Grounded Theory - analysing interviews

Avatar for Noctu

Hi guys,
I'll be starting my recruitment soon and will be using GT as my methodology. I'll be conducting semi-structured interviews and observations.

For all those who used GT: I have got it in my head from somewhere (!) that you must transcribe and analyse each interview BEFORE conducting the next one? i.e. make it a completely iterative process which allows the interview guide to evolve based on the findings from the previous interview?

Or am I completely wrong and it can all be done 'whenever'? I'm trying to work out my schedule of interviewing and am thinking that if I need to analyse each interview before I can conduct the next one, I'll probably only be able to conduct 1, possibly 2 interviews a week...

Oh also another question, how have you guys got around the premise that one should not read any literature before embarking on the research (yeah right!), I've heard of pragmatist grounded theory but ideally I'd like to stay as true to 'traditional' GT as is possible in modern day funded research ,-)


======= Date Modified 12 Feb 2012 20:22:54 =======
======= Date Modified 12 Feb 2012 15:57:08 =======
Hi Noctu

I used GT and also did semi-structured interviews and [non-participant] observations!

I didn't interpret things quite the same as you - while yes, it is an iterative process etc etc, if you were to transcribe and analyse each interview before proceeding to the next, you would never be finished, imo. I took it to mean that transcription and analysis happened concurrent to interviews and observations taking place. My understanding of the literature is that while of course you shouldn't use GT in an 'a la carte' manner it is supposed to be sufficiently flexible and iterative so as to be 'iterative' and flexible!! Transcription can be pretty sole destroying and if I didn't have the 'carrot' of a real live interview to look forward to every now and again I would have gone stone-mad! Other forms of analyis require that all methods are conducted before analysis can start, so in that way GT is different.

Re: a purist GT approach - what you are describing is the Glaser way of doing things, ie to hold off reading the literature so that you go into the field with a 'tabula rasa [blank slate]. I argued against this and chose the Strauss-Corbin version whereby I felt they were more realistic and accepted that rarely is anybody a blank slate and inevitably you bring prior knowledge into the field.

There are some good readings on GT (recommended to me mainly by Olivia of this forum). Most, if not all are freely available but come back to me if you can't find them:

Roy Suddaby - What grounded theory is not
Karen Wilson Scott - Relating Categories in Grounded Theory Analysis : Using a Conditional Relationship Guide and Reflective Coding Matrix
Corbin and Strauss - Grounded theory research: Procedures, canons, and evaluative criteria [I can email you this one if you need it]

also if you're not aware of, have a look at glaser's website -

Edit: I was asked quite a lot about GT in the viva but more in relation to my core category rather than the process per se. The examiners seemed to accept that I had done it correctly and didn't query why I followed Corbin and Strauss and not Glaser and Strauss (1967) seminal text that Olivia mentions. I must have explained that much sufficiently at least :$ I enjoyed the GT discussion bit of the viva the most!


Hello and great choice of methodology! :D Grounded theory is a great methodology to use, and I think you will be amazed at the results you get from using it. That said, you are right to want to get the steps right on its use...otherwise, you are not really doing grounded theory, but some weak imitation. If you want to stick to the "classic" or "traditional" grounded theory then you should get the original texts on it by Glasser and Strauss-- also be aware that they later had a split themselves, and Strass and Corbin went on to develop further strands, as have other people. So be clear which strand of it you want to use, and stick close to the texts on that when designing your methods.

I think what you are asking about relates to the "constant comparative" part of the method, as well as theoretical sampling and theoretical saturation. These all have to do with data collection and data analysis as an ongoing part of your work, occuring simultaneously. You should have worked out what you are going to do with the data from the interview ( ie take notes, or do word for word transcription, and from there, what steps you are going to go through in your analysis of the data as you continue to collect it-- interviews being data).

One question comes about in knowing when you have theoretical saturation, have you done enough interviews? Does your data allow you to start to form a theory ( look in the original texts for the multi-level analysis that you do of your data to develop a theory) and if there are gaps in what you are putting together, do you need to do interviews on a specific part of your data, etc.

Be sure you have worked out the steps of your methodology BEFORE you start your data collection. Be very familiar not only with the steps of the strand of GT that you have chosen, but the philosophy and aims behind it. Only then, IMO, are you ready to start interview data collection.

Avatar for Noctu

Thank you, thank you, thank you... both of you!

Funnily enough my supervisor is an 'expert' on GT but I don't want to ask her for fear of looking stupid (I don't want her to lose confidence in me, silly as it sounds), and sometimes books just make you feel like you're wading through academic syrup rather than explaining it clearly and fairly painlessly 8-) I will check out those texts Ady and I think I'll formulate a clear step by step plan of exactly what I'm going to do once I have my interview and observation transcripts. I may be asking for help again though lol, hope you don't mind ;-)