Sneaks' mixed methods mess continues (walminski!! help!)

Avatar for sneaks

Ok so I've decided I'm pragmatist. Just about to write that section up and I've had a look through Creswell's book.

Its banging on about different types of enquiry....sequential, transformative, priority etc etc

But I don't think my study fits any of them. I have...

1 qualtiative study (analysed for themes)

1 qual and quant study (actually this is really the same data from study 1 analysed with a qual/quant technique)

1 quant study. (questionnaire with stats and stuff)

My problem is I don't really integrate the findings anywhere - I just use qual and quant as they were the best way of getting at the problem - which I understand is 'pragmatic'

Just worried I can't say "I used a sequential, qual priority....." blah!


Sneaks, it is not uncommon to use triangulation (normally the term for mixed methods: forget the tri bit, it could still refer to two!). I think that qualitative and quantitative analysis should are not opposites but should compliment/complement each other. Let that be your rationale whichever way you go about it. (up)

Avatar for sneaks

I think I'm just struggling, because although my 1st 2 studies are connected - i.e. using the same data, just for different results, my last study is not really connected at all. Its on the same kind of participants, but its really looking at a different question. Its stand alone. and that's what is confusing me.

I can't say my questionnaire was developed from the previous interviews or anything :-(


I think your rationale should justify the use of the questionnaire on the 3rd study.

Avatar for sneaks

======= Date Modified 24 Feb 2011 12:00:59 =======
ok, so the questionnaire in Study 3 builds on some of the issues rased in study 1, but its not like I developed a questionnaire from the themes

I basically have used a pre-existing questionnaire and used it in a unique setting and I use it to investigate 2 issues that arose in my interviews. I suppose that's kind of linked?? but really I'm writing them up (well haven't yet but sup wants me to do this) reasonably separately - so they are pretty much stand alone.

Maybe I should ditch the stand alone aspect and talk about how they tie together?


one of the things that keeps coming up when I talk about thesis with people that already got a PhD is that a thesis should tell a story. They told me that you need to show why the work in chapter 5 followed chapter 4, etc, so that the examiner can see why these things developed. They don't have to be chronologically ordered (mine are all over the place!) but they need to make sense as a package.

I can see why your sup have said to write them as seperate things for now, since you have mostly been writing up your work for publication. I have been doing the same. However, for the thesis, I've had to connect all the different studies back together, so that it tells a story.

So on that basis, I would say you'd be better off stressing more about how the studies link :)

Avatar for sneaks

you see, this is what I've said to sup - but she's continuing with the whole "they say different things and you don't need to link them" whereas I like to see threads being pulled through the thesis

Its beginning to dawn on me that my supervisor may well be a complete moron. (turkey)


lol I think that is normal. In a PhD, a process in which we learn to become professional researchers, I think the student need to become the teacher at some stage. In fact, I think that is the evidence of one becoming a fully fledged researcher - one that can think for themselves :)

It's your thesis at the end of the day, you should do what you feel is right. Have you seen any thesis in your fields that have talked about the different studies as being completely seperate?



I really don't mean this to sound as it might... the perils of forum posting... but the thought that struck me is that if the terms in the Creswell book are causing you grief,  use a different book for reference.

There's zillions out there, instead of crow-barring your work into an ill suited framework. 

It sounds like you've developed a wonderful story and the elusive golden thread is there. You did a then b and because of what emerged from a you did c. a and b are connected anyway and c emerged... what more can you ask for.  Brilliant.

Sorry if that sounds sharp, not intended.


Avatar for sneaks

======= Date Modified 24 Feb 2011 13:01:18 =======
not sharp at all :-) I welcome the comments because I'm getting F all help through the 'formal' channels.

I'm using creswell, because I can get my brain around it - but have no fear I've also got up about 20 windows all with different articles and books on mixed methods, its just taking me a while to get through it all and i only have about 3 hours to write this section argh! May have to take tomorrow to do it too!

Have now decided to include a little picture of my studies and how the methods link (up) I like pictures!

ETA: @pink numbers - now that you mention it no, even the previous supervisee has links. gah! I'm going to ignore ALL sup's advice from now on!


I think this is probably one of the most difficult bits really, YOU know how and why you have done it, in a particular way, you just have to put it onto words. At the beginning of my methodology I have put a chart (I like charts and tables!) which shows how all the bits link together and how the various approaches I have made are reflected in the contents of the chapters. I thought this was necessary as it doesn't follow the method -result -conclusion path but is much more like a 3D jigsaw, or a whodunit? This makes it difficult to say 'I used xyz approach' because surely research has to be fluid, mine has taken lots of different turns as areas that appeared to be peripheral have turned out to be much more important and are now taking centre stage. you could say that as a result of the findings of a and b, c appeared to be the next step because....., or a particular answer/analysis showed that one area appeared to require further investigation, and c was the way to obtain the info needed. You could then use your pics to illustrate this. :-).


Hi Sneaks, don't worry at all about you research not fitting Creswell's and Plano-Clark's typology. You can develop your own mixed methods design or adapt one - there's that level of flexibility with mixed methods. Remember that Tashakkori and Teddlie have actually classified 40 different designs, so you can have a look at theirs. We can classify yours if we know a bit more about it, using priority, order and timing. Did each phase build on the other one? Did you use the qual and qual/quan phase to build the questionnaire in phase 3?


Hi Just seen this thread after I've worked out that I'm a pragmatist. I've seen some mentions of Cresswell. Can anyone suggest anyone /books to read on pragmatism / mixed methods? (I'm a newbie PhD starting in September).


@LBaines - you could try the main one, which is Mixed Methods in Social and Behavioural Sciences. It's still early and I'm waiting for the caffeine to kick in and one by Creswell and Plano-Clark..also one by Jennifer Greene. Type mixed methods into Amazon and they should pop up.

@Sneaks...Looking at what you've said above again. I think you have a sequential exploratory mixed methods design of some type, fully mixed. You do have some integration, which is a tricky business for mixed methods (Bryman, 2006, 2007). For instance, your qualitative themes from the first phase have actually been used to generate quantitative items in your questionnaire. Integration can occur in your style of writing and the way you weave your qual and quan findings together, also in the discussion. There's many ways that integration can occur, and the difficulty of actually achieving integration can depend on your mixed methods design.

Avatar for sneaks

======= Date Modified 25 Feb 2011 14:09:38 =======
thanks wal. For clarification...

Study 1: interviews - found 6 themes - purely qualitative

Study 2: same interviews, but looked at *how* they were describing stuff, rather than what they were describing - coded stuff and compared it quantitatively (with stats)

Study 3: picked up 2 of the things out of study 1, then used this finding, along with an existing hypothesis in the literature - I didn't use the themes to create the questionnaire - all the questionnaires I used were already out there in the literature. I just used them to test a concept that emerged in study 1 (if you get what I'm on about you deserve an award!)