How to conduct a qualitative interview

Qualitative vs. quantitative research

Qualitative as well as quantitative research have their pros and cons. In a quantitative survey, the process of creating questionnaires usually takes longer, because questions should be set in stone and are not adaptable afterward. Therefore, it is essential to state the most adequate questions to meet the research goals, with the ideal formulations and answer possibilities, so the statistical evaluation works smoothly. The main advantage of quantitative research is that it is easier to compare certain groups of people because the sample is usually larger than in qualitative research and often the requirement of representativity is given.

Due to the principle of openness in qualitative research, it is possible to react to interview partners more flexibly. In many situations, it is also allowed to ask additional questions or adapt them during an interview. In this case, the interview guidelines can be viewed as orientation. The researchers should be open to surprises and allow the interview partners to set their own thematic priorities despite prepared questions. The drawback here is that analyzing interviews takes a lot of time per person compared to quantitative tools.

An example from the content strategy world: For the creation of personas, one could choose quantitative methods like questionnaires. This would be sufficient for the process and facilitate an overview of the topic or field in question. With qualitative methods, such as interviews, the researcher could go into detail and adapt the questions considering the position and relevancy of the interviewees.

When to use qualitative methods?

“With qualitative social research it is possible to find out what lies behind little-known phenomena and it can be used to generate surprising discoveries and shed light on details that are harder to see with quantitative methodology”, Sackl-Sharif states in her first session. When it comes to content strategy, it is often the case that the weak points are rather hidden and not easy to spot at first sight. This makes qualitative interviews even more interesting for the field.

The decision if a qualitative research method is the best approach for a project, is affected by different criteria:

• Depth of knowledge about the topic: If the researcher is an insider to the field and has a lot of in-depth knowledge, it may be easier to collect details in the interviews and to speak “the right language”. But being an insider could also create blind spots that outsiders don’t have, which should be reflected during the analysis process. As an outsider, it may take more time to get in contact with the field or to learn the language of the research group.
• Existing research and literature: The amount and quality of already existing knowledge are important to decide if it’s necessary to explore the field in a wide sense or focus on details that are yet to be discovered. When a new field is about to be explored, qualitative methods like interviews or observations can help to gather first insights and to orient oneself within a new field.
• Theoretical background of the study: No method without methodology. It’s always link the method with the theoretical background of a study. For content strategists, it is important to reflect if the used methods fit the goals of content strategy in general and the goals of the study.
• Available resources: As always, time, money, staff, and competencies also play big parts when it comes to choosing the best-fitting methodology. As a content strategist, the researcher is most likely limited by the resources given by the ordering company or organization.

Susanne Sackl-Sharif teaching the students of COS21 in Graz.

Conducting qualitative interviews

Step 1: The topic

For qualitative interviews, it is crucial to understand the core of the research question and already discover possible pain and gain points that should be tackled during the interviews. Up-front research concerning existing knowledge about the topic is one part of the process, clarifying the objective is the other. And choosing the right strategy is the next important step.

Step 2: The interview form

Three main strategies are especially interesting for content strategists when it comes to gathering verbal data in research projects:

1. Narrative interviews: The interviewee should be invited to talk about their story. This strategy is recommended in biographical research or when it comes to more sensitive topics. But pay attention: This kind of interview produces an amount of unstructured textual material and therefore needs higher effort in transcribing and analyzing it.

2. Group interviews: Group discussions such as focus groups could be a good fit for the topic of research because they simulate an “everyday situation” and are not isolated like one-to-one interviews. Extreme or incorrect opinions usually get corrected by the other participants and they validate their opinions together. Group interviews are particularly interesting when it comes to problem-solving and finding solutions, and are often used in marketing. For content strategists, this strategy could be used for the evaluation of a website or social media channel.

3. Semi-structured interviews: Semi-structured interviews are based on question-answer models. They can be used by content strategists to create personas. An example is the problem-centered interview, which concerns a relevant social problem or, specifically, a smaller unit of this problem, and should include social developments. Another strategy is interviews with experts, who have privileged knowledge on a topic. The professional person is at the center of the interview.

Step 3: The interview questions

1. Narrative interviews: In narrative interviews, it is important to prepare the interviewees and invite them to tell their own stories. Most important for this strategy is the entry question: It should generate a whole story about how everything started and the most important stages relevant to the research topic. The interviewees should not be interrupted and should set the end point of their narration by themselves. When they are done, it’s time for clarifying uncertainties and details. Afterward, the researcher can ask some more abstract and theoretical questions that might be of interest to the research topic.

2. Group interview: At the beginning of a group interview, the moderator should explain the procedure, clarify the expectations of the participants and define their role clearly. Afterward, all members of the group should be introduced as a warming-up exercise, before it comes to the discussion stimulus. The stimulus could be a provocative question, a short film, a website, a social media channel, or an inciting theory, depending on the research topic.
   The moderator should not only take care of the formal direction, such as time management, but also of the steering during the process. In these kinds of discussions, the participants mainly talk to each other and therefore go in certain directions over time. If the course of the discussion goes in the wrong direction or opinion leaders dominate the group, the moderator should be prepared to steer the discussion and try to influence the dynamics with certain questions. Some tips from Sackl-Sharif considering this were to ask for other opinions before going to the next question, or address specific (non-dominant) participants directly when the next topic occurs. Additionally, it is always wise to state in the beginning that the researcher is interested in different opinions and perspectives.

3. Semi-structured interviews: An elaborated guideline is especially important when it comes to one-to-one interviews. For this, Sackl-Sharif provided the students with helpful tips:

   1. It is good to have three or four main topics with one main question that covers the most important details of this topic.

   2. It is always preferable to produce spontaneous narratives, so the main questions should be very openly phrased.

   3. There should not be any yes/no questions. Instead of general phrasing, the questions could start with “what”, “how”, “to what extent”, etc.

   4. Questions should be short, precise, and not too direct or too abstract.

   5. There should only be one question in a sentence.

   6. The background of the interviewees should always be considered, and the language should be easy to understand

   7. The last question is preferably something like “What did I forget?” or “What would you like to address?”

The lecturer also mentioned the SPSS principle by German researcher Cornelia Helfferich. In this method, the researcher first collects (S for “sammeln”) up to fifty questions, then checks (P for “prüfen”) them to eliminate fact and pure information questions as well as questions that don’t generate open answers. Afterward, the sorting (S) phase starts, where the questions get categorized into 1-4 bundles. Then, the subsume (S) phase completes the process by finding a single, simple narrative request, which combines the questions of each bundle.

Step 4: The interview partners

When it comes to finding the best-fitting interviewees, there is no golden rule. One could always have bad luck and end up with participants who are reluctant to share or do not show up at all. Generally speaking, “good” interview partners are the ones who are very talkative, try to find answers to every question, and honestly admit if they don’t have an answer. Interviewees should ideally surprise with their information and opinions and help develop a new theory. They should also fit the chosen strategy, research question, and interview technique.

If it comes to an expert interview, there is a rule: Try to find the highest variation of experts to gather a maximum variation of opinions and statements. The researcher should be aware to choose real experts though, so participants should have privileged access to knowledge in the concerning field.

In group discussions, the participants could either consist of people who exist in real life or artificial groups who are put together by the researcher depending on certain characteristics. The participants could also either be homogeneous, so have some similarities, or heterogeneous and have differing features. When it comes to putting a group together, the researcher should always think of the consequences of their choices: Is it smart to include the CEO of a company in the discussion? What is the ratio between dominant and reserved participants? Are the interviewees too similar or too different?

Step 5: The analysis

As stated before, analyzing data from qualitative social research projects usually takes more time and effort compared to quantitative surveys. The goal here is to find categories for the data and identify relevant passages.

The categorization can happen in two ways: Inductive, where categories are developed from the material, and deductive, where the categories are derived from already existing theories and empirical results before going through the material. According to Uwe Flick, who suggests inductive analysis, a short summary of each participant and their views is also recommended during the process. An approach Sackl-Sharif suggested when it comes to narrative interviews that concern sensitive topics, is the importance of reviewing it ethically and asking the interviewee again if all the information can be used.

The lecturer highly recommended the software “MAXQDA” which is especially suitable for the analysis of qualitative research or mixed methods.

Where to go from here #

Authors who were frequently mentioned during the lectures were Uwe Flick and Cornelia Helfferich. Details are in the reference section.

References #

Flick, Uwe (2018). Managing the Quality of Qualitative Research (2nd ed.). London/Thousand Oaks, CA/Dehli: Sage.

Cornelia Helfferich (2005). Die Qualität qualitativer Daten. Manual für die Durchführung qualitativer Interviews. Lehrbuch, Wiesbaden.