To those who still see qualitative and quantitative methods as fundamentally different and opposed to each other, I’d like to offer two examples of research that have integrated qualitative and quantitative methods in a unique and meaningful way. In the first case, researchers used qualitative research to assess the measurement error in surveys of illicit behavior among street youth in Liberia. These researchers found that they could verify the results of self-reported data through observations and informal interviews, for relatively little cost. In the second case, researchers developed comprehensive life trajectories through qualitative interview techniques, to understand how major life events coincide with financial well-being. Researchers used this qualitative research to provide the more nuanced and detailed insights that their larger panel study failed to achieve.
The second discussion I witnessed was among a much smaller group of researchers clustered together for an entirely different purpose. Here, we listened to the representative of one prominent development research funding institution discuss the realities of mixed methods research. According to the representative, while an increasing number of evaluations are including qualitative research, often times the qualitative component is given less attention. Frequently, the qualitative work is done to help draft a survey better (a valuable use!), but isn’t used in the larger study to answer more complex or nuanced questions. Moreover, the qualitative research is often not well designed or thought through. I found this discussion particularly relevant to the theme of this series.
It is not always obvious how to make qualitative research rigorous, how to make it meaningful or how to do it well. Truthfully, like all other types of research, qualitative research is an art and one generally improves with experience. Still, it is helpful to have some guidelines. I will cover some main points that I find helpful for keeping qualitative research rigorous. Keep in mind that I am not an expert, just someone who values qualitative work and has learned a bit along the way. I should also point out that most of what I talk about will refer to interviews, because that is what I do. But I think it can be applied to other methods as well. This post starts with a discussion of sample size and the next post will cover techniques for data collection and analysis. For those who are interested in learning more, I’ll provide resources along the way and at the end of each post.
Picking the Right Sample Size
Qualitative research has some ambiguous qualities. One of them has to do with sample size. Unlike studies that capture changes in quantities, qualitative studies have no clear formula for picking the “appropriate” number of respondents. Indeed, there is not even a universally agreed upon method for selecting a sample. Some say sample size doesn’t matter; others say 30 is the magic number; while an entirely different approach suggests one should keep going until the responses start repeating each other. Sample size, considered so crucial to the construction of a sound quantitative study, immediately proves itself to be a stumbling block for establishing “rigor” in qualitative research.
Typically, we engage in qualitative research because we want to understand a phenomenon better. Our questions can’t be answered with a yes or no or multiple choice response; we want to know “what?”, “why?” or “how?” The trick here is that each observation is unique. One person or program site or village is not going to provide perfect information about the next. (This is, of course, the same variation between observations that we consider when we conduct power calculations or think of external validity.) People differ in how they understand and perceive the world around them, they differ in opinions, they differ in so many ways on the very things we want them to help us understand. If you conduct one lengthy interview with one person, you will surely understand the phenomenon as applied to this one person, but you won’t have the full range of responses. In fact, you may have just accidentally interviewed the one outlier in the community.
The key question is: how much does the full range of responses matter? If it doesn’t matter at all, then the number of observations you choose for your qualitative research is probably not very important. If you are hoping to capture the full range of responses, or very close to it, then the situation becomes more tricky. If you use the “keep going until responses start repeating” approach, you just may miss out on the full range. Given what we know about probability, there’s a good chance that you may interview several people in a row who give similar responses due to…well, chance. So can you be sure that the repetitiveness is an indication that you’ve got the full range – or at least most of it?
The best bet is to avoid “rules of thumb” and to make a thoughtful, well-reasoned justification for your sample size, which is clearly explained upfront. Selection of sample size should match your goals and the type of research you are doing. There is a lot of guidance out there for identifying the appropriate sample size, which may vary according to your project’s goals and approach. One enterprising marketing consultant offers a formula for calculating sample size, based on the probability of missing infrequent phenomena. Although this approach has not yet made its way to academic circles, it certainly can be appealing to those who crave the satisfaction that only mathematical formulas can provide.
Note: Two researchers from the National Centre for Research Methods in the UK tried to address this sample size question by reaching out to leading qualitative scholars and soliciting their input. They obtained a variety of responses. It is an interesting read and provides a nice foundation for thinking about sample sizes in qualitative research.
Further reading:
Patton, Michael Quinn (2002). Qualitative evaluation and research methods. Thousand Oaks, CA: Sage Publications.
Anwuegbuzie, Anthony J. & Leech, Nancy L., “A Call for Qualitative Power Analysis,” Quality and Quantity 41 (2007): 105-121.
Marshall, Bryan, et al., “Does Sample Size Matter in Qualitative Research? A Review of Qualitative Interviews in IS Research,” Journal of Computer Information Systems (2013): 11-22.
Can Qualitative Research be Rigorous? Part 1: What is Qualitative Research?