Hello all:

Sorry to bore you but this was a last resort in case there was even one person who can help as I am quite literally about to give up on my phd...which is devastating.

You see I collected stories from participants who had reported that exercise/sport was a part of their recovery from addiction (recruited them by asking, "has exercise played a role in your recovery" and interviewed them by simply asking them to tell me about their experiences with addiction, recovery and exercie/sport and then probed them on topics they raised). I have transcribed about 40 interviews lasting 2-4 hours which I felt really spoke to the richness of the data, but as discussed below this may also relate to me trying to do too much in one study.

My original conjecture was that identity changes associated with adopting an athletic identity might support that recovery and I was going to use identity as a central sensitizing concept--while being open to other arising themes-- in perhaps a grounded theory or narrative approach. My tentative 'grand tour' question was 'how participants conveyed a sense of who they are in their sport stories, versus their stories of addiction'

Now I have come to the realization--albeit obvious in retrospect--that I am looking at two processes (i.e., recovery and exercise/sport involvement) that could and have been considered in isolation (and have) in other studies adopting a grounded theory approach (which generally focuses on one specific process not two) and even identity change is typically studied in the context of one process (e.g., recovery) which simplifies the process of making the necessary links to theory, and generally allowing them to make generalizations across the data on that basis (e.g., typologies of identity changes).

In contrast, I fear that I have collected such a complicated sample from the outset (involving recovery and sport involvement) that I will not be able to tie the data and theory together in a logical whole. Just as an example, I am left wondering, "how do I account for the fact that some were very involved in some forms of treatment (with consequences for their identity as well) while others were not", "how do I account for the fact that while all identified exercise as part of their recovery, some started exercising before their recovery while others started exercising to support their recovery once it had already begun". As you can see, it would have been a lot simpler to have stuck to looking at one defined process and now I fear that in my overzealousness, and due to a lack of methodological foresight, I have dug myself into a hole that I won't be able to get out of.

Does anyone have any advice?

P.S.: while discussing with my committee, I also fear that by pointing to these problems with my sample I will effectively be cutting off any last opportunities there might be to save things by framing my sample, approach and research question in terms of what I have collected.

Please don't give up! You have soo much data! And you seem like you are interested in the project. this is all part of the PhD proces I'm afraid.

I'm not really in your area, but have you considered doing a simple thematic analysis to see what the key themes are? that is usually a good place to start.


It all depends on what your questions are. It could be that you can use completely different analysis techniques?

You seem to know the weaknesses of your work, which is a good thing; your examiners perhaps won't see those weaknesses as rubbishing the whole study in the same way you do.

What's your supervisor's take on this?

This is too specialised for me to offer detailed advice, but I echo the question about what your supervisor's views are. My research uncovered different things from what I was expecting originally, and my supervisor helped me to turn things around to make the most of what I'd found. Could that be possible in your case?

Sneaks, I was going to say that myself. Correct me if I am wrong, but you have used open-ended interviews with verbal probing to generate the data - did you purposively sample? Sorry, if that sounds a bit obvious. 40 interviews of 2 - 4 hours duration is an incredible amount of data, so you can't give up now.

Why not adopt template analysis to identify the themes for the 2 processes that you have identified and then corroborate your findings with the other studies to validate them and perhaps extend the findings?

You just have to find a methodology to fit your data (I know it's usually the other way around), change your research questions and, though it may be tough, you'll be okay. Template or thematic analysis is a good way of doing this because its not as rigid in demand as many of the other qualitative approaches. I was going to use it for my research. I've got some papers and would be happy to e-mail them to you.

http://www.hud.ac.uk/hhs/research/template_analysis/intro.htm

Karl,

you could see it the other way round: it is actually the strength of your project that the data provided you with other information as you had expected. Why do a project if you know already in advance what will come out of it?

You have a large amount of data, now it is a question of presenting it. As others have mentioned you could just describe main themes that come out of it. Overall you can conclude that the whole concept between addiction and sport is non-liniair and complex, this project have provided some insight in this complexity and that has only been possible due your open minded qualitative approach!

I totally agree with Rick.

I really recommend you to read carefully the books I mentioned in the other thread (or othe books dealing with the grounded theory). They explain precisely how you should work with your data. Btw, you will be able to read them partially in books.google.com

You should forget existing theories and preconceptions now and start coding and organising the data and creating concepts. By doing that, the theory arises from the data. At first, do not expect to find anything.

After the analysis, you can link your results to existing theories. It has been even suggested that one should write literature review after the research process in order to prevent existing preconceptions to interfere with the data and results originating from it.


