When we conduct interviews, we look to gain enough data that can offer a reliable picture of the issue being examined. Depending on the field, interview numbers in a study can be very small (I have even heard of a published study with only 1 interview), to >20.
- What is the best way to determine data saturation - i.e. that conducting further interviews will be unlikely to reveal pertinent new data ?
Hello Nicholas
This is quite an informative paper about this subject:
Fusch, P. I., & Ness, L. R. (2015). Are we there yet? Data saturation in qualitative research. The Qualitative Report, 20(9), 1408.
http://tqr.nova.edu/wp-content/uploads/2015/09/fusch1.pdf
This also might be of interest:
Mason, M. (2010, August). Sample size and saturation in PhD studies using qualitative interviews. In Forum Qualitative Sozialforschung/Forum: Qualitative Social Research (Vol. 11, No. 3).
https://www.researchgate.net/profile/Mark_Mason7/publication/47408617_Sample_Size_and_Saturation_in_PhD_Studies_Using_Qualitative_Interviews/links/55141f270cf2eda0df30424b.pdf
This might be helpful, but I haven't read the full text:
Bowen, G. A. (2008). Naturalistic inquiry and the saturation concept: a research note. Qualitative research, 8(1), 137-152.
https://www.researchgate.net/publication/249731235_Naturalistic_Inquiry_and_the_Saturation_Concept_A_Research_Note
Again, I have not read the full text, but this paper questions the concept of data saturation:
O'Reilly, M., & Parker, N. (2012). ‘Unsatisfactory Saturation’: a critical exploration of the notion of saturated sample sizes in qualitative research. Qualitative Research, 1468794112446106.
https://www.researchgate.net/publication/257413822_Unsatisfactory_Saturation'_A_critical_exploration_of_the_notion_of_saturated_sample_sizes_in_qualitative_research
Very best wishes with your PhD,
Mary
Article Sample Size and Saturation in PhD Studies Using Qualitative Interviews
Article Naturalistic Inquiry and the Saturation Concept: A Research Note
Article Unsatisfactory Saturation’: A critical exploration of the no...
The O'Riley paper is very good and discusses how the idea of saturation was introduced linked to GT, but it seems to have been transferred blindly to other forms of qualitative research. Mason talks of how PhD researchers tend to submit theses based on overly-large amounts of interviews "just to be on the safe side". We seem to be picking nice round numbers as opposed to any consideration whether the data demonstrates what it claims to.
A good paper is: Baker, S. E., and Edwards, R. (2012) How many qualitative interviews is enough? Expert voices and early career reflections on sampling and cases in qualitative research. Southampton: ESRC National Centre for Research Methods, University of
Southampton. In this, they recount how an expert who they approached did not offer their full opinion, but rather turned the tables by asking 'Well how many expert opinions are you looking for to answer your question?* :-)
Hopefully there will be some practical examples of how RG members have used small numbers of interviews to demonstrate a particular issue, and the ways they have justified this.
In my opinion, the idea of qualitative data saturation is linked to generalizability of data and "sample size." However, qualitative data is not generalizable, by its definition. Qualitative methodology is about understanding the experiences and opinions of the people interviewed.
In my experience, data saturation is mainly advocated by quantitative researchers who do not understand qualitative methodology.
Michael, I agree. In GT it makes perfect sense, but when treated almost as a quantitative measure of quality, it simply asks for it to be misapplied. Some people see 'science', 'research', 'methodology' etc, & instantly impose some form of rule in the belief that without it, there is no rigor.
There definitely is a difference between theoretical saturation in grounded theory and the idea that no new codes are appearing, which I prefer to call code saturation. Recently, the later idea has received a certain amount of more quantitative attention, as in the attached
Data saturation, code saturation and theoretical saturation are interlinked, but nevertheless different concepts. Theoretical saturation is related to the conceptualisation work in grounded theory. It means that your story line is found, that you story can be told, that you produced your theoretical element from your analysis. Data saturation has to do with sampling. If you need new cases for expanding your theory through constant comparison, you look for new cases and data (purposeful sampling). Code saturation is on a more descriptive level. It mirrors both your data and your theoretical sensitivity.
What's everyone views on saturation table in qualitative research? Is it worth spending the time and effort creating one to justify saturation?
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size