At a student conference recently, a man in a plaid jacket with elbow patches was very upset about my poster.  He crossed his arms across his chest and made a lot of noises that sounded like “Hmph!”  I asked if he had any questions.  “Where’s your control?? How can you say anything about your results without having a control group?”

Both the natural sciences and social sciences both share common roots in the discipline of philosophy, but the theoretical underpinnings and assumptions of those two fields are completely different.  I don’t know what happened when science and psychology were young siblings, but man, those are two are separate monsters now.

Just so you know, I have never taken the Qualitative Methods course.  I have never conducted interviews, or analyzed qualitative data.  But I find myself in situations where I need to explain and defend this type of research, as I plan to interview citizen science volunteers about their experience in the program.  I am learning about the difference between qualitative and quantitative data which each step of my project, but there is a LOT I still need to learn.

Here’s what I know about interviews:

  1. They are an opportunity for the participant to think about and answer questions they literally may have never thought about before.  Participants create/reflect on reality and make sense of their experience on the spot, and share that with the interviewer.   Participants are not revealing something to you that necessarily exists already.  Interviewers are not “looking into the mind” of the participant.
  2. It’s important to avoid leading questions, or questions where the answer is built in.  Asking a volunteer “Tell me how you got interested in volunteering…” assumes they were interested when they started volunteering.  Instead, you can ask them to provide a narrative of the time they started volunteering.  When volunteers respond to the prompt “Tell me about when you started volunteering with this program…”  they may tell you what interested them about it, and you can follow up using their language for clarification.  Follow-up and probing questions are the most important.  Good default probes include “Tell me more about that,” and “What do you mean by that?”
  3. You don’t necessarily set the sample size ahead of time, but wait for data saturation.  Let’s say you do 12 interviews and participants all give completely different answers.  You do 12 more interviews and you get fewer new types of responses.  You do 12 more and you don’t get any new types of responses.  You might be done!  Check for new discrepant evidence against your existing claims or patterns.
  4. Reporting qualitative data involves going through your analysis claim by claim, and supporting each claim with (4-5 paragraphs of) supporting evidence from the interviews.  I’ve read that there’s no one right way to analyze qualitative data, and your claims will be valid as long as they represent consistent themes or patterns that are supported by evidence.  Inter-rater reliability is another way to check the validity of claims.

And to the man in the plaid jacket, there are plenty of fields within the natural sciences that are similar to qualitative research in that they are descriptive, like geology or archeology, or in that it may be impossible to have a control, like astronomy.

Let me know what your experience is defending qualitative research, and what your favorite resources are for conducting interviews!


Print Friendly, PDF & Email

One thought on ““Say whaaaat?” about interviews

  1. Great reflections! I find that I more often have to “defend” qualitative research when working with those who don’t know much about it or whose work is quant-focused. Being able to do so is an essential skill in the business and government sectors where people often rely on Big Data to make decisions. You just have to be able to explain what the methods involve and the sort of valuable insights they can provide. Examples are always helpful.

    There are so many benefits to qual that quant can never provide, but the inverse is also true. I think many of the best projects use mixed-methods to look at research questions from multiple perspectives, but I will always be an advocate for this often misunderstood approach to research.

Leave a reply