This is an interesting question as it is not appropriate to quantify qualitative data using quantitative methods. Qualitative research does not look at variables however focuses on events and incidents that are specific to the needs of the study. The richness and volume of the data is reflected in the small participant numbers. Therefore the selection of participants into qualitative studies such as phenomenology requires participants to have experience the phenomenon of interest and recruitment ceases when data saturation is reached and this is evident when no new themes are forthcoming.
There are at least two ways you might be using randomization: random selection of the participants or random assignment to categories such as who gets interviewed. Random selection (the equivalent of random sampling in surveys) is seldom worthwhile in qualitative studies because the sample sizes are so small that you cannot create meaningful generalizations from them. In more technical terms, with small Ns you produce large standard errors, which means that you have very wide confidence intervals around any estimate that you produce. Thus, while your estimates might be generalizable in theory, in practice they are so unstable that they are virtually worthless.
In qualitative research, the typical response to this problem is to use "purposive" (or, purposeful) sampling instead of random sampling. This consists of setting criteria for which kinds of informants would be most useful for your purposes. The simplest way to think of this is that if you are going to be interviewing only 10 or 15 people, then you should make careful choices about which people to interview.
In contrast, with random assignment (which is typically used in experimental designs), you would have a set of people who are all equally eligible to be interviewed, according to your purposive criteria, and then you would randomize the people that you did or did not talk to. The goal here would be to eliminate "selection" effects where you might choose people in ways that would bias your sample. In other words, you would first purposively select a larger sample of potential participants, and then randomly select that actual participants from within that group.
While this idea of random assignment makes sense from an abstract, methodological point of view, I have seldom seen qualitative researchers use it in practice. Instead, the emphasis is on a careful use of purposive sampling.
Thank you very much Rachel and David, It seems, you very interested and experienced in qualitative studies. Hope will be in touch continuously.
Since the goals and purposes of quantitative and qualitative research methods are polar opposites, you pose an interesting question. The goal of quantitative research is to generalize results to the population from which the sample was drawn, while the goals of qualitative inquiry is not to generalize, but to understand and interpret phenomena from the viewpoint of the human experience. In order to generalize, sample size and randomization is crucial. As far as randomization goes, the argument for it is that a heterogeneous sample will reduce sample selection bias, which can skew results and cause you to come to erroneous conclusions. Random selection allows the sample to be more representative of the population from which the sample was drawn (Aguinis& Gottfredson, 2010; Vogt, 2007).
Ref:
Aguinis, H., & Gottfredson, R. K. (2010). Best-practice recommendations for estimating interaction effects using moderated multiple regression. Journal of Organizational Behavior, 31(6), 776-786.
Vogt, P.W., (2007). Quantitative research methods for professionals. Boston, MA: Allyn and Bacon.
The way we approach a particular research method e.g. quantitative, qualitative or mixed methods is depending on the research problem and research objective. If the research objective is to generalize the findings to the entire population, then randomization of sampled respondents are appropriate. If the research objective is to study subject(s) in an in-depth manner as well as the phenomena etc. then randomization of sample is not required. However, when you are performing mixed method research e.g. sequential explanatory study (quantitative then follow by qualitative research), then you need to select the subject(s) for interview from the sample pool of respondents conveniently / randomly that you have collected during the quantitative research. Another words, you can't choose any subject for interview which is not from the sampled pool even though the subject is relevant & convenient but s/he didn't involve in responding to your earlier quantitative survey.
- Badges
- Science method
- Similar topics
- Methodology
- Research Methodology
- Qualitative Research