Guidelines for amount of qualitative data


Hi all,

Firstly a thank you. I have lurked these boards for three years and your discussions have been so helpful. I am coming up to my three year deadline in October. Unfortunately I'm going to over run. :-( I have written a first draft of four chapters out of eight and still have a load of analysis and the bulk of my literature review to do (is that really far behind?!).

Anyhow, I've had some feedback on my thematic analysis, and my supervisor has asked me to justify the amount of data I acquired. To be honest I just kept getting it until it looked like a lot - so I have no idea what to say! Are there any guidelines about the amount of data needed? For some context, I gathered responses on an online forum, and used thematic analysis to gather categories of interest for the following chapter so it was only a preliminary study (hence I'm worried I may not have collected enough).

Thanks for any help. PS, are all the other 3rd years way ahead of me? How long does it take to do the lit review? My supervisor told me not to write one in the first year - thought that was great at the time, but really wish I'd gone against him now!


how long is a piece of string???? Really I think it depends on the topic and how you have gone about things. I have done 60 interviews and ten observations and it hasn't really come up if that this is enough or not so I'm presuming that it is! A colleague has done <30 interviews and has just submitted. He has been told to be ready for questions about the smallish numbers of interviews he has done in the viva; if he can defend it okay, that should be enough. I know somebody who only spoke to five people but they were repeated interviews, really life histories.

Why not read up on the concept of theoretical saturation to justify what you have? Also, is your methodology sufficiently innovative for you to claim that whatever number you have is enough?


Thanks for you reply. In a way it is probably good that there is no guidelines, but I just worry about justifying everything. I feel a lot more prepared if I have a reference saying something is acceptable - but I guess this is related to my own lack of confidence. I'm a quantitative researcher by nature and this qualitative chapter has driven me crazy. I've enjoyed it, but I'm constantly second guessing myself. I think I'll ask my supervisor to elaborate on his comments just to be sure of what he meant to. Thanks again.


My data was collected both online and by getting others who went to meetings to hand out and collect questionnaires for me. I've got 145 questionnaires, plus a mass of other data (somewhere in the region of 5,000 separate entries, but I am only going to use about 400 or so of those (It took too long to get even those off and anonymised to even consider doing more unless someone says I ought to do that. They are all separate e-mails about subjects and have to be divided into categories and then subject and answers so that the type of question and the number of answers produced, plus the origin of the original question can befound and related to other things, it is a long, long job:$) and I've also got comments from others. However the total population I'm looking at is relatively small so I have about about 3% of the total population for the questionniares, and a similar number for each of the other sources . Others who have worked with other related areas have published results that use only between 5 and 50 of a population of around 450,000 and that seems to be OK. One ofthe students here has been told to put in something like 'the size of the sample used means that the results may be viewed as an indication of xyz in the total population' to make it clear that you realise your sample may be smaller than that which could be obtained by others with more contacts/money/time etc. although if any of your findings can be linked with other research this may of course add more weight to your findings, and of course the size of your sample compared with the population in your particular pool may make it more robust. I think the answer is the sample is big enough when you have enough to analyse it in the way you want, and therefore will depend on your individual study. If anyone queries mine I shall just point out the number of people in these similar surveys to the size of their potential pool and make the comparison.

I have just over two years of a part time PhD and would have been a bit further along if I had not had a very rocky year due to supervisor problems, I've three chapters nearly done in draft/ close to finished form, plus another two fairly advanced and two more started. there will be one more and the conclusion after that. Doesn't sound nearly long enough!!!! :-(. My lit review is ongoing too as I'm writing it as I get to each chapter and decide what to include in it. Which is good really as a fair bit of what I put in to start with has been left out - all that reading too, which is a bit of a shame. How long it will take depends on how you are writing it, my thesis has a lot of comment on the literature all the way thorugh, so my lit review is basically what I chose and how I found it, together with what I think is the main contribution to this research, what papers etc. I left out and why - still looks like ending up as way too long and will need a lot of pruning to cut it down to size. Yours may have a different approach which may affect the length.

Avatar for sneaks

I have 40 interviews and then I also have a questionnaire with 350 respondents. I think it just depends though - I know someone who had 5 interviews but analysed the life out of them and get the PhD.


Quote From sneaks:

I have 40 interviews and then I also have a questionnaire with 350 respondents. I think it just depends though - I know someone who had 5 interviews but analysed the life out of them and get the PhD.

I did 15 interviews and analysed the life out of them (and got the PhD) as no one had ever analysed interviews from this group of stakeholders!


Thank you for your replies everyone. I feel much more confident about my results, as I have done a huge amount of analysis on it (my supervisor suggested perhaps too much).
Also thank you Joyce for your comments on the literature review. I think I really need to look at restructuring the whole thing. Perhaps when I've done that it will come along a bit easier. :-)