Help! Mixed methods people!

Avatar for sneaks

======= Date Modified 26 Jul 2011 13:19:19 =======
I have 3 studies....

Study 1 = fine
Study 2 = fine
Study 3 picks up on an idea found in Study 1, but because of this idea I went ferreting about in the literature and it also tests a key hypothesis from existing literature.

I've kind of tried to introduce these studies like 'and study 3 builds on an idea from study 1' - I wanted it to be a bit of a 'reveal' - so it looks like study 3 was decided upon AFTER study 1 if that makes sense, so it wasn't pre-decided when I started the research.

Now my sup reckons I need a research q for each study. So I'm now confused - if I put the researchq for the last study in my intro, then it will look like it hasn't built on the ideas of study 1 but it was all decided upon before I'd even started my first bit of data collection. :-(

So I'm not sure how to introduce study 3 and when/where to do it and in how much detail - should I be upfront about what it is about from the beginning (and therefore telling the reader the findings of study 1 before introducing study 1?)

If you're following any of this and can help, please do.


Two points (which sort of pull in opposite directions, but may help):
(1) My sup has actively talked me out of the idea of any "reveals". When I protest that this will make it boring, he says that thesis examiners aren't looking for a thrill, and, even if they are, they want to be thrilled by the quality of the work, not by a surprise twist. The number one thing they want is clarity, i.e. (a) this is what I'm going to do (b) this is me doing it (c) this is what I did. And each of these parts should fit together coherently so it's not a case of (a) this is what I'm going to do (or is it...? wait and see) (b) I was doing it but then something exciting happened (woo!) and then I went down a totally different road (c) look at where we have ended up - what a magical mystery tour!
(2) However, I think it is true that to achieve this coherence you have to be faithful to a certain extent to the shape of your research process. It needs to make narrative and logical sense.
So, what I would do in your situation is say (to give a mad example):
* In study 1 I tackle the question of why icecream vans make less money on rainy days. In the course of this study, as shown in chapter 2, I discovered that, contrary to the existing literature, the preponderance of pistachio icecream in vans is a contributory factor to the climate/profit relation.
* In study 2 I approach the question of why fast food vans at festivals charge £4 for a bottle of water.
* In study 3 I build on the discovery made during the course of study 1 regarding pistachio icecream, to compare nut-content, climate, location and profit in mobile catering.
Could this work? Give some sense of what happens in study 1 without going into huge amounts of detail, so that the question for study 3 doesn't appear to come out of nowhere.

Avatar for sneaks

ah I'm SO stealing all of your ice cream ideas!

Yes that would work - Its kind of like what i have, but probably is a bit stronger - at the moment mine's like 'study 1 looks at xyz, study 2 looks at abc and study 3 looks at some of the things from study 1' = a bit lame!

So yes, I maybe need to strenghten it - I'm about to write out my reserach questions and then I guess I need to sit down and write sections for the lit review, method, each chapter and then the discussion related to them


introduce the questions for study 1 and study 2, then say something like study 3 developed from the analysis ofthe results of study 1 which revealed a possible connection with the study by X, and put the ref for the study you found. I wouldn't think of doing too much of a 'reveal' examiners probably like things simple rather than embarking on a 'whodunnit?' type of approach! :-)

Avatar for sneaks

thanks for the advice - I hate these kind of 'this is what I'm doing/done and why' sections. bleugh! For some reason I'm really good at them with other peoples work, but not my own.

Another question:

Can you have an 'over arching research question' that doesn't really specifically get answered and then more specific ones tackled by each study?


I've got an overarching question, and have the
usual intro, lit review (both still ongoing) and methodology chapters, then I
have two chapters that provide basic info that I think is needed to understand
where I'm coming from. Aftert that I'm returning to the main question which I've
divided into three chapters each with its own question and aim which I will work
through relating the contents to its own question. the final chapter will bring
all these questions together and tie them to the overarching question. At which
point I should be saying 'TA DA' but will probably be thinking that there is a
grave error somewhere along the line which will invalidate everything I've

Avatar for sneaks

I ditched the overarching one in the end - it was annoying me, because I wanted to say how each research question was addressed and I couldn't really say it specifically, just 'well, so errrm, the thesis kind of addresses the overarching one'.

So ditched it and had a sentence about the overall aim of the thesis instead.

I am sure this will all be ripped apart by sup anyway


I've pulled mine apart many times and turned it upside down too! I think it is part of the process

I now - well in September - will have a completely new supervisory team AGAIN so I am expecting they will want more changes anyway. I managed to persuade them not to give me the one who when I first started this process decided I should pursue something they wanted which was miles away from where I wanted to be, but now I have to explain it yet again. The only plus to this is by the time I get the thing finished I should be able to rattle off a very good summary! :-)