the occurrence of psychologically pertinent themes in each communication unit, e.g., one hundred words. The themes that are counted vary with the purposes of the interview, and a detailed application of an anxiety scale has been described.


Often when the purpose of the interview is highly specific, arbitrary restrictions of the content of the interview may be justified. This would obviously be true in a census interview or a simple poll of buying habits, political opinions, or voting intent. In a similar way the taking of testimony is often a highly restricted interview. In many situations the employment interview is equally restricted, but it is possible that such restriction misserves its purpose and may obscure more than it reveals about the candidate.

It is possible to conduct an interview in a somewhat structured manner, thereby guaranteeing that certain types of information are included without being narrowly restrictive or resorting to a question—answer type inquiry (Cline 1955). This is often done by being certain that a standard set of provocative questions is introduced during the course of the interview. Psychiatric case history procedures may follow such a pattern. It is also possible to give the interview an effective structure if the interviewer has well in mind a standard set of questions for which he desires answers. On this basis he can be selectively attentive to the material that is pertinent to his interest. As the interview progresses to its close, he can guide the discussion over remaining areas of pertinence so that answers to his questions may be secured without making the informant fully aware of the selective nature of his interest and without biasing the informant’s responses. Thus a standard body of information is secured without the stilted or arbitrary employment of a standard inquiry. This procedure can be developed to yield scores that have a high order of reliability and validity. For example, Wittenborn and his associates have developed a standard set of questions which may be answered by a social worker on the basis of her interview with the family informants of mental hospital patients. These scores, based on a factor analytic reduction of data, have been found to be useful in identifying patients who will respond favorably or unfavorably to a given treatment (Wittenborn & May 1966).

Unstructured interviews

There are also unstructured approaches to the interview. These are, in effect, exploratory in the sense that their content and course are expected to be reflective of the interest, needs, and moods of the respondents and not determined by the a priori interests of the interviewer. The use of such an approach is usually based on the premise that the affectively significant material will find expression in a tolerant, acceptant, protected atmosphere (Rogers 1942). If affective requirements and the huge freight of content which they carry are considered to be important for the purposes of the appraisal, then the acceptant, unstructured approach has much to commend it; the interviewer’s task becomes one of encouraging the flow of such material without precipitating an accelerated cathartic reaction which can generate a subsequent defensive refractory phase often known as “resistance.” In general, however, the unstructured approach does little to heighten the vigilance of the respondent and generates very little defensive resistance. The permissiveness of the interview situation is never complete. Even the most nondirective interviewer is somewhat selective in the material that elicits from him expressions of sympathy, assurance, understanding, interest, or acceptance. Since such feedback communication from the interviewer tends selectively to reinforce certain kinds of materials and not others, the un-structured interview is never without control, and the literature showing that the response of the interviewee is greatly influenced by his perception of the interviewer is quite substantial.

In some interviews, particularly those employed in psychotherapy, an attempt is made to secure a free association of the respondent’s thoughts and feelings. Presumably in this situation the course of the content is guided primarily by the interplay of the individual’s needs and his defensive responses to them. Since it appears that the interviewee requires guiding, coaching, encouragement, and correction while attempting to associate freely, a free association may not be entirely “free” and may rest largely on what the interviewer’s theoretical persuasion or personal viewpoint leads him to consider to be “free.”

Of increasing significance is the use of technical aids in securing a record of the interview. Complete recordings which may be subsequently reviewed, transcribed, or scored are now commonplace. In some situations the use of sound movies is considered to be practical, and closed-circuit television permits the simultaneous use of more than one observer. These devices may have their greatest value in obviating problems of reliability and validity which are always a part of the participant observer situation, both because of the practical limits of the interviewer’s observational acuity and the defensive limitations that are generated in him as he responds to the interviewee and is involved in the emerging flow of information.

Much is possible within an interview procedure, but there are barriers to an invariably satisfactory use of the interview as a method of personality appraisal. The greatest barrier appears to lie not in the interview itself but in the fact that right answers to wrong questions do not lead to satisfaction. In the interview method of personality appraisal, it is probable that failure is more often due to the fact that the wrong questions were asked than to the possibility that the wrong answers to these questions may have been proffered. This is the criterion problem. If as much care were taken in formulating the questions that the appraiser is expected to answer as the appraiser ordinarily exercises in providing his answers, the interview as a method of personality appraisal could be viewed with greater confidence. Unfortunately, the same interview, if not the same responses within the interview, must often be used not only to formulate the incisive questions but to provide the definitive answers. Such confounding may lead to unverifiable results because in the appraisal interview, as in all other inquiries, one cannot use the same data both to generate an inference and to test its implications. The difficulties that prompt these reservations are not insuperable. While the traditional use of standard psychometric devices may have attained its maturity, the use of the interview, guided by increasing knowledge of the personality and aided by the datareducing capacity of modern computers, may now be approaching the threshold of its great development.

J. R. Wittenborn

[Directly related are the entriesPersonality measurement; Projective methods. Other relevant material may be found inPersonality; Psychoanalysis, article ontherapeutic methods; andPsychometrics.]


