How to choose the right sample size for a qualitative study… and convince your supervisor that you know what you’re doing
The question of how many participants are enough for a qualitative interview is, in my opinion, one of the most difficult questions to find an answer to in the literature. In fact, many authors who set out to find specific guidelines on the ideal sample size in qualitative research in the literature have also concluded that these are “virtually non-existent” (Guest, Bunce and Johnson, 2005: 59). This is particularly unfortunate, given that as a student planning to undertake your research, one of the things that will be most likely to be asked of you is to indicate, and justify, the number of participants in your planned study (this also includes your PhD proposal in which you are expected to give as much detail of the study as possible).
If you, then, turn to the literature, hoping to find advice from some of the great minds in research methodology, you are likely to find them evading the question and often hiding behind the term “saturation” which refers to the point at which gathering new data does not provide any new theoretical insights into the studied phenomenon. Although the concept of saturation may also be controversial, not least because the longer you explore, analyse and reflect on your data, you are always likely to find something “new” in it, it has come to be the guiding concept in establishing sample size in many qualitative studies. As Guest, Bunce and Johnson (2005) rightly point out, however
“although the idea of saturation is helpful at the conceptual level, it provides little practical guidance for estimating sample sizes for robust research prior to data collection”
(Guest, Bunce and Johnson, 2005: 59)
In other words – how in the world are we supposed to know when we will reach saturation PRIOR TO THE STUDY???
My advice is to use the available literature on the point of saturation and use it to justify your decision regarding the sample size. I did it for my PhD study, as I was growing frustrated that I really have to justify my decision to include 20 participants for an interview, even though I had read dozens of reports in which this number, or smaller, was common (“are you going to interview 20 participants just because others did?”). I just felt that this would be enough, and my common sense, which as I learnt throughout my PhD was the last thing that anyone would care about, was telling me the same thing. In order to support my decision with the literature, however, and considering that there are hardly any guidelines for establishing sample size, I decided to try to reach some sort of conclusion as to how many participants are enough to reach saturation and use it as my main argument for establishing the size of the sample.
So what does the literature tell us about this? Just as there is not single answer as to what sample size is sufficient, there is no single answer to the question of what sample size is sufficient to reach theoretical saturation. Such factors as heterogeneity of the studied population, the scope of the study and the adopted methods and their application (e.g. the length of the interviews) are believed, however, to have a central role in achieving this (cf. Baker and Edwards, 2012; Guest, Bunce and Johnson, 2005; Mason, 2010). Mason’s (2010) analysis of 560 PhD studies that adopted a qualitative interview as their main method revealed that the most common sample size in qualitative research is between 15 and 50 participants, with 20 being the average sample size in grounded theory studies (which was also the type of study I was undertaking). Guest, Bunce and Johnson (2005) used data from their own study to conclude that 88% of the codes they developed when analysing the data from 60 qualitative interviews were created by the time 12 interviews had been conducted.
These findings helped me in arguing that my initial sample size was going to be 20. “Given the detailed design of the study, which includes triangulation of the data and methods”, I argued, “I believe that this number will enable me to make valid judgements about the general trends emerging in the data”. I also stated that I am planning to recruit more participants, should the saturation not occur.
I hope that this article will help you in your quest to determine the sample size for your study and give you an idea of how you can go about arguing that it is a well thought-through decision. Do remember, however, that 20 participants may be enough for one study and not enough, or too many, for another. The point of this article was not to argue that 20 participants is a universally right number for a qualitative study, but rather to point to the fact that there is no such universally right number and that you are not the only one struggling to find guidelines regarding the interview sample size, as well as to put forward the concept of saturation as one of possible principles that may guide you in deciding how many participants to recruit for your study.
If you have any questions regarding this topic, comment below or send me a message through my Facebook page.
- UPDATE – see my Facebook page for my response to the question about the relevance of “saturation” for Phenomenological research
References:
Baker, S. & Edwards, R. (eds., 2012). How many qualitative interviews is enough? Expert voices and early career reflections on sampling and cases in qualitative research. National Centre for Research Methods, 1-42.
Guest, G., Bunce, A. & Johnson, L. (2005). How many interviews are enough? An experiment with data saturation and variability. Field Methods, 18 (1), 59-82.
Mason, M. (2010). Sample Size and Saturation in PhD Studies Using Qualitative Interviews. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 11 (3).
