27th January 2012 at 12:17 pm #2761
I am part way through the data collection stage of a PhD using an arts-based methodology involving meeting partcipants once a week over 12 weeks. This process will happen twice. The participants are women who have experienced domestic violence and the location is community-based – factors that mean the number of participants can vary week to week depending on what else is happening in the partcipants lives. The first time around there were 3 consitent participants. This time around there may be twice that amount. Obviously this is lower than would be deemed acceptable within a piece of quantitaive work. My question is: what are the justifications for having a low number of consitant participants within a piece of research that involves meeting particpants with the frequency described using the arts to express responses to questions? How far can the depth over breadth argument be extended before validity can’t be justified. Any body else been in similar circumstance or are aware of over similar PhD’s that have been deemed acceptable with a similar method and partcipant involvement. Thanks.27th January 2012 at 2:50 pm #2765Paula DawidowiczMember
If I understand you correctly, you’re asking two questions–why you can have a smaller sample and how small your sample can be. The nice thing about a qualitative design is that by its nature it’s not supposed to be generalizable to other groups. You’re looking at the nature of your specific group to get an in-depth picture of its members. The expectation of readers is that your exploratory or explanatory picture gives them solid insights into what can be the case so they can, in turn, consider the applicability of that information to the individuals/groups they are themselves examining. That’s why you can have a smaller group–different expectation/purpose for the data you gain. As for numbers, with qualitative research there are not set formulas or number expectations. This can vary by design used, by population size, and by your goals themselves. If you’re using the same methodology–same questions, same activities–in your study with each of your groups, you can conduct repeated groups until you get an acceptable number of participants. The number of participants considered acceptable is determined by you and your committee. I have some guidelines I use with my students, but they’re not the same as other faculty’s expectations so I don’t want to give you specific numbers here. The good news is that, with your projected participant numbers, it sounds like you’ll get a reasonable number without a problem. (FYI, narratives often use as few as three people, and special ed research may examine one student, albeit while speaking to others about that student.) And to answer your last question, yes, many students have used small numbers of participants and succeeded and, at the same time, many qualitative researchers (not just students) have struggled to get a large enough number of participants to complete their studies. You are not alone in this.
I hope this helps,
Paula Dawidowicz27th January 2012 at 2:59 pm #2764Fred GarlickMember
Calculating sample size can be tricky because one needs to quite a lot about the prevalence of whatever it is you are looking for in the population and of course what it is you hope to produce as an outcome. But in this case it does not seem sensible to think that way because you seem to be using small case studies or vignettes and they don’t naturally lend themselves to statistical processing. What you can do is describe each oft the vignettes as away of explaining domestic violence in these case and the possibly draw some general conclusions but the value of this kind of research is your evaluation of the cases you looked at because I doubt it will be possible to draw statistically valid conclusions because I guess that each case is more or less unique. In short it is not necessarily the volume of cases but the time and quality you put into the ones you have.1st February 2012 at 2:47 pm #2763
thank you, your response is very helpful. I knew that numbers were never going to be vast (and with the method adopted not necessary or appropriate) and also somewhat unpredictable from week to week. More a sudden state of panic about numbers. I know I can justify the numbers in light of the choice of method and it is good to be reassured that this is a common issue. Some good examples of similar sized projects within my literature review will help I think. Thanks again for the very quick reply.1st February 2012 at 2:51 pm #2762
thanks for your prompt reply. I realy like what you say about vignettes – unique stories for each person that might overlap with each other (that I do see happening within the group). The comment about time and effort is also very helpful. The method certainly seems to allow the partciapnts to show their stories as well as tell them, which in itself seems to be beneficial for them (a whole other are to discuss). Thanks again for the quick response.
- The forum ‘Default Forum’ is closed to new topics and replies.