Anderson, C. W. 1960 The Relation Between Speaking Times and Decision in the Employment Interview. Journal of Applied Psychology 44:267-268.

Anisfeld, moshe; Munoz, Stanley R., and Lambert, Wallace E. 1963 The Structure and Dynamics of the Ethnic Attitudes of Jewish Adolescents. Journal of Abnormal and Social Psychology 66:31-36.

Bieri, James 1955 Cognitive Complexity–Simplicity and Predictive Behavior. Journal of Abnormal and Social Psychology 51:263-268.

Bolster, B. I.; and Springbett, B. M. 1961 The Reaction of Interviewers to Favorable and Unfavorable Information. Journal of Applied Psychology 45:97-103.

Boomer, Donald S. 1963 Speech Disturbance and Body Movement in Interviews. Journal of Nervous and Mental Disease 136:263-266.

Cairns, Robert B.; and Lewis, Michael 1962 Dependency and the Reinforcement Value of a Verbal Stimulus. Journal of Consulting Psychology 26:1-8.

Campbell, Donald T. et al. 1964 Varieties of Projection in Trait Attribution. Psychological Monographs 78, no. 15.

Chapple, Eliot D. 1949 The Interaction Chronograph: Its Evolution and Present Application. Personnel 25: 295-307.

Cline, Victor B. 1955 Ability to Judge Personality Assessed With a Stress Interview and Sound-film Technique. Journal of Abnormal and Social Psychology 50:183-187.

Cohen, Bertram D. et al. 1954 Experimental Manipulation of Verbal Behavior. Journal of Experimental Psychology 47:106-110.

Cronbach, Lee J. 1955 Processes Affecting Scores on

“Understanding of Others” and “Assumed Similarity.” Psychological Bulletin 52:177-193.

Cuadra, Carlos A.: and Albaugh, William P. 1956 Sources of Ambiguity in Psychological Reports. Journal of Clinical Psychology 12:109-115.

Dibner, Andrew S. 1956 Cue-counting: A Measure of Anxiety in Interviews. Journal of Consulting Psychology 20:475-478.

Dittmann, Allen T. 1962 The Relationship Between Body Movements and Moods in Interviews. Journal of Consulting Psychology 26:480.

Dohrenwend, Barbara S.; and Richardson, Stephen A. 1963 Directiveness and Nondirectiveness in Research Interviewing: A Reformulation of the Problem. Psychological Bulletin 60:475-485.

Domey, Richard G.; Duckworth, James E.; and Morandi, Anthony J. 1964 Taxonomies and Correlates of Physique. Psychological Bulletin 62:411-426.

Douvan, Elizabeth 1956 Social Status and Success Strivings. Journal of Abnormal and Social Psychology 52:219-223.

Ekman, Paul 1964 Body Position, Facial Expression, and Verbal Behavior During Interviews. Journal of Abnormal and Social Psychology 68:295-301.

Eldred, Stanley H.; and Price, Douglas B. 1958 A Linguistic Evaluation of Feeling States in Psychotherapy. Psychiatry 21:115-121.

Explorations in Personality: A Clinical and Experimental Study of Fifty Men of College Age, by Henry A. Murray et al. 1938 London and New York: Oxford Univ. Press.

Feinberg, Mortimer R. 1953 Relation of Background Experience to Social Acceptance. Journal of Abnormal and Social Psychology 48:206-214.

Fenichel, Otto 1945 The Psychoanalytic Theory of Neurosis. New York: Norton.

Ferguson, Donald C.; and Buss, Arnold H. 1960 Operant Conditioning of Hostile Verbs in Relation to Experimenter and Subject Characteristics. Journal of Consulting Psychology 24:324-327.

Gage, N. L.; and Cronbach, Lee J. 1955 Conceptual and Methodological Problems in Interpersonal Perception. Psychological Review 62:411-422.

Giedt, F. Harold 1955 Comparison of Visual Content and Auditory Cues in Interviewing. Journal of Consulting Psychology 19:407-416.

Gleser, Goldine C.; Gottschalk, L. A.; and Springer, K. J. 1961 An Anxiety Scale Applicable to Verbal Samples. Archives of General Psychiatry 5:593-605.

Greenspoon, Joel 1955 The Reinforcing Effect of Two Spoken Sounds on the Frequency of Two Responses. American Journal of Psychology 68:409-416.

Gynther, Ruth A. 1957 The Effects of Anxiety and of Situational Stress on Communicative Efficiency. Journal of Abnormal and Social Psychology 54:274-276.

Haan, Norma 1964 The Relationship of Ego Functioning and Intelligence to Social Status and Social Mobility. Journal of Abnormal and Social Psychology 69: 594-605.

Handel, Gerald 1965 Psychological Study of Whole Families. Psychological Bulletin 63:19-41.

Hathaway, Starke R. 1956 Clinical Intuition and Inferential Accuracy. Journal of Personality 24:223-250.

Heller, Kenneth; Myers, R. A.; and Kline, L. V. 1963 Interviewer Behavior as a Function of Standardized Client Roles. Journal of Consulting Psychology 27: 117-122.

