Hello
I would truly value guidance on the best approach to combining inductive and deductive analysis of qualitative data. I will first analyse my data using Braum and Clarke's (2006) thematic analysis process. Once complete, what process(s) do you recommend to deductively analyse the data noting it has been inductively analysed.
Thanks :-)
Can you say more about why you want to start with inductive coding and then move to a more deductive version?
In general, you could treat the coding system you derive from the inductive approach as a "codebook" and then recode the data. Or you could divide the data in half and generate the codebook from the first half.
Either way, I am not sure what the deductive phase would add.
Britt Anne O'Keefe
I think this would be good if you have a large amount of data. In the first stage, you may sampling the data ( half of the dataset, as David L Morgan suggested, OR some representative portions ), analyze them inductively to create your codebook. Then, in the second stage, the generated codebook can be used by you or (other coders) to analyze the rest of the data deductively based on the initial coding. Remember to check the Inter-rater reliability ( IRR).
I am not sure, however, assuming that you have already completed thematic analysis (inductive) and trying to find correlations between themes (could be treated as 'variables') if any (deductive).
If you are intended to research two themes (variables), of which, one is independent (with varying degrees in quality) and the other one is the dependent theme (variable), you could perform a correlational analysis based on any existing theoretical/hypothetical framework (e.g. degree of happiness and spending on self or spending on others)
Inductive data usually provides you with some evidence-based, rather than a priori, theorizing. So in this case, you can look at the themes/trends that exist in your data to design surveys, for example, with survey questions driven by those inductive data. That way, you can now apply what you learned qualitatively to an analysis of potential generalizeability. This is one approach, but there can be numerous, depending on your qualitative methods, your results, existing data on the topic, what you are seeking to understand, etc…
Additionally, you can look at potentially applying an abductive-inductive-deductive (A-I-D) qualitative approach with a builtin Peircean self-correcting methodology as per the work of David Boje et al. (4th Wave Grounded Theory), or the novel design of built-in quantitative methodology into the A-I-D Peircean self-correcting work (5th Wave Grounded Theory).
- Badges
- Science topic
- Similar topics
- Mathematics