Those who criticize qualitative research very often say that the findings are not representative. Is it true? It's true that the purpose of qualitative study is not to generalize findings but if there is any way that might help to enhance its strength.
Discussion of generalizability in qualitative research typically involve a comparison to sampling in survey research, where the goal is to get a precise estimate of a population value for a mean or a correlation etc. Since qualitative research does not involve measurement, that kind of generalization is not possible.
In this case, the more relevant issue is whether you can take the broad conclusions you reach from interviewing a small number of people and then say that they apply to the larger set of people "represented" by those participants. From a statistical point of view, the answer is no, on a variety of technical grounds.
But for qualitative research I prefer not to concentrate on statistical generalizability but instead consider what Lincoln and Guba (1985) called transferability. In other words, given the results you have and the ways that you obtained those results, where else do you think you could apply them?
No. As a matter of fact FGD is most reliable source of qualitative data. When question is asked in the group, there is no scope of receiving non representative response. As responses are in consensus form. If some one says some thing which is not true, other members argue /oppose and do not agree. This natural check in the form of group behavior, ensures you get information to which everyone agree
Discussion of generalizability in qualitative research typically involve a comparison to sampling in survey research, where the goal is to get a precise estimate of a population value for a mean or a correlation etc. Since qualitative research does not involve measurement, that kind of generalization is not possible.
In this case, the more relevant issue is whether you can take the broad conclusions you reach from interviewing a small number of people and then say that they apply to the larger set of people "represented" by those participants. From a statistical point of view, the answer is no, on a variety of technical grounds.
But for qualitative research I prefer not to concentrate on statistical generalizability but instead consider what Lincoln and Guba (1985) called transferability. In other words, given the results you have and the ways that you obtained those results, where else do you think you could apply them?
I use qualitative research to capture information which can't be captured through a typical quantification of information format. So qualitative information is actually used to strengthen findings based on quantitative research
- Badges
- Science method
- Similar topics
- Methodology
- Research Methodology
- Qualitative Research