Helper, Malcolm M. 1958 Parental Evaluations of Children and Children’s Selfevaluations. Journal of Abnormal and Social Psychology 56:190-194.

Hildum, Donald C.; and Brown, Roger W. 1956 Verbal Reinforcement and Interviewer Bias. Journal of Abnormal and Social Psychology 53:108-111.

Jones, Edward E. 1954 Authoritarianism as a Determinant of First-impression Formation. Journal of Personality 23:107-127.

Jourard, Sidney M. 1957 Identification, Parent-cathexis, and Self-esteem. Journal of Consulting Psychology 21:375-380.

Kanfer, Frederick; and Mcbrearty, John F. 1962 Minimal Social Reinforcement and Interview Content. Journal of Clinical Psychology 18:210-215.

Kanfer, Frederick et al. 1960 Experimental Modification of Interviewer Content in Standardized Interviews. Journal of Consulting Psychology 24:528-536.

Koch, Helen L. 1956 Attitudes of Young Children Toward Their Peers as Related to Certain Characteristics of Their Siblings. Psychological Monographs 70, no. 19.

Koltuv, Barbara B. 1962 Some Characteristics of Intrajudge Trait Intercorrelations. Psychological Monographs 76, no. 33.

Kramer, Ernest 1963 Judgment of Personal Characteristics and Emotions From Nonverbal Properties of Speech. Psychological Bulletin 60:408-420.

Krout, Maurice H. 1935 Autistic Gestures: An Experimental Study in Symbolic Movement. Psychological Monographs 46, no. 4.

Landfield, Alvin W. 1955 Self-predictive Orientation and the Movement Interpretation of Threat. Journal of Abnormal and Social Psychology 51:434-438.

Lundy, Richard M. 1956 Assimilative Projection and Accuracy of Prediction in Interpersonal Perceptions. Journal of Abnormal and Social Psychology 52:33-38.

Luriia, Aleksandr R. 1932 The Nature of Human Conflicts; or, Emotion, Conflict and Will: An Objective Study of Disorganization and Control of Human Behaviour. New York: Liveright.

McArthur, Charles 1955 Personality Differences Between Middle and Upper Classes. Journal of Abnormal and Social Psychology 50:247-254.

Mahl, George F. 1956 Disturbances and Silences in the Patient’s Speech in Psychotherapy. Journal of Abnormal and Social Psychology 53:1-15.

Malmo, Robert B.; Smith, A. A.; and Kohlmeyer, W. A. 1956 Motor Manifestation of Conflict in Interview: A Case Study. Journal of Abnormal and Social Psychology 52:268-271.

Manis, Melvin 1958 Personal Adjustment, Assumed Similarity to Parents, and Inferred Parental-evaluations of the Self. Journal of Consulting Psychology 22:481-485.

Matarazzo, Joseph D.; Saslow, G.; and Guze, S. B. 1956 Stability of Interaction Patterns During Interviews: A Replication. Journal of Consulting Psychology 20:267-274.

Mendelson, Myer 1960 Psychoanalytic Concepts of Depression. Springfield, 111.: Thomas.

Moos, Rudolf H. 1963 The Retention and Generalization of Operant Conditioning Effects in an Interview Situation. Journal of Abnormal and Social Psychology 66:52-58.

Munroe, Ruth L. 1955 Schools of Psychoanalytic Thought. New York: Dryden.

Ort, Robert S. 1952 A Study of Role-conflicts as Related to Class Level. Journal of Abnormal and Social Psychology 47:425-432.

Payne, Donald E.; and Mussen, Paul H. 1956 Parent-Child Relations and Father Identification Among Adolescent Boys. Journal of Abnormal and Social Psychology 52:358-362.

Peck, Robert F. 1958 Family Patterns Correlated With Adolescent Personality Structure. Journal of Abnormal and Social Psychology 57:347-350.

Phillips, Jeanne S. et al. 1961 Relationships Between Descriptive Content and Interaction Behavior in Interviews. Journal of Consulting Psychology 25:260-266.

Pope, Benjamin; and Siegman, Aron W. 1962 The Effect of Therapist Verbal Activity Level and Specificity on Patient Productivity and Speech Disturbance in the Initial Interview. Journal of Consulting Psychology 26:489.

Rankin, Robert E.; and Campbell, Donald T. 1955 Galvanic Skin Responses to Negro and White Experimenters. Journal of Abnormal and Social Psychology 51:30-33.

Rogers, Carl R. 1942 Counseling and Psychotherapy. Boston: Houghton Mifflin.

Rosen, Bernard C. 1956 The Achievement Syndrome: A Psychocultural Dimension of Social Stratification. American Sociological Review 21:203-211.

Sainsbury, P. 1955 Gestural Movement During Psychiatric Interview. Psychosomatic Medicine 17:458-469.

Sarason, Irwin G. 1962 Individual Differences, Situational Variables, and Personality Research. Journal of Abnormal and Social Psychology 65:376-380.

Sarason, Irwin G.; and Campbell, James M. 1962 Anxiety and the Verbal Conditioning of Mildly Hostile Verbs. Journal of Consulting Psychology 26:213-216.

