I agree with your sense that most versions of grounded theory would prefer to alternate data collection and analysis. Beyond that, there is increasing convergence among a number of approaches about how to handle the data once it has been collected: begin with line-by-line coding, group those codes into substantive categories, and then develop a more conceptual summary.

That admittedly crude summary would cover both Thematic Analysis and Framework analysis. In particular, since Braun and Clarke's version of Thematic Analysis assumes that all of the data is collected prior to analysis, that would be a reasonable alternative for you consider. Google scholar gives their 2006 article more that 15,000 citations, so you would hardly be alone in making that choice.

Keep in mind the Grounded Theory mantra "all's data". So do not limit yourself to your interviews. You can use documents, newspaper clips etc as far as you need it with your ongoing analysis.

I agree with an idea of @Charles Berg. Dear @Raman try do not be extremely "procedural" as M. Hammersley once said. 

I agree with your sense that most versions of grounded theory would prefer to alternate data collection and analysis. Beyond that, there is increasing convergence among a number of approaches about how to handle the data once it has been collected: begin with line-by-line coding, group those codes into substantive categories, and then develop a more conceptual summary.

That admittedly crude summary would cover both Thematic Analysis and Framework analysis. In particular, since Braun and Clarke's version of Thematic Analysis assumes that all of the data is collected prior to analysis, that would be a reasonable alternative for you consider. Google scholar gives their 2006 article more that 15,000 citations, so you would hardly be alone in making that choice.

I think content analysis will be helpful too

I think content analysis will be helpful too

Since you have collected the data, GT is not suitable , based on recommendations in literature. You can think about the following options:

1. Thematic analysis proposed by Braun and Clarke, as suggested by prof. Morgan.

2. General inductive approach proposed by David R. Thomas.

Both approaches are well explained in the articles available for free download.

Dr. Morgan and others

Really appreciate your responses. Wonderful advices. I have three follow-on questions.

1) Is Thematic analysis a respected / favoured method for Doctorate research? I was told that sometimes TA is not considered that rigorous a method for doctorate level research thesis. Is this a misconception? 

2) On the other hand, I have data of 74 interviews and over 30 documents in total and I was also told that for such massive data and scale, TA is an appropriate choice. Please guide if that's a general understanding in the research community (and hence would be of examiner too). ,

3) Further, I am looking to create a graphical model or some kind of visual model of patterns through   data analysis. While GT inherently expects that kind of model as output, TA does not emphasize the outcome in that format. Can I still use TA and report results in the form of a graphical or visual model?

Any guidance will be highly useful. As usual thanks to Dr, Morgon doing great job in guiding new researchers like me. 

Raman

Before analysis, I'd start by looking for where you have missing pieces of data.  With any type of qualitative research, as you have indicated, you collect data, analyze and then collect more data based on the analysis.  As you look at the early interviews now, what data from the interviews would you have pursued if you had had more time?  When you find the trends or concepts with missing date, if you cannot find participants, I support Charles Berg's suggestion about looking at other sources of data besides interviews.  Triangulation is always a good idea with qualitative research.   In one study I did of previously unconscious patients, for example, a few subjects told me they felt cold.  It went over my head at first.  Then when I had another subject say they felt cold, I asked him to elaborate on feeling cold.  He was freezing, he said, like below zero cold. When I studied critical care nurses, one nurse would describe her approach but might comment she knew a nurse who had a different approach.  I'd go seek out that other nurse. 

From what you've described those are  the type of process piece that might be missing.  If you find the thread or elaboration through other sources that would be helpful.  Also any missing pieces should be reported in your write up and recommended for follow up in other research studies.  It is much better for you to acknowledge missing pieces of information that could have been drawn out.  It will show your committee you understand the limitations of what you have and what in more ideal situations could have been done. 

Good luck

Yes, TA is widely recognized and acceptable, as demonstrated by its 15,000 citations.

In terms of the number of interviews, etc., that is a large amount of data for any form of qualitative analysis. Certainly you do do it with TA, but it will be a lot of work (which would be even more so with Grounded Theory).

Diagrams may be common in GT, but there certainly is no limitation to using them there. I have attached a reference on using diagrams in the equivalent of TA (although it predates the Braun & Clark article).

Dr Morgan. I had exactly same thing and same article in mind which inspired me to have diagrams in TA, Thanks for pointing out the dates. I think it is great reference. Thanks for the information. Very useful as always. 

I think you can still use coding, but maybe you have to switch from GT thinking to other perspectives like discourse-analytical coding. Regarding the amount of your data and the question of visualization I think you could use a software like ATLAS.ti. I had good experiences in organizing and analyzing data. You can also handle different data sources like text, pictures, audio and video with the software. There is also a visual tool for networks of codes. This was really helpful for my analysis and I was able develop first visual models of the data.

I agree with Moritz and think you can still use coding. I had a slightly similar situation recently and was guided by my academic supervisors to continue with GT and Discourse Analysis but make claims only to using this "style" of analysis methodology. Mine was a pilot study and therefore I was aiming at a tentative explanation of my observations. Colin Robson in "Real World Research" 3rd edition 2011 talks about this sort of approach and more widely about these sorts of challenges in the 'real world of research' which I found very helpful.  As an aside, I used NVivo to analyse rich and varied data from children and again like Moritz found it really helpful.