How many interviews have you conducted?

How do I conduct expert interviews?

Many students want to do their bachelor thesis or a larger seminar paperempirically work and decide to do a so-called Expert interview respectively. The following article explains what an expert interview is, what special features have to be taken into account when collecting and evaluating the data and warns of frequent “pitfalls” that students stumble upon.

Conducting and evaluating (expert) interviews is an extremely complex process of qualitative social research that knows numerous subtleties that cannot be dealt with in detail here. It is therefore urgently recommended that you read the numerous specialist literature available on the subject (e.g. Bogner, Littig, Menz, 2014; Flick, 2009; Mayring, 2010).

What is an expert interview and who is considered an "expert"?

An expert interview (or of course: an expert interview) is usually a so-called partially structured interview. “Partially structured” means that the conversation with the expert is based on a guideline. However, the questions to the experts are neither standardized nor self-contained (e.g. like in a questionnaire survey in which, for example, consent is asked: yes / no, fully agree, partly / partly, disagree, etc.) nor is it a so-called narrative interview, in which the interviewee chats freely “from the sewing box” without being controlled by the interviewer. For an expert interview, a guide must be prepared that contains all the topics and questions to be discussed. However, this document serves only as a working aid for the interviewer and is used by the interviewee Not handed out. It is allowed to deviate from the wording of the question as well as the order of the questions.

Formally, an expert interview is one of the qualitative methods of empirical social research. What exactly this means becomes apparent when evaluating the interview material.

Preparation for the expert interview

Interviewing an expert does not release the author of a paper from the obligation to acquire expert knowledge himself. On the contrary: Before the interview can be conducted, the interviewer must acquire an expertise himself. This is the only way to create sensible key questions in advance and to choose a suitable expert.

Who is an expert anyway?

The question of who can be considered an expert is a contentious issue within qualitative social research. Flick (2009) advocates a broad expert term:

The expertise is therefore independent of academic titles or the like and depends on the research area or the research-leading question. The expert is usually interviewed in his role as a role / function holder. Depending on the research topic and question, a member of a youthful subculture, a drug user or a public authority employee can be an “expert”.

The answer to the question of who or what 'experts' are depends on the subject of investigation and the related theoretical-analytical research approach. (...) As experts one could describe those people who are particularly competent as "experts" with regard to an interesting issue.

Deeke, 1995, pp. 7f .; quoted in Flick 2009: 214

Once an expert has been found, an appointment has to be made. Since the success of a work stands and falls with good quality, usable interviews, it is crucial to agree on a binding appointment and, if necessary, to remind the interviewee of the appointment again shortly before the agreed interview date. Often there is a desire to find out the specific interview questions in advance from the interlocutor. This should be done whenever possible Not must be complied with, as this would affect the spontaneity of the interview situation. For the same reason, an interview conducted face-to-face is preferable to an email interview. It is also advisable to draw the interview partner's attention to the fact that the conversation should be recorded for evaluation purposes. In this way, the expert can already adjust to the situation.

A discussion guide must be drawn up prior to the interview. It serves as a reminder for the interviewer so that no important questions are forgotten in the course of the conversation. The guideline depicts an idealized course of the conversation that will probably not be possible in this form. In the case of partially standardized surveys, the order of the questions is variable and can / must be adapted to the course of the conversation. In order to get a quick and discreet overview of the outstanding questions, it is advisable to group the questions on the one hand by main topic and on the other hand to print out the questions in large font size (nothing disturbs the flow of the conversation more than when the interview lasts for a few seconds Scroll through notes to find the next question).

As part of the preparation for the interview, the interviewer must also think about the start of the interview. What information about yourself and the research project should / must be disclosed? It is important to find a balance between a suitable attunement to the topic and the unconditional avoidance of tendentious influencing of the interview partner.

Finally, it is important to test the recording device and to familiarize yourself with it. How loud can background noises be heard in the room? How long does the recorder battery last? How big is the storage capacity of the device? Can the created recordings be easily transferred to the computer and processed further? etc. The recording function of cell phones achieves quite useful results here. However, a professional dictation machine is particularly useful for difficult recording situations (lots of background noise, an interview has to be conducted outdoors, etc.).

Conducting an expert interview

The expert who makes himself available for an interview usually does so without receiving any counter-guidance. It goes without saying, therefore, that the interview should be carried out whenever possible according to the wishes of the interviewee; i.e. this usually determines the time and place of the interview. When determining the location, however, care should be taken to ensure that as little disruptive background noise as possible has a negative impact on the recording quality. A street café is therefore rather unsuitable as a place to talk, just like a lounge that is used by many employees, etc.

At the beginning of the interview it is essential to clarify whether the interview partner anonymity should / can be assured. Since this person was selected in his function as a role bearer and less as an individual, this is usually not a problem. It is more difficult if the specific role allows a clear conclusion about the person (e.g. head of authority X, employee of Y responsible for Area Z). The degree of anonymity has an impact on the response behavior. A conversation partner who trusts the interviewer that his anonymity will be preserved is much more likely to comment on critical / controversial topics.

Before the interview begins, it must be pointed out again that the conversation will be recorded. Only after the interview partner agrees to the recording again will that be Recording device turned on. It should be positioned so that good sound quality is guaranteed, but at the same time the person you are speaking to is not too distracted by the device in front of him.

If you plan to publish the research results beyond a university-internal publication (e.g. in a specialist journal), it is necessary to obtain a written letter of acceptance to be able to use interview passages (if possible in anonymous form, see above).

Once these questions have been clarified, the interview can actually begin.

If you are planning to conduct several interviews, it is advisable to name the location, time and interview partner so that the recordings can be clearly assigned later during the evaluation (“Today is DD.MM.YY. We are in the premises of company X. Opposite me sits the head of department Y, Mr. Meier, who has kindly agreed to answer some questions about Z. ”).

In order to get the interview partner in the mood for the conversation, the interviewer leads on to the actual topic. This Conversation impulse is to be set with care, as it determines the further course of the conversation. In general, one should strive for the greatest possible objectivity and refrain from making any judgmental statements (e.g. "As you already know, I do research on the subject of drug prohibition. I have a few questions for you about this."). Depending on the research topic / question, it can also be conceivable that one specifically tries to steer the course of the conversation (e.g. "In recent years the number of drug deaths has increased, as has the number of first-time drug users registered by the police. This was Reason for me to initiate a research project on the legalization of cannabis. I would now like to ask you a few questions about this. "). One can imagine that the conversation in the above examples runs very differently depending on the chosen entry point.

The problem of social desirability

In empirical social research, social desirability is used to describe the fact that test subjects adjust their response behavior to the anticipated attitude of the questioner and supposedly answer “as requested”. An example of a response behavior that is shaped by social desirability can be found, for example, in studies of sexual behavior. When asked about the number of previous sexual partners, male respondents usually give a significantly higher number than female respondents. This is presumably due to the fact that in Western cultures men are more likely to have promiscuity than women who are socially expected to be sexual abstinence or at least monogamous. A woman who truthfully reports about several past sexual partners would break with these normative social ideas. In order to avert the impending loss of reputation, this woman may decide to specify a lower number of sexual partners. The same effect can also be responsible for the fact that male respondents overestimate the number of their past sexual partners.

The problem of social desirability is particularly evident when it comes to issues that (can) be subject to moral judgment (e.g. how do you feel about the death penalty? Have you ever committed a criminal offense in the past? How do you feel about corporal punishment of children? Etc.) .

Basically, it is important to ask open questions. The aim of an expert interview must be that the expert reveals as much of his expert knowledge as possible. However, this does not work if he answers closed questions with “yes” or “no”. In addition, suggestive questions should be avoided at all costs (e.g. "You are certainly also of the opinion that ... Or maybe not?").

Ideally, a natural flow of conversation arises after a short time, which the interviewer should carefully steer but not interrupt. With a slight nod of the head or an affirmative “Yes”, the person you are speaking to can be encouraged to continue speaking. Meanwhile, the interviewer keeps an eye on the guidelines and checks which topics have already been discussed. Since this is a partially structured interview, the intended sequence of topics does not have to be adhered to. Let the expert talk in peace at his own pace and following his own logic. When the expert is finished with his explanations, you can always come back to a topic and ask questions ("You mentioned earlier that you ... Can you tell us more about that?").

Transcription and evaluation of an expert interview

Transcribing expert interviews

Once the interview has been held with one or more experts, the real work begins. The recorded conversations are now listened to and written down word for word. This activity is known as that Transcribe (Latin transkribere = to rewrite). You will quickly find that this is an extremely time-consuming endeavor. One of the reasons for this is that the spoken word differs greatly from the written language. Sentences are not brought to an end, verbs are missing at the end of the sentence or endings do not correspond to the grammatical rules and many pauses and filler words fragment the flow of words. The transcription takes place here word for word.

Depending on the research project and question, it can even be crucial to also note down non-linguistic utterances. For example, if the interviewee pauses for a long time before he or she starts to respond, it may be due to insecurity, inner conflict, or resentment. But laughing, crying, lowering or raising the voice can / should also be noted when transcribing.

The transcribed interviews are attached to the appendix of the thesis. You will each receive a unique numbering / labeling, separate page numbering and, above all, line numbering. In this way, a source reference to passages from the interviews can be clearly established (e.g. "Interview 1, p. 3, line 18")

Software to support the transcription and evaluation of expert interviews

The transcription and the subsequent evaluation of expert interviews can easily be done with conventional Office programs.

However, programs specially tailored to these tasks make work considerably easier. The f4transkript program for transcribing expert interviews and the MAXQDA program for evaluating transcribed interviews have proven to be particularly effective. Both programs are chargeable, but offer moderately priced, time-limited licenses for students.

The evaluation of expert interviews

The content-analytical evaluation of interviews is a data-reducing process, i.e. the transcribed interviews must now be checked for their meaning and summarized. This work step is also known as Coding.

This is done by creating categories or codes. Which and how many categories / codes are formed depends on the expert interviews and the research question. By reading the expert interviews several times and interpreting statements, individual words, sentences or even entire passages are assigned to a category. This evaluation step is the most important in the research process. When categorizing, one should be guided by the material and resist the temptation to achieve a desired result through misinterpretation or overinterpretation of statements. In cases of doubt, certainty is provided by the so-called intercoder validity, i.e. checking whether comparable results are achieved when the material is coded by two independent people.

After around a third of the material has been coded, the categories are collected and checked again, and categories are added or deleted if necessary. After determining the category scheme, the entire material is then subjected to an examination.

The evaluation of expert interviews in the manner described is basically a qualitative research method. But of course there are also points of contact with quantitative research approaches. For example, word mentions could be counted and statistically evaluated.

Literature and further sources

  • Bogner, A .; Littig, B .; Menz, W. (2014) Interviews with experts. A practice-oriented introduction. Wiesbaden: Springer VS.
  • Flick, U. (2009). Qualitative social research. An introduction. (2nd ed.). Reinbek near Hamburg: Rowohlt Taschenbuch Verlag.
  • Forum: Qualitative Social Research. Online at:
  • Mayring, P. (2010). Qualitative content analysis. Basics and Techniques. Weinheim: Beltz.

Category: ResearchTags: expert interview, research, qualitative method