Secord, Paul F.; and Muthard, John E. 1955 Personalities in Faces: IV. A Descriptive Analysis of the Perception of Women’s Faces and the Identification of Some Physiognomic Determinants. Journal of Psychology 39:269-278.

Shagass, Charles; and Malmo, Robert B. 1954 Psychodynamic Themes and Localized Muscular Tension During Psychotherapy. Psychosomatic Medicine 16: 295-314.

Shrauger, Sid; and Altrocchi, John 1964 The Personality of the Receiver as a Factor in Person Perception. Psychological Bulletin 62:289-308.

Siller, Jerome 1957 Socioeconomic Status and Conceptual Thinking. Journal of Abnormal and Social Psychology 55:365-371.

Taft, Ronald 1955 The Ability to Judge People. Psychological Bulletin 52:1-23.

Verplanck, William S. 1955 The Control of the Content of Conversation: Reinforcement of Statements of Opinion. Journal of Abnormal and Social Psychology 51:668-676.

Weintraub, Walter; and Aronson, H. 1964 The Application of Verbal Behavior Analysis to the Study of Psychological Defense Mechanisms. 2: Speech Pattern Associated With Impulsive Behavior. Journal of Nervous and Mental Disease 139:75-82.

Williams, Juanita H. 1964 Conditioning of Verbalization: A Review.Psychological Bulletin 62:383-393.

Wittenborn, J. R.; and May, Philip R. A. (editors) 1966 Prediction of Response to Pharmacotherapy. Springfield, 111.: Thomas. → See especially Chapter 7 on “Factors Which Qualify the Response to Iproniazid and to Imipramine,” by J. R. Wittenborn.

Wittenborn, J. R. et al. 1962 A Standard Interview as an Aid to Assessment. Journal of Nervous and Mental Disease 134:353-360.

Wittich, John J. 1955 The Generality of the Prediction of Self Reports. Journal of Consulting Psychology 19:445-448.

Yonge, K. A. 1956 The Value of the Interview: An Orientation and a Pilot Study. Journal of Applied Psychology 40:25-31.

Zborowski, Mark (1951) 1955 The Children of the Covenant. Pages 352-374 in David C. McClelland (editor), Studies in Motivation. New York: Appleton. → First published in Volume 29 of Social Forces.

Zigler, Edward; and Kanzer, Paul 1962 The Effectiveness of Two Classes of Verbal Reinforcers on the Performance of Middle- and Lower-class Children. Journal of Personality 30:157-163.


The therapeutic interview, like any interview, is a structured social situation involving two persons who communicate with each other primarily by means of a shared language. Although communication by verbal symbols is emphasized, it may also be mediated to some extent by bodily movements, tone of voice, rate of speech, inflection, etc. The therapeutic interview differs from other interview situations in its purpose. It is designed to influence the interviewee’s feelings, attitudes, or beliefs so as to produce a reduction or amelioration of intrapsychic and/or interpersonal conflict. While the therapeutic interview has evolved as a clinical instrument of major importance in psychiatry, psychoanalysis, clinical psychology, and psychiatric social work, there are significant implications for the study of interpersonal processes in a wide variety of two-person situations.

Contrasts between therapeutic and other forms of interviews

The purposes of an interview may vary widely. The situation is commonly regarded as being under the control of one person (the interviewer), who elicits information, expressions of opinion or belief, attitudes, etc. from the second person (the interviewee). Some of the many aims may be to obtain information that the interviewee may possess (as in legal proceedings); to learn his opinions, attitudes, and views on given topics (as in a press interview); to appraise his mental or physical qualifications (as in an employment interview); to gather data for research purposes (as in opinion polling); or to study his mental functioning (as in a psychodiagnostic interview). In all of these situations, the interviewee furnishes the interviewer with something which the latter presumably needs for a purpose that is usually, although not necessarily, understood by the interviewee. At times (in research studies, for example), it may be desirable or even necessary to disguise the true purpose of an interview. The data elicited from the interviewee are typically considered to be under his conscious control. He is free to give information or to withhold it, although some interviews may be specifically designed to gather data that the interviewee is either reluctant to reveal (as in a crime-detection interview) or of which he is not consciously aware (as in a psychodiagnostic interview).

Although the dividing lines are not always sharp, the therapeutic interview differs from the foregoing situations in important respects. Perhaps the most significant difference is that the therapeutic interview is always undertaken with the explicit understanding that the person to be benefited by the interaction is the interviewee, not the interviewer. The interviewer functions in the role of an expert who, by virtue of special training and experience, attempts to provide help in solving a problem in living or in alleviating an emotional disturbance for which the other person consults him. The role relationship, therefore, is one of a professional person to a client who expects to pay for the professional service he receives. It resembles, but is not identical with, the physician–patient relationship. A major difference lies in the role assigned to the patient. In the medical model, the patient passively receives the ministrations of the physician. In contrast, the psychotherapeutic interview generally assigns to the patient a rather active role: he is encouraged by the therapist to become a collaborator and to retain full responsibility for the conduct of his life. In important respects, the relationship is one between equals, one of whom possesses superior knowledge. Following general usage and for lack of better words, the terms “therapist” and “patient” are employed in this article.

Structure of the therapeutic relationship

Central to the conception of the therapeutic interview is the development of an interpersonal relationship between patient and therapist. The patient usually approaches the therapist by presenting a “problem” for which he desires help. Often he is anxious, depressed, troubled, or otherwise in distress. By accepting the patient for “treatment,” the therapist implicitly (although rarely explicitly) promises relief. Usually he does not give direct advice but approaches the problem indirectly. The therapeutic value of a single interview is rarely tangible; rather the ameliorative influence evolves from repeated contacts between therapist and patient. Thus the single therapeutic interview is a link in a chain of meetings often extending over considerable periods of time.

The meetings are structured with regard to time, place, financial arrangements, and the like. Furthermore, the character of the therapist’s role behavior is rigorously defined and adhered to. While being attentive, maintaining a friendly interest and a nonjudgmental attitude, and conveying respect, the therapist attempts to remain detached and emotionally uninvolved. His verbal communications are sparing and confined to occasional comments, questions, clarifications, and “interpretations.”

The patient, on the other hand, is free to talk about any topic he desires; indeed, he is fully encouraged to express his feelings, attitudes, and concerns. It is his responsibility to introduce topics for discussion and to maintain verbal communication. Thus there results a one-sided personal relationship within a highly impersonal framework, which is unlike any other human relationship in that it fosters an atmosphere facilitating the largest possible degree of self-revelation and disclosure on the part of the patient. The patient is assured by the therapist of the confidentiality of all communications.

Dynamics of the therapeutic interview

The therapeutic interview is designed to bring about “corrections” in the patient’s life experience. Through a variety of techniques that are the province of various systems of psychotherapy (including psychoanalysis), the therapist creates a climate in which the patient has an opportunity to develop greater trust in another person. To accomplish this, he has to unlearn patterns of feelings and behavior that he developed in childhood to cope with early interpersonal difficulties. For example, he may learn that he tends to provoke others into rejecting him, or he may come to realize that he tends to dominate people. These patterns are often exceedingly intricate and refractory to change. Great skill and patience are, therefore, prerequisites for the therapist. In his work he is greatly aided by the patient’s unconscious tendency to recreate in any interpersonal situation—and particularly in the therapeutic interview—those early life experiences which have remained problematic and troublesome. This tendency is encompassed by Freud’s concept of the transference, which, when properly dealt with by the therapist, becomes the most potent therapeutic force.

The model of the psychoanalytic situation has been chosen to delineate dynamic events in the therapeutic interviews, although other forms of psychotherapy conceptualize the transactions between patient and therapist in different terms.

The arrangements of the therapeutic interview create a “tilted relationship” (Greenacre 1954) in which the communicative flow tends to be directed from the patient to the therapist. The setting capitalizes on important aspects of all human interactions: (1) Whenever two people are repeatedly alone together, some sort of positive or negative emotional bond will develop between them. The depth and intensity of this bond will be partly influenced by the frequency and the length of periods they are in each other’s company. (2) In everyone there is a need for human “contact,” presumably rooted in the earliest mother–child relationship. The basic need is for sensory contact (warm touch of another body), but it later extends to other sense modalities. (3) The emotional bond develops particularly rapidly if (a) the two persons are alone together, so that the feelings and spontaneous expressions of one person are not deflected by a group, and (b) the patient is troubled and comes to the other person for help, the other person having the characteristics of an expert or an authority. (4) Since the patient is admittedly in need of help and the situation is structured in such a way that the therapist is relatively passive, nonparticipating, and more clearly aware of the kinds of transactions to be expected in a situation of this kind, the relationship does not become one of “mutual warming” rather the patient develops a receptive–expectant (dependent) attitude toward the therapist. The very structure of the therapeutic setting is conducive to a reinstatement of the child–parent relationship; that is, conditions are created in which the patient can take advantage of the tendency inherent in everyone to regress to more primitive, childlike modes of relating to important authority figures.

On the other hand, the tendency to regress is opposed by the realistic aspects of the patient–therapist relationship, which is and always remains a relationship between two adults. Irrespective of the patient’s tendency to distort his perception of the therapist and to treat him as a parental figure, the fact remains that the therapist is a real person. His maturity, fair-mindedness, reasonableness, and tact are characteristics that pertain to the present, not the past.

This dual aspect of the relationship, aided by the therapist’s interpretations, brings about a process that in psychoanalysis is called ‘splitting of the ego.‘ Thus, the patient is forced to re-experience and to some extent re-enact the past in the presence of the therapist (that is, the emergence of transference feelings is encouraged), but he is urged at the same time to become a dispassionate observer of these processes and to view them, in alliance with the therapist, as anachronistic occurrences that have largely lost their utility in the present. Accordingly, the patient eventually learns to identify with the therapist as a representative of adult reality and to abdicate his own allegiance to his neurotic impulses and patterns that had dominated him in the past. In this way the patient achieves mastery and conscious control over hitherto ego-alien impulses and strivings, or, stated otherwise, by identifying with the therapist’s mature ego, the patient gradually acquires strength of his own (Freud 1940).

Furthermore, by assuming an attitude of neutrality, benevolence, and interest, the therapist gradually wins the patient’s confidence. However, while he does not punish, neither does he gratify the patient’s excessive and often contradictory needs. The inevitable result is that sooner or later the patient will experience frustration, because the experiencing and the verbal expression of his wishes and expectations are stimulated, but his wishes are not gratified. Frustration is accompanied by impatience, anger, resentment, and hostility, which tend to be directed against the therapist as the person who has become the instigator of the patient’s significant feelings and emotions. However, the patient’s expectations and wishes rarely express themselves directly and in clearly recognizable form. More typically, they are disguised, rationalized, and distorted in a variety of ways. In other words, the patient unwittingly fights against their recognition.

The manner in which these expectations and wishes express themselves follows the patterns the patient has learned since early childhood in dealing with frustrating situations. For example, he may directly attack the therapist for alleged lack of interest, passivity, and the like. Or, he may try to ingratiate himself by submitting to the therapist for the purpose of winning his favor and acceptance. When all of these maneuvers fail, his anger and hostility against the therapist may be openly experienced. If the feelings become too strong or painful, the patient may become tempted to break off the therapy.

The therapist’s approach to this chain of events takes essentially the following course: First, he attempts to demonstrate to the patient that there are feelings emerging in the therapeutic situation that the patient is fighting to keep out of awareness; that is, the therapist identifies the patient’s resistances, which are manifested by blocking, talking about trivia, and other diversionary tactics. Second, he tries to identify the character of the resistance and the feelings against which it is directed. Once this interpretive task is accomplished, and as the patient progressively gains more awareness of his behavior, he gradually abandons the resistance and instead experiences the painful affect, which thus far has been warded off. As this process repeats itself, increasingly deeper layers of the patient’s defensive system are mobilized, dealt with, and restructured. The study of the patient’s dreams and other fantasy materials often proves very valuable in this connection. Gradually the patient improves, and the returns from continuing therapy diminish. Therapy then comes to a close, but only after the patient’s dependency on the therapist has been resolved. The process of “weaning” the patient from what he has come to experience as a helpful relationship is frequently very difficult, opposed as it is by strong unconscious forces within the patient (and sometimes within the therapist as well).

Theoretical differences

Although theforegoing account represents a schematic account of the process of therapeutic work in terms of the psychoanalytic model, it should be emphasized that there is no unanimity of opinion concerning the essential ingredients of the therapeutic action in the two person relationship. Similarly, there are marked differences beween therapists with respect to concepts, techniques, and goals in the therapeutic interview. These divergences derive from different conceptions of personality, personality development, and personality change. That is, differences in the therapist’s attitudes and techniques are (at least in part) dictated by the theoretical orientation to which he subscribes (Ford & Urban 1963).

In general, theories of therapeutic action may be divided into those that view the action primarily in terms of the personal relationship between the two participants and those that rely more heavily on technical procedures followed by the therapist to bring about a specific objective. The former group, among whose exponents are the phenomenologists (like Carl Rogers, the existentialists, and adherents of Daseinanalyse), stress the patient’s self-realization, self-actualization, and freedom to grow, which are believed to be promoted by an interpersonal relationship in which the therapist attempts to understand and to empathize with the patient. Once the therapist’s unconditional positive regard and empathic understanding are communicated to the patient (Rogers 1961), he is then able to free himself from the defensive and self-defeating and constricting tendencies that impede his free personality functioning.

The second group of theorists (exemplified by therapists subscribing to one or another form of learning theory) conceptualize the role of the therapist as that of a person who directly and actively attacks and conditions the maladaptive patterns of the patient by deliberately encouraging him to make responses antagonistic to anxiety-provoking situations. Freudian psychoanalysis seems to occupy an intermediate position between these two poles.

A common element in all therapeutic approaches, as Frank (1961) has pointed out, is the existence of a firm belief on the part of the patient in the superior ability of the therapist as a professional helper, and a similarly strong conviction on the part of the therapist that he is able and willing to work with the patient toward the achievement of a particular therapeutic goal. It certainly seems plausible that a therapeutic relationship in which the patient experiences the therapist’s acceptance, respect, dedication, and positive regard leads to the development of a high degree of interpersonal trust. Once this condition has been created, the therapeutic relationship may come to serve as a vehicle for mediating learning experiences of various kinds. These may consist in the extinction of maladaptive patterns of interpersonal relatedness and the learning of more mature and satisfying and less conflictual ones. The therapeutic interview thus constitutes a model for learning experiences in the interpersonal realm. The patient learns how to learn and how to benefit from interpersonal experiences. Clearly, to have therapeutic value, any learning in the therapy situation must generalize to the patient’s life outside of therapy. Much remains to be learned about this process of translation.

The therapist as participant and observer

The therapist operates within a complex social field, and it is one of the unique aspects of his role that he is both participant and observer (Sullivan 1954). What the therapist learns about the patient’s feelings, attitudes, moods, symptoms, conflicts, difficulties in living, etc., is filtered through and influenced by the patient’s relationship to the therapist. The therapist, as an objective observer, gathers his data by participating with the patient in a personal relationship, albeit a highly specialized one. He must alternate between the roles of observer and participant. As observer he gathers data that hopefully can be turned to therapeutic advantage, but the data come to him only because he is a participant. As participant, too, he attempts to influence the character of the interaction for the benefit of the patient. He is more likely to be successful in this attempt when the patient’s feelings have been mobilized in the situation. At such times, the social field has become fluid and more amenable to modification. In the absence of feeling on the patient’s part in the here and now, the therapist’s messages are perhaps chiefly didactic. It has been said that the patient needs an experience, not an explanation. Accordingly, the essence of the therapeutic interview is the patient’semotional experience with the therapist—what the therapist comes to mean to him on a feeling level.

Limited objectivity of clinical data

It is important to note that there can be no purely “objective” data in an interpersonal situation. To be sure, the patient’s verbalizations and other modes of communication are or can become objective data. But the patient’s state of mind, the meaning of his feelings, as communicated more or less directly to the therapist, must be inferred. In so doing the therapist relies on his clinical experience and theoretical knowledge, but perhaps more importantly on empathy and introspection. Thus, the therapist himself is the most sensitive instrument of observation. However, a concomitant of this process is the observer’s subjectivity, which constitutes a potential pitfall in objective research.

Research and methodological considerations

Since communication between patient and therapist is mediated primarily by linguistic symbols, a first requirement for objective research is to devise conceptual tools that permit the investigator to abstract and quantify relevant aspects of the verbal interchange. Obviously, there are numerous ways of accomplishing this end, and the measures the investigator decides upon are as noteworthy for what they leave out as for what they include. The selection is dictated by theoretical as well as practical considerations, and in a sense it represents a prejudgment of what is important to measure. A system of analysis, therefore, is not a neutral yardstick, and the descriptive measures representing its yield are accordingly circumscribed.

During the past twenty years a large number of content-analysis systems have been developed— some applicable to patient communications, some to therapist communications, some to both. (For references, see Strupp 1962; Auld& Murray 1955; Marsden 1965). Some systems are anchored in a particular theory of psychotherapy, others are theoretically neutral. Ideally, one would like a system that is highly objective and equally sensitive to therapist activities irrespective of theoretical orientation. But these requirements are difficult to meet and perhaps even mutually exclusive.

Of the technical problems inherent in measuring the content of the message exchanged between patient and therapist, probably none is more difficult than the assessment of latent motives and meanings. A large part of the difficulty arises from the fact that (1) linguistic symbols are inadequate to deal with the phenomena to be described, defined, and measured, and (2) psychotherapy as a technique relies heavily on verbal symbols to bring about changes in the affective processes of another person.

In recent years, transactions in therapeutic interviews have been studied from the vantage points of many different theories and by means of a variety of techniques. Apart from the various theories of psychotherapy proper (Ford & Urban 1963), other new departures include communication theory (Ruesch & Bateson 1951); sociology (Lennard & Bernstein 1960); anthropological linguistics (Pittenger, Hockett & Danehy 1960); kinesics, i.e., the study of body motions having communicational value (Dittmann & Renneker 1963).

Technique and the therapist’s personality

Technique in the interview may be viewed as a theory translated into action. This translation, however, is not made in a vacuum, nor are techniques as a rule slavishly applied. The translation is made by the therapist, who, as a sensitive participant–observer, is responsive to the patient’s attitudes, moods, verbal and nonverbal communications, etc. A number of studies have taken the therapist’s communications in actual interviews as a point of departure and investigated various correlates attributable to therapist variables (theoretical orientation, level of experience, etc.). Such studies have been facilitated by the advent of sound recording techniques, combined with a greater willingness on the part of therapists to tolerate external observers who admittedly represent an intrusion into the privacy of a unique and confidential human relationship. From the standpoint of research operations, the development of a variety of systems permitting the quantification of the verbal (and, to some extent, the nonverbal) communication content of the messages exchanged between patient and therapist has rendered research more feasible. In addition to studies dealing with the naturally occurring events in psychotherapy, there have been experimental or quasi-experimental investigations of the therapist’s contribution. While such work is more rigorous and permits a wider sampling of patient and therapist variables, its relevance (validity) to actual psychotherapy remains an open question (Conference on Research in Psychotherapy ... 1962).

At present, it is impossible to assess the nature of the therapist’s personal influence in any form of psychotherapy, and it is one of the important objectives of research to elucidate the effects of this variable.


The primary purpose of research is to attain increasingly objective and specific information about the most effective utilization of the patient-therapist relationship for maximum therapeutic gain. This statement implies the need for extending knowledge concerning the precise nature of the psychological influence exerted by the therapist in interaction with a particular patient, including data about the modes of interaction between the two participants, the technical operations employed by the therapist, and the effects of other social, cultural, and psychological factors impinging upon the therapeutic process. The danger of invading the privacy of the therapeutic relationship is a special problem in psychotherapy research.

Modern therapeutic interviewing attempts a systematic and self-conscious manipulation of variables in a human relationship and has thus become, or is striving to become, a scientific discipline. The therapist, for his part, is an applied scientist. In this role, he is an observer of the patient’s attitudes, feelings, and patterns of behavior, particularly as they are expressed toward him. He tries to be aware of the impact of these variables upon him as a person and as a participant observer; he notes the immediate or long-range effects of his therapeutic interventions (the therapeutic method) as well as his attitudes, feelings, and behaviors upon the patient; and he is engaged in the process of refining his observations and specifying the nature of his influence. In this endeavor, the psychotherapist (as an applied scientist) joins hands with the psychologist (as a “basic” scientist).

Hans H. Strupp

[Other relevant material may be found inMental disorders, treatment of; Psychiatry; Psychoanalysis, article ontherapeutic methods.]


Alexander, Franz 1950 Analysis of the Therapeutic Factors in Psychoanalytic Treatment. Psychoanalytic Quarterly 19:482–500.

Auld, Frank Jr.; and Murray, Edward J. 1955 Content Analysis Studies of Psychotherapy. Psychological Bulletin 52:377–395.

Bellak, Leopold; and Smith, M. Brewster 1956 An Experimental Exploration of the Psychoanalytic Process. Psychoanalytic Quarterly 25:385–414.

Conference on Research in Psychotherapy 1958 1959 Research in Psychotherapy: Proceedings. Edited by Eli A. Rubenstein and M. B. Parloff. Washington: American Psychological Association.

Conference on Research in Psychotherapy 1961 1962 Research in Psychotherapy: Proceedings. Edited by Hans H. Strupp and L. Luborsky. Washington: American Psychological Association.

Dittmann, allen T.; and Renneker, Richard 1963 Kinesic Research and Therapeutic Processes: Further Discussion. Pages 140–160 in Symposium on Expression of the Emotions in Man, New York, 1960, Expression of the Emotions in Man. New York: International Universities Press.

Ford, Donald H.; and Urban, Hugh B. 1963 Systems of Psychotherapy: A Comparative Study. New York: Wiley.

Frank, Jerome D. 1959 Problems of Controls in Psychotherapy as Exemplified by the Psychotherapy Research Project of the Phipps Psychiatric Clinic. Pages 10–26 in Conference on Research in Psychotherapy, 1958, Research in Psychotherapy: Proceedings. Edited by Eli A. Rubenstein and M. B. Parloff. Washington: American Psychological Association.

Frank, Jerome D. 1961 Persuasion and Healing. Baltimore: Johns Hopkins Press.

Freud, Sigmund (1940) 1949 An Outline of Psychoanalysis. New York: Norton. → First published in German as Abriss der Psychoanalyse.

Fromm-Reichmann, Frieda 1950 Principles of Intensive Psychotherapy. Univ. of Chicago Press.

Gill, Merton; Newman, Richard; and Redlich, Fredrick C. 1954 The Initial Interview in Psychiatric Practice. New York: International Universities Press.

Glad, Donald D. 1959 Operational Values in Psychotherapy. New York: Oxford Univ. Press.

Greenacre, Phyllis 1954 The Role of Transference: Practical Considerations in Relation to Psychoanalytic Therapy. Journal of the American Psychoanalytic Association 2:671–684.

Harper, Robert A. 1959 Psychoanalysis and Psychotherapy: 36 Systems. Englewood Cliffs, N.J.: Prentice-Hall.

Janis, Irving L. 1958 The Psychoanalytic Interview as an Observational Method. Pages 149–181 in Gardner Lindzey (editor), Assessment of Human Motives. New York: Rinehart.

Lennard, Henry L.; and Bernstein, Arnold 1960 The Anatomy of Psychotherapy. New York: Columbia Univ. Press.

Marsden, Kenneth G. 1965 Content-analysis Studies of Therapeutic Interviews: 1954 to 1964. Psychological Bulletin 63:298–321.

Menninger, Karl A. 1958 Theory of Psychoanalytic Technique. New York: Basic Books.

Mowrer, Orval H. (editor) 1953 Psychotherapy: Theory and Research. New York: Ronald Press.

Plttenger, Robert E.; Hockett, Charles F.; and Danehy, John J. 1960 The First Five Minutes: A Sample of Microscopic Interview Analysis. Ithaca, N.Y.: Martineau.

Rogers, Carl R. 1957 The Necessary and Sufficient Conditions of Therapeutic Personality Change. Journal of Consulting Psychology 21:95–103.

Rogers, Carl R. 1961 On Becoming a Person: A Therapist’s View of Psychotherapy. Boston: Houghton Mifflin.

Ruesch, Jurgen; and Bateson, Gregory 1951 Communication: The Social Matrix of Psychiatry. New York: Norton.

Strupp, Hans H. 1962 Patient-Doctor Relationships: The Psychotherapist in the Therapeutic Process. Pages 576–615 in Arthur Bachrach (editor), Experimental Foundations of Clinical Psychology. New York: Basic Books.

Sullivan, Harry Stack 1954 The Psychiatric Interview. New York: Norton.

Wolberg, Lewis R. 1954 The Technique of Psychotherapy. New York: Grune.