The term “interview” allows for several definitions. In this article, all forms of the socioscientific interview – also called survey – are dealt with. The interview – along with content analysis and observation – is one of the three basic empirical instruments of data collection. It is defined as a planned and systematic situation in which knowledge is gained by human subjects. The aim is to generate individual answers, which in their entirety lead to the clarification of a (scientific) question. This method has been in use since the end of the eighteenth century. Significant impulses for its establishment and development in the field of mass communication were set by several mainly sociological studies. These were conducted in both scientific (by a group around Paul F. Lazarsfeld) and applied (by George Gallup) contexts. Socio-scientific interviews are conducted to elicit subjective evaluations, judgments, attitudes or interpretations of an issue. Interviews are also conducted to measure both cognitive contents and behavior that cannot be observed. In mass communication, interviews are central to measuring the process, evaluating, and studying the effects of mass media use, as well as to examine the state and trend of public opinion.
The interview’s basic idea is to give respondents the opportunity to express their personal views. In principle, this method is comparable to everyday communication. In an interview, however, this communication follows certain defined rules. Using specific linguistic stimuli, the respondents are motivated to react in a specific way (stimulus–response approach) – either in their own words or by using multiple-choice answers. The ruling principle is to measure communication by communication. This is at the same time the great strength and the great weakness of the method: the interview’s strength lies in its ability to measure a person’s mental state; its weakness is the respondent’s reactivity, which is caused by the communicative situation and the interaction between interviewer and respondent. An interview has two prerequisites: first, respondents must be articulate; second, they must be able and willing to provide information on the subject in question. Thus, they have to be conscious of their answers, or at least become conscious of them during the interview.
The Situation Of The Interview
It is the social situation of the interview that sets this method apart from other forms of data collection. Interviewer and interviewee meet – either in person or via the telephone – in a seemingly everyday situation. The interaction, however, is not the same as in a normal conversation, where social roles are usually not so clear-cut. As opposed to natural social situations, it is necessary to avoid establishing social standards that create conformity. This, however, is hardly possible because besides the actual question – which is considered the stimulus – several other parameters of the situation (e.g., mood, atmosphere etc.) can exert significant side-effects. Every question initiates a cognitive process in the interviewee that leads to an answer. More often than not, this process is related to certain personal assessments or judgments. Thus, instead of simply recalling attributes (i.e., reacting to the stimulus), the respondent acts according to the situation. That means the respondent perceives the parameters of the situation, rates them according to perceived possible consequences and expectations, and finally chooses the reaction that – in his or her opinion – best serves the goals.
Depending on the subject in question, specific situational parameters can have a “distorting” effect. Such a bias may stem from the interviewer or the respondent. An answer is described as biased when it is influenced by such a parameter and not just by the respondent’s attitude. Biasing attributes – the so-called interviewer effect – can have various forms. They may be apparent: such as the age, gender, education and profession, and ethnic or religious affiliation of the interviewer. Psychosocial attributes, however, can be of influence as well – the expectations and attitudes of the interviewer, for instance. In addition, the conduct of the interview itself, including verbal conditioning, eye contact, body language, and gestures, may have influence. To minimize such effects interviewers must be thoroughly trained. Some authors suggest that interviewers should be matched closely to respondents in vital personal characteristics wherever possible. As far as the respondents are concerned, particularities of the real situation – such as the presence of a third person – have effects on the answering behavior. Another form of bias is due to so-called response sets that are independent of any content. Most important here are acquiescence (the respondents try to give the seemingly accommodating answer) or the tendency for evasion. Another important influencing factor is the phenomenon of social desirability, which can be both person-related and situation-related. Social desirability may distort answers because of a perceived but unspoken norm. None of these three phenomena can be avoided completely. Intelligent wording of the questionnaire, however, can at least minimize the problem.
