Hi Anki - Aino Petterson is also exploring Nvivo. She has a different set of issues (she is using text data as source, from Twitter). Could be good to hook up though. Cheers!

Thank you, Robbie - will do. Hopefully it won't be a case of the blind leading the blind!

For thematic coding, you do need to transcribe (or have someone else transcribe) each interview. And you should set up NVivo so that you analyze each participant separately. With regard to context, you should first find patterns in your coding via NVivo, and then go over the relevant data segments to interpret those patterns.

Have you read Braun and Clarke (2006) on thematic analysis? Their 6-step process should be a useful guide for you.

Hi Anki,

1. You can use otranscribe to fasten your transcribing; it did help me a lot. Transcribe yourself as far as possible as it gives you the opportunity to create and sift through your codes lending you a very deep insight into your narratives.

2. I believe setting NVivo to per participant would be preferable if each narrative is equivalent to unit of analysis.

Good luck.

Thank you, David & Devasheesh!

I am definitely doing the transcribing; I was more wondering how faithful I needed to be to all the 'uhm' and 'err's since content is only what matters. I have found mentions on researchgate that you can skip these, but I haven't found an actual source in the literature confirming this. Hence my hesitance on it.

As for coding I am currently trying a hierarchical structure - e.g. I sort all in-vivo & descriptive codes so they either relate to situations that elicit this emotion or (separately) to behavioural outcomes. I am concerned though that the behaviour extracts may not make sense out of context - but perhaps this is not a problem if you, as the researcher, remembers the context of each interview?

I will have a look at otranscribe - thanks, Devasheesh! I have indeed found Braun & Clarke, David (it's an excellent and very helpful article).

The question of transcribing "um" and "er" etc. depends on how useful that information is for your analysis. If it doesn't matter for understanding your substantive interests, then standard verbatim transcription should be just fine. I follow that approach, with notations for either laugher or long (2 seconds or more) pauses.

Using a hierarchical structure for your coding is often a wise choice, since it groups a larger number of initial codes into a smaller, more useful set of categories. The next step is to work with those categories to think about the data thematically -- usually in terms of a few (perhaps 3-5) broad themes that will tell the reader about what is important in your data.

I think NVIVO will help you to arrange things, to see the links and connections. But real analysis is to you. You yourself have to do it. You have to do coding and subcoding based on your research questions and lit review. NVIVO makes your work more arranged, helps to make framework among the themes. Best wishes for you

Thanks Dr David L Morgan for his comprehensive elaboration on convenient transcription method ; second, I would like to thank him as I really benefited from the documents that he posted i.e. Braun and Clarke. thank again!

Then, If I have some thing to add, from my short experience with qualitative data analysis, is that you need to quote "um" and "er" as separate responses which describing a specific themes.

In other words, to be transcribed as introductory or preliminary narrative comments that describing or orienting quotes so as to highlight or foster specific theme.

Finally, I agree with colleagues that Nvivo is just help you in grouping and categorizing data. in your case better to code per participant.

All the best

Hi Anki- I am doing thematic analysis in NVivo.What in my opinion is never upfront is the preparation of text for coding in NVivo- please pay particular attention to that because there is no straight forward reference. For me it was like trial and error and very painful- once the data is in,walla you can do what Dr Morgan and Others have suggested. Try reading Saldana(1999).

Thank you, everyone & especially Dr Morgan! Can highly recommend oTranscribe after using it - it did improve my speed & accuracy. :-)

Thank you I will try oTranscribe next time- I transcribed all my interviews, a very painstaking and hallowing experience!!!