I am conducting a two faceted study, with a quantitative questionnaire package and highly structured open ended questions, as an attempt for triangulation. I am not quite familiar with qualitative analysis software such as NVIVO or MAXQDA. Would excel be an option for organizing and coding highly structured open ended responses? I provide an example data excel file and my data collection tool. Its in Turkish yet you may understand the structure, I guess. Or would you suggest learning some basic features of standard qualitative analysis software packages?
If you are planning to do quantitative coding I see absolutely no reason to use qualitative software.
It may depend on how sure you are about your coding scheme. If you want to develop new codes you might want a possibility to play with the data, write memos, link them to your coding scheme ... for this kind of inductive work I see an argument for some sort of coding software.
But if your coding scheme is rather clear excel is surely not a bad choice.
Excel seems OK, if you want to do basic descriptive statistics. For more powerful analysis, you'd probably want to use a dedicated statistics package, like SPSS.
If you use Excel, one difficulty you might have is estimating the reliability of your data. Ideally, you would want to use Krippendorff's alpha for that, for which you really need SPSS. Using Excel, you'd have to go for some other metric, such as perhaps Cronbach's a, which are OK, but sub-optimal.
Echoing Kenneth, I see no reason to use CAQDAS for this task. Excel is easier to use and more powerful.
Thanks to all for your useful answers. I plan to use Excel for only open ended questions and export those codes as strings and ratings as numberic variables into SPSS. I plan to conduct a validity check by correlating the ratings to the anologous subscale scores. For instance by looking at the correlation of the attitude towards face to face counseling and attitude ratings of open-ended questions... Does this sound good?
In my eyes sounds reasonable. As soon as you have your quantitized ratings you can of course additionally do the full programm of checking dimensionality and scaling (Cronbach's Alpha, factor analytical checks ...). But keep in mind that the idea of validity checking through correlations (and the other mentioned methods) all assume that you are actually measuring the same thing.
In fact I have a previously validated instrument on the quantitative side and I have two open ended questions tapping to the same area. Perhaps this can justify validity of the two items only. The others have different yet closely related aims... Maybe I should still check the inter-rater reliability with Cohen's Kappa.
Probably the most common way handing this kind of data is by creating a series of dummy variables for each of your codes in the qualitative data, where 1 means a topic was mentioned and 0 means it was not. If this is the case in your study, then Excel would work just fine as a way of capturing that kind of data.
Dr Morgan, Thank you for your useful reply. When I enter those data as dummy variables, as you describe, what is the best way to report those data using excel or SPSS in your opinion?
If you already have your basic dataset in SPSS, then I would add the dummy variables from Excel to that. In particular, if you have the same cases sorted in the same order in the two datasets, then you can cut and paste the columns from Excel into SPSS.
I am assuming that these are independent variables, in the sense of what difference would it make it if a respondent did or did not mention a topic in an open-ended question. In that case, the simplest analysis would a t-test, and a more complex analysis would involve entering the dummy variables into a regression.
Thank you very much for these suggestions, This would integrate the data and enable many different analyses I guess I can also make use of descriptives in the form of matrices, frequency tables, etