GT info help required - Olivia??


I used GT for data collection and analysis. My supv is arguing (with me!) that the 3 sub categories I identified are enough and that I don't need a core category. But, I think I do! After the worry of 'would I identify a core category' I am actually genuinely satisfied with the one I extracted from my data. How can I argue for my substantive GT theory without having a core category? How would it go down with external examiners if my theory formed from sub categories, not core?

Any GT people out there???


Hi Ady
I will do my best to try to give an answer. There are different strands of grounded theory out there- and it sounds like you are using the Strauss-Corbin methodology. Their books stress the importance of the central or core category ( they seem to refer to these interchangeably). I much prefer the second edition of Strauss-Corbin to the third edition for describing the use of a central/core category in producing a theory from your research. They explain ( pg 146, 2nd edition) that "A central category has analytical power. What gives it that power is its ability to pull the other categories together to form an explanatory whole."

Roy Suddaby has an excellent article on "What Grounded Theory is Not" ( 2006) in The Academy of Management journal. He argues that it is very important for grounded theory research to adhere to the methodology of grounded theory ( the phases of coding and abstraction to arrive at your own grounded theory). Otherwise, research that claims to be grounded theory is in fact, NOT. It might be inductive and theory generating, but it is not "grounded theory". Therefore, to be doing grounded theory, you need to adhere to the methods of grounded theory research.

And if you are following Strauss-Corbin, you need a core or central category. You simply cannot go on with the "selective coding" phase of your work without this as they say, "The first step in integration is deciding on a central category" ( 2nd edition, pg. 146). Without this central or core category, how do you then proceed with selective coding?

In other words, at some point in YOUR inductive work, YOU have derived a central or core category that is a fundamental piece of your eventual grounded theory. Cite Strauss-Corbin 2nd and 3rd editions---you NEED this category to emerge in your selective coding phase. That should be end of--you are adhering to the methods of grounded theory work, and no one who has even the remotest familiarity with grounded theory should even question the presence of your central/core category.

Your supervisor, to me, is interfering in your inductive work. It is YOUR inductive work, not hers. Maybe she would have arrived at different categories or conclusions--so be it. Inductive work is not the same as deductive work--and grounded theory was developed to get AWAY from the heavy positivism infused methods of the day.

Grounded theory is my opinion is tainted if you have someone trying to impose a category ( or lack of) from the outside. This is not a process that arrived out of your data study--it is a supervisor trying to exercise control over your inductive research. That spoils and corrupts the methodology.

I think that you may simply need to stand back and hold your ground ( no pun intended) with your supervisor.

(contining in second post as I appear to be getting long winded! :$


I recall a somewhat similiar showdown with my own supervisor over my grounded theory methods--and I just said, in essence, "Back off. I am doing the methods with close attention to the ways in which they should be done, and if you interfere, you are messing with the process. Moreover, this is MY work and not yours. Go away."

I think its valid for a supervisor to ensure your processes adhere to grounded theory work--but certainly NOT for them to interfere in the content of your coding. You have not strayed outside the acceptable ways to perform grounded theory work--you are sticking to the very "classic" Strauss-Corbin methods.

If you are of the mindset that grounded theory is not simply any old inductive research or review of interview data but rather is a methodology that follows a set of phases of coding--then stick to your position over the work. Politely and firmly tell your supervisor that perhaps you and he/she will have to simply agree to disagree, because your work relies ( per Strass-Corbin or whomever you are using for the GT) on the presence of this category--that you respect his/her opinion, but that you also feel that they are interfering in your data analysis and mucking up the process by so doing....



Thank you so much, your reply is more than I could have hoped for :-).

You're right, I am using Strauss and Corbin not Glaser and Strauss. I own the 2008 edition but only got the 1990 edition from the library so I decided to stick to 2008. There are a few differences but I thought it better to try not to mix the two.

To be fair to my supervisor he used Glaser and Strauss in the mid eighties so he is a GT person but he is not familiar with the spilt in the two camps or the later editions of the GT method[s]. So yes, I will stick to my guns, you're right. He says I'm too apologetic overall so it's something I need to work on come the viva - not one of my fortes!

Thanks again Olivia for such a speedy and utterly helpful reply - much appreciated. I have 'starred' you in the past so can't do so again I'm afraid but please accept these virtual gifts.



Hi Ady, glad I was able to help. GT has different "churches" it seems, and people can get quite attached to one or the other sides...even Glaser and Strauss fell out over how to do it in the end! If I recall, Glaser did not like the way in which Strauss-Corbin took things, and then enter the constructivists like Charmaz!!! I do not know if you cover, or have considered covering, the development of the different strands of GT--I did this in my methodology chapter, justifying my own use of the constructivist method, but showing an understanding of the methodology as a whole as well. Along the PhD journey I encountered people who got very wound up over my non use of Strauss-Corbin and using constructivist GT...shrug...I would say, unless they were willing to discredit the entire constructivist GT, then there was very little to respond to as I had chosen the constructivist strand for reasons XYZ...

Some of them indeed would have liked to have discredited the entire constructivist strand, and I would just say, well, that is beyond my own thesis, and the constructivist strand is certainly academically accepted, so....shrug...

More of agreeing to disagree...

And thanks for the virtual gifts--they are lovely :)


To update - I stuck to my guns (metaphorical ones) and argued for how I have done my GT research, paraphrasing you heavily Olivia! Didn't quite use the words "back off" but told him that to dispense with the core category was "going too far away from key tenets of GT etc etc. My supervisor listened and replied "Fair enough"

Me - :$


======= Date Modified 24 Apr 2012 03:59:03 =======
======= Date Modified 24 Apr 2012 03:55:41 =======
very nice discussion...even I am using GTM...I have used Strauss & Corbin...I am stuck...coz I have four categories( core categories)..and see most of the initial codes...relate to all of there any problem? Secondly...if I do comparison within these four categories...central phenomena becomes it ok to label this as the core category?
can you suggest me some literature regarding it? Thanks.


Hi, I am struggling with kinda the same thing but in my case it is around the need for a core category when drawing upon constructivist grounded theory approach? I am really not sure whether this is appropriate, or whether I am methodologically blurring earlier approaches to grounded theory with Charmaz? Would be most grateful for any guidance. Many thanks


======= Date Modified 02 May 2012 08:53:49 =======
Hi Ady and redridinghood- I'm also using Charmazian GT and try to develop a core category, which is not easy. Saying that, Charmaz is quite relaxed about the whole core category thing- she seems content with establishing the 'basic social processes' pg20. I think when taking a constructionist/interpretivist position you acknowledge that there are multiple truths and realities, rather than one 'true' category- this is too absolute and perhaps fits with the Glaserian (objectivist) GT.

A core social process is useful when trying to organise and sort your theory. My supervisor said 'if there was one word that you could use to describe what are participants are say/doing etc, what would it be- and that's your core category'.... I still haven't constructed it yet!

I'll probably still persist with trying to find one, but I feel I could argue in viva if I had 2 or three core social processes...



Thanks PhDee, gosh this thread seems like an age ago. I am now post viva and indeed there was a 'discussion' in the viva about the core category!


Oops should have looked at the date! Can you share some details about what you were asked in relation to the core cat:-)


Quote From PhDee:

Oops should have looked at the date! Can you share some details about what you were asked in relation to the core cat:-)

discussion in viva pretty much mirrors that which I describe in my original post [ie pre viva]:$. i was much more pro the absolute need for a core category than they were! However, it was a discussion, to and fro between them and me and one which I think I handled okay.