I'm currently conducting a qualitative research using semi-structured questionnaire interviews with a sample size of 72 participants. What is the best tool to be used to extract codes and themes from the huge amount of data that I have?
I suggest the following book: Qualitative Data Analysis: An Expanded Sourcebook by Matthew B. Miles, A. Michael Huberman. It can give you many ways to analyse your data.
I do recommend the old version (1994), rather than the last version (2013).
Dear Ibrahim, I can relate with the voluminous amount of data that you are working through. I am currently working with unstructured data from more than 80 participants and fully appreciate the amount of work that goes into it. Your question pertaining to a 'tool' to extract codes and themes: I started using Atlas.ti in February last year. It took some time (2 to 3 days) to get a functional understanding of how it works. After watching the videos on the Atlas.ti Youtube channel and reviewing the resources on the website (http://atlasti.com/), I was absolutely convinced. The aforementioned resources takes you from start to finish and the rest is up to you. There are also online training and talks that you can engage with. I am sure there are other views out there (i.e. NVivo), but my personal experience was pleasant and I am continuing in my journey to exploit the potential offered by Atlas.ti.
How to Code Qualitative Data?. Available from: https://www.researchgate.net/post/How_to_Code_Qualitative_Data/1 [accessed May 10, 2017].
I am not sure what you mean by "extract" codes, since the usual approach is for qualitative researchers to "attach" codes to data segments. This means that you need to develop a codebook which contains the set of codes you would be using and standards for when you would apply each code, One widely used method for this kind of coding is Thematic Analysis (Braun & Clarke, 2006).
As far as software tools go, all of the major packages provide similar features, as well as extensive training videos etc., including ATLAS.ti, Dedoose, MAXQDA, and NVivo.
I presently use QDA Miner for this task ( https://provalisresearch.com/products/qualitative-data-analysis-software/ ). WordStat and SimStat are tools included in the package. The library of my university offers it to students and professors, maybe yours too ?
I think Strauss, A. Qualitative Analysis for social Scientists would be a great help to learn basics about codes, memos, theoretical sampling, comparisons and to develop theory through qualitative analysis.
How to analyse the Data depends on your Research-Interest. Often used is Content Analyse. For an Overview: http://journals.sagepub.com/doi/abs/10.1177/1049732305276687?journalCode=qhra
There are different kinds of the CA, you have to figure out, what´s the most suitable for your Interests.You want to know more about a special feeling or emotion? Than use the problem-centered or one of the in-depth-methodes.
Because of your large Number of 72 Participants - I would consider you want to analyse Expert-Interviews? There ist a special Method when analysing them, because you only need the special knowledge of them.
Along with the other good suggestions above, another source to consider is H. Russell Bernard, "Research Methods in Anthropology." A couple of years ago, I downloaded the 4th Edition (published 2006) as a free Google eBook; it is such a good resource, however, that I will be buying the hard-copy when I return to the US this summer. Pages 505-521 specifically address the use of 'Content Analysis' techniques for the type of coding you are wanting to do.
It largely depends on your research objectives. If your study is theory-driven, i suggest you employ both inductive and deductive coding techniques - inductive coding first before the deductive coding. With the large volume of data that you have , i suggest you do thematic analysis. But, remember that your analytical strategy would have to be informed by your study design. I strongly suggest you read: Morse, J.M. (2015) Analytical Strategies and Sample Size.
As for a suitable software, i suggest you get NVivo v10 or v11.
Aquad is a free qualitative research tool and works with all kinds of data without transcribing.
The actual coding still comes from you, software just makes it easier to analyse. I recommend Patton (2015) as a summary on what software can and cannot do and why it is a good idea to use it.
I did this for my PhD (published in 2002) and did it manually - printed transcripts and coloured highlighters. I had 97 interviews - all of an hour or more... It takes time and can be tedious, but it brings you closer to the data. I looked at qual analysis software - Nvivo and nud.ist then - but felt it was pushing me is directions that I didn't want to go... Of course the software is better these days...
What method you use will depend on your research questions and overall methodological approach. The best, most practical "how to" book I have come across is "The Coding Manual for Qualitative Researchers" by Johnny Saldana. It details a whole range of coding strategies. I have recommended this book to every qualitative researcher I know!
I did it manually -yes, it is painstaking and time consuming process but you are in control of the process. I used many highlighters to "coloured coding" my themes, sub-themes and categories. I enjoyed the process because I learned a skill of coding and analysing the data and not a machine. This was the milestone of my study a "lifelong learning" skill. I done is manually- word by word and a skill to do it.
Exploratory and/or confirmatory work can also be done using a package such as Leximancer which applies natural language processing methods to your text data. The software generates 'concept's and 'themes' only from the text you supply (and very quickly too). Then you can see if hand-coding produces similar or different results and explore why...several colleagues have used both methods and found the comparison useful in their theses and research work.