Data analysis and presentation for a mixed method study

[Tabassum Ferdous]

Hi all

I am doing a concurrent mixed method study in ‘Health education for chronic disease management’ where quali is the principal and qanti is supportive and I have situated the study in the interpretive-constructivist paradigm. In a published article I have justified the use of MMR in int-cons paradigm. Now at the end of the study I have a few concerns and I need your expert suggestions in regards to these 2 queries.

 

1. Since the study has been conducted in a small regional city, and my participants were culturally and linguistically diverse; I have been able to collect both quali and quanti data from 27 and only quali data from 17. If I include all 27 for the quanti analysis then can I say that this quali sample is embedded/nested in the original sample of 27?

 

2. My second concern is about answering research questions. My two research questions are subdivided in to seven sub-questions, some of them are answered by only quali or quanti and some by both. Can I present my findings by answering sub-questions followed by emerging themes?

 

Looking forward to your suggestions.

[Pat Bazeley]

Tabassum,

I wouldn’t bother trying to label with embedded or nested or whatever – just simply state that you had 27 with full data and a further 17 with qual data only  (or whatever the numbers were – it’s not clear from what you wrote, because in trying to apply the label you contradicted what you had already written). Clear statements of what was done in methods are always more important than labels.

Re the second question: surely the issue here is what you want to be the ‘take-home’ message for your article. You then organise your material in whatever way is appropriate to make the points you want to make. That will rarely be achieved by listing questions with nothing more than retrieved ‘themes’ (especially if they are just descriptive categories listing what people said in response to a question).

Pat

[Tabassum Ferdous]

Hi Pat

I am extremely grateful for your response. Looks like now I have got a clear direction for writing my thesis. I have become more excited to see that the response was provided by someone I have already referred in my writings.

Thanks again

Tabassum

[Muhammad Mazhar Khan]

Hi Tabassum:

I am just thinking about your first post/inquiry regarding your R design and other issues. You know I have very strong feeling that you seem more than concerned with justification of tools, methods, and the overall design. Today, to my shallow knowledge, it is important to see a R context from the lens of your RQs rather than paradigm. So we’re in an era of “free-floating” research methods, where paradigm war has subsided to a great extent.

Hence, it’s nice that you have done so much research. So I’d suggest that you can present your R from the best angle you see it to be presented. I am sure it does not matter “as long as your R makes sense”. I am sure it does. Regards, Mazhar

[Tabassum Ferdous]

Hi Mazhar, Thanks you so very much for your assurance. Hope to discuss more about MMR in future. Regards, Tabassum

[Muhammad Mazhar Khan]

You’re welcome. Forgive me but would you mind letting me know what makes you so conscious about justification of every method “within a particular paradigm”? Have you been trained this way in your coursework? Just wondering.

[Tabassum Ferdous]

Think it is my supervisor’s influence, the way she has been guiding me.

[Vanessa Arcilla]

Hi Tabassum,

I agree with Pat regarding the reporting of your participants. In your second question, you may somehow present your findings in two ways, either by answering your subquestions (regardless of being quan or qual) or by focusing more on your two RQs, wherein you reconcile your quan and qual findings. Or you have trouble reconciling them, you may want to have an integration area in your findings where you may discuss complementarity or contradiction of your quan and qual findings.

But since your model is QUAL-quan, you may have to put more weight on your qual results and discuss how these are supported, basically, by your quan findings.

 

[Author]

Posts