Hi,

I have three studies, each of which contain their own discussion section that goes into detail. Conclusions and recommendations are made in each chapter too.

I'm just wondering in that case what goes into my overall discussion chapter of the thesis? I can bring all three studies together, and come up with some synthesized recommendations, but that's about it. I don't want to repeat everything I've already written about earlier.

Any ideas? What are you doing/plan to do?

Sounds spot on. Also reflect back to your introduction chapter/literature review to make sure that they complement each other in terms of the story you're telling overall.

It's a good place to highlight any common themes that emerged in the actual research processes, and highlight any challenges if you haven't done so already.

I'm planning the same approach - bringing it all together so it fits more as a 'body' of work, looking back to the literature review to see how/if my studies helped address the gaps, limitations of the work and suggestions for future research.

Hi I have a question that is sort of on this theme. I've also got three separate data chapters and have written a discussion chapter to link them. When my supervisors went through the final data chapter and discussion chapter they became a bit overwhelmed and made some suggestions to simplify. After a lot of thinking, I came up with a sort of simplified paradigm to help with the third (main) data chapter that has meant that I would reshape the discussion chapter (not writing new stuff but presenting it in a different format based on the paradigm that reworked my 3rd data chapter).

Now when I skim through the 1st and 2nd data chapter, I can see all these links and threads that foreshadow my organisation for the final big data chapter and discussion chapter. What I am now wondering, is do I need to go back to these earlier chapters of data and reorganise them along the final data chapters paradigm, or can I just highlight the 'threads' a bit more. Is it okay for these to be on the same topic but be organised quite differently based on the surveys and questions. There is no way I would have written these originally using the final organisation, as I had to do all the research to get to this point-if this makes sense. From your comment Hazy Jane, it seems as if I can leave these two earlier chapters organised a little differently, as long as the whole thing is cohesive and the discussion chapter makes all of the links? (Hope this all makes sense-today I am pulling my hair out and was really relieved to see that you had asked this question Hugh).

Pjlu, what you describe sounds correct to me.

May I ask which paradigm you used? You could pm me if you prefer. It looks like you've found a lovely solution :)

Thanks HazyJane and Chickpea.

Based on what you have said, I think I should write up my literature review and introduction before I tackle the final discussion chapter and conclusion.

I've also seen some thesis that combine the final discussion and conclusion into one chapter. And I saw one thesis that skipped overall discussion and went straight to conclusion.

I'm planning on making my conclusion very short, because I think I'll just waffle and repeat parts of my discussion if I aim for anything more than that.

That's a good idea, I think I'll do the same too. I think I might combine the final discussion an conclusion into one chapter. I'm worried about repetition.

Hi Hugh, by paradigm I mean an organising or conceptual framework for my findings. My study is a mixed methods and uses basic descriptive statistics plus qualitative data through surveys and questionnaires and then has a chapter that includes purely qualitative data from participant interviews. I've used a social constructionist/interpretive epistemology overall and thematic analysis techniques (Braun & Clarke, 2006) for the qualitative data. However, the first two data chapters included most of the descriptive statistics with some qualitative stuff while the 3rd chapter data processing took ages-coding, creating tables, aggregating masses of codes into categories and then merging categories into themes. I've worked a while on trying to link the data chapters, especially as one was a bit different-a sideways look at the topic (a larger survey I used when I panicked and thought I wasn't going to have enough data)

I originally wanted to put in everything but could see that from an outsider's perspective, it was just too busy and too complex. So my supervisors questioned my major themes a little (and I began reducing them). Then all of a sudden I saw that really I could simplify both my themes and my original research questions and once I could do this, I could take all of the data and reorganise it into the new themes and take these and the collapsed questions and use these as organisers to contain all of my categories and data more effectively in the discussion chapter.

It has meant a lot of reorganisation but doesn't change the findings or rather it sort of puts it in a different light. What I was agonising over was whether I then had to go back and do this with the very detailed descriptive statistic chapters-or whether I could just highlight the themes a little which were emerging even then, without me really realising. It all sounds exciting but it isn't really-but I think it will improve the final product. I have actually completed a full draft including conclusion and am writing final drafts of each chapter. Good luck with yours btw and hope this is helpful.

This is an excellent post just published about writing up the discussion chapter

Thanks for the link, Hugh - I've never used a 'diagram' approach to deciding what should go into a chapter - might be worth a try!