Standardization Of The Interview
There are many types of interviews. They can be distinguished on several dimensions: the degree of standardization, the number of persons interviewed at the same time, the chosen method, and the frequency of administration. With regard to standardization, one can differentiate between completely standardized, semi-standardized, and nonstandardized interviews. Thus, the roles of both interviewer and interviewee are framed by the type of interview. In the case of a completely standardized interview, the course of the interview is precisely predetermined. Both the questions and the multiple-choice answers are accurately phrased. The interviewer is obliged to follow the instructions accurately. Thus, all respondents receive the same stimulus so that their responses are based on exactly the same conditions. Attempts are made to control or eliminate any factors that might influence the responses, such as unwanted additional stimuli from the interviewer, or changes during the communication situation caused by the presence of a third person or similar. Semi-standardized or semi-structured interviews are often called guided interviews or, because they are often conducted by experts, expert interviews. This type of interview is less well defined. The course of the interview is planned, but the interviewer is not bound to follow it strictly. The basic questions are written down beforehand, but only to remind the interviewer to raise the topics. He or she can decide upon the question order and the wording – just as the situation and the interviewee demand. Thus, he or she is able to adapt to the situation whenever necessary. At the same time, the respondents have ample scope to develop their answers and to contribute to the interview situation – and the research question. The advantage is that such an interview leaves room for spontaneity, unforeseen notions, subjective wording, and the individual behavior of the respondents. Despite the increased scope they allow, semi-structured interviews are comparable to one another. Non-standardized interviews are not structured at all and more closely resemble a “normal conversation” than other types of interviews. Only the essential subject is defined beforehand. The course and content are dependent on the situation and on the persons participating in the interview.
These three basic forms of interview can be conducted in different situations and in different ways. In principle, all forms can be used for individual as well as for group interviews. The standardized group interview, however, is restricted to filling out a standardized questionnaire. It is important to note that the social situation is different when an interview is conducted in a group situation. In addition to the relationship between interviewer and interviewee, interactions among group members can influence the situation dramatically. In principle, the same interview can be conducted only once with each respondent (cf. the problem of reactance of the respondents and its effects on the reliability). Therefore, most interviews are conducted as one-time cross-sectional surveys. Nevertheless, some issues, particularly those on development and diversification, require repeated measurement. There are two basic approaches: (1) repeated interview on the same subject with different respondents (trend or tracking study) and (2) repeated interview of the same respondents at different times, which is called a panel.
Interview Modes
The traditional form is the oral interview conducted personally face-to-face. Depending on the research goal and the interviewees, other forms of communication are possible: via telephone or in a written form. Within these types of interview, there are different alternatives, particularly because of the increasing use of computers and the Internet (online interview). The choice of method depends first and foremost on the kind of information required and on the subject. The degree of standardization also plays an important role. The higher the degree, the simpler is the use of supplementary media such as paper, telephone or computers. Whereas guided interviews can also be conducted via telephone, this channel does not apply to open qualitative interviews. The choice of method is also determined by the budget, the time schedule of the study, and the information relevant to choosing the sample. In a face-to-face interview, the interviewer and respondent face each other. The interviewer asks the questions and records the respondent’s answers. Thus, in a personal interview the social situation has the strongest effect. Face-to-face interviews offer a wide range of possibilities. The social presence allows for complex questions and those requiring mutual trust. Apart from linguistic stimuli, the interviewer is able to present nonverbal, visual, and optical stimuli (e.g., lists, brands, logos, products, objects or other supporting material). At the same time, however, the interviewer is particularly responsible for the correct conduct of the interview. It is difficult to control the interview situation from outside; this involves both the presentation of questions and the logging of answers. Measures to secure the quality of the sample – particularly relevant in quantitative surveys – are also highly dependent on the interviewer.
In recent years, the computer has become increasingly important for conducting surveys, mostly for quantitative interviews. If computers replace paper and pencil questionnaire in personal interviews, the technique is called Computer-Assisted Personal Interviewing (CAPI). Normally, in a CAPI interview the respondents do not sit in front of a computer monitor, but the interviewer enters the answers for them on a keyboard (usually a laptop or hand-held computer). Some answers, however, can be keyed in by the respondents themselves to ensure privacy. The use of computers in a face-to-face situation offers both advantages and disadvantages. The advantages are the technical processing of the questionnaire and the accurate presentation of the questions. The disadvantages are a consequence of the medium itself – the possibility of technical failure, and the presence of the technical device itself altering the communication situation. In telephone interviews, interviewer and respondent talk to each other on the phone. Interviewers can fill out paper questionnaires or the interview can be conducted as a “Computer-Assisted Telephone Interview” (CATI). As telephone interviews are conducted swiftly, they are mainly used to measure spontaneous reactions to current events, to represent a current public opinion or a trend in the population. The telephone can also be used for expert interviews, particularly to facilitate recruitment and appointment. Apart from these aspects, the main advantages of the telephone are the speed of data collection and the low costs. A disadvantage, however, is that it is not possible to use any stimuli other than linguistic ones. Complexity as well as questionnaire length has to be adapted to the interview situation. Thus, the range of application is somewhat limited. In the case of international comparative studies, it has to be considered that in many countries outside the western world, telephone penetration is far from 100 percent.
The self-administered questionnaire differs most from personal and telephone interviews by the absence of an interviewer. The social situation is thus completely different, defined only by the respondents and their social environment. That means that the use of written interviews depends on the literacy of the country in which the interview is conducted. The questionnaire, together with a cover letter, can either be sent via mail to a sample of respondents or it can be delivered to them in person. In the case of an audience survey at a public event, interviewers may also hand over questionnaires directly to potential respondents. It is also possible to insert a questionnaire in a journal (reader survey) or leave it in a public place for people to take away (e.g., for visitors’ surveys). If completion of the questionnaire has to be “controlled,” it can be distributed to individuals or groups in a “classroom survey,” and respondents complete the questionnaire in the presence of a supervisor. A special type of interview is the survey via the Internet, the so-called online survey. The questions are presented online – i.e., on the world wide web or via email. This demands, however, that the respondents are familiar with the technique and that they have easy access to it. Written presentation can only be used in standardized or semistandardized surveys. In answering the questions, the respondents are left to their own devices. They decide on the time and order of completion. Questions measuring spontaneous reactions cannot be used here whereas questions demanding longer contemplation or even research of cognitive contents or facts work quite well. At the same time, though, the respondents’ high self-determination is a disadvantage of the written interview. Selfselection of the respondents, i.e., the decision to participate, happens without any further personal persuasion to cooperate. The high level of nonresponse threatens the sample quality. Additionally, some specific types of interviews have been established in media and communication research. A Delphi study is a specific form of group communication with experts to predict particular circumstances. With the help of a (usually standardized) Delphi questionnaire, selected experts interact with each other anonymously. The goal of a Delphi survey is to jointly develop answers to specific problems by a repeated exchange of the results achieved. Mostly the aims are prognoses about the future (of communication, for example). A central difficulty of surveys in communication and media research is that they are mostly conducted independently of the actual media use in question. Usually, the respondents are interviewed after watching television or reading a newspaper, which demands a high ability to remember and to generalize. This is a problem because media use often takes place in low-involvement situations. Therefore, respondents may have difficulty reproducing their attitudes, behavior, motives or even emotional states with validity. Diaries can be used to solve this problem. A diary study is a written form of survey, which records the daily (media) behavior. The advantage of this method is that the respondents are only dependent on their short-term memory. The comparatively high effort necessary for this time-consuming method – for both the researcher and the respondents – is problematic, however. To avoid problems of retrospective measurement, the “Experience Sampling Method” (ESM) can be applied. The ESM requires the repeated application of a written questionnaire. Respondents are reminded (via telephone or SMS). The times of measurement are predetermined and chosen at random. Unlike the diary method, ESM mostly analyzes the quality of media use and tries to relate this to personal characteristics and/or situations.
Sampling And Methodological Problems
Almost without exception, all quantitative (and to a lesser degree, qualitative) surveys are conducted with samples. Thus, the quality of a survey is also highly dependent on the sampling method. Determination of the reliability and validity of an interview are problematic in principle, due to the high reactivity of this survey method. That means that the instrument has influence on the measured object, which is in consequence modified. It is impossible to conduct the same interview with the same respondents without the first interview influencing the second. Also, processing and problems of comprehension of the respondents influence the validity of the measurement. These problems cannot be solved – but they can be addressed with the use of high-quality interviews or by conducting meta-analyses.
References:
- Denzin, N. K., & Lincoln, Y. S. (eds.) (2000). Handbook of qualitative research, 2nd edn. Thousand Oaks, CA: Sage.
- Flick, U. (2006). An introduction to qualitative research, 3rd edn. Thousand Oaks, CA: Sage.
- Gordon, R. L. (1987). Interviewing: Strategy, techniques and tactics. Chicago, IL: Dorsey Press.
- Gubrium, J. F., & Holstein, J. A. (eds.) (2002). Handbook of interview research: Context and method. Thousand Oaks, CA: Sage.
- Sudman, S., & Bradburn, N. M. (1983). Asking questions: A practical guide to questionnaire design. San Francisco, CA: Jossey-Bass.
- Sudman, S., Bradburn, N. M., & Schwarz, N. (1996). Thinking about answers: The application of cognitive processes to survey methodology. San Francisco, CA: Jossey-Bass.