Well, this is only one perception of how a reseach process could work.

First of all I want to thank all of you dearly for your comments, encouragement and advice.  They really struck a chord with me and I followed up on each and every one of the leads you provided. Yes it is an area I am very passionate about, both professionally and personally, and so I am trying to remind myself of that in the hopes of moving forward. 

After examining all of the options you have provided I have decided that perhaps taking the grounded theory route would be best in turning this into a case of making new and unexpected discoveries.

The only thing I am a bit worried about still is the fact that most grounded theory studies seem to be framed around one process that they then study in detail in order to arrive at some general conceptualization about (e.g., a typologies of processes of recovery from addiction) . 

By considering the specific role of exercise in recovery from addiction, however, it seems as though I have actually complicated things by taking on two processes (i.e., recovery and exercise involvement) and/or have taken my study to a substantive level that might make more general conceptualizations more difficult and/or might lead me to focus more on the nuts and bolts of how and why exercise might work. Of course, this substantive focus came as an innevitable result of my specific interest in this topic that would not have been possible through a focus upon more general processes...so it is a bit of a catch 22.  Based upon an initial scan of teh data, I am worried that I will be challenged to elevate it to a broader theoretical level as I will be mired in thousands of descriptive codes as to how exercise might fit into recovery.

Does that make sense to any of you? 

If so do you have any suggestions as to studies that have tackled this or perhaps ideas as to how I might frame a grand tour question that would allow me to avoid this trap.

Am I making any sense at all?

Thanks again!

P.S.: it was a purposive sample in the sense that I did recruit people based upon the fact that they had recovered from alcohol or drugs and felt exercise/sport had played a role in this

Karl - are you planning to do further studies?

My PhD - there wasn't much in the area, so I did a qualitative study very similar to yours in order to 'set the scene'. I analysed the data in two different ways from my first study, and have gone on to build the themes into a questionnaire, for my second (or third if you count the data analysis x 2) and am doing a mixed qual/quant approach for my 3rd/4th study.

So what I'm saying is that this doesn't have to be the be (bee? bzzzzz) all and end all - it can be a building block or a scene setter. YOu can then go and expland your work or use it as a way to explain the process of your phd i.e. this study made me stop looking at this cos there was a dead end, or this study made me realise these important factors so I therefore went on to.....

I hadn't even thought of that Sneaks! Looking at as potentially one stage of a larger process certainly leaves the door open for more opportunities and in doing so may also take the the pressure off in terms of me thinking I have to have it all nailed down in the first go. Having said that it would still be nice to be able to find a way to milk the data I already have, particularly as there is so much of it....over 30 interviews at 2-5 hrs per interviews is a lot of paper! Also, as deadlines loom near, I am not sure I have time for more data collection...hmmm, will have to give it some thought. Coincidentally, what is your research on Sneaks?

What deadlines do you have? Is it for the entire completion or upgrade or something similar??

I would say you should definitely look into analysing the data in two different ways, as you have so much of it. My area, without completely revealing my identity is around a certain section of society and their social mobility kinda (sneaks being very vague here)

At my university 90% of candidates who submit a thesis get their PhD, but only 50% submit. So, if that's true elsewhere, even if you think your approach has flaws, just write it up as best you can and submit something. All projects have flaws; the important thing is that you acknowledge them. There's no such thing as a perfect thesis.

Just my $0.02

You might want to take a look at another thread which relates to 'Failed PhD - Any Advice?' You will read about a litany of academics causing significant problems for students in reserach. You've learned useful techniques and know something about how to write up results at this point. Please take a look at this thread and be shocked at the closed mindedness of universities towards students when it suits their own political agenda. If you genuinely feel that you can't go on...then don't, why hurt yourself? The PhD doesn't define you as a person, its a part of who you are, you're a hard worker and obviously clever.

I think you should choose a narrative approach that focuses upon how exercise has contributed to the recovery of the participants (if indeed exercise has). The narratives could be shaped by Arthur Frank's 'The Wounded Story Teller' - which deals with Restitution / Hope / Chaos. The stories of your participants should be able to be applied to a narrative, although they will be shaped by time and space.

You have so much data, that you should easily be able to manipulate it into a thesis. The work of Andrew C. Sparkes and Brett Smith on Spinal Chord injury amongst rugby players would also be worth considering in demonstrating how your findings would look.

Two books worth reading are 'Narrative Methods for the Human Sciences' by Reissman (the standard narrative text) and 'Analysing Narrative Reality' by Gubrium and Holstein.

Best of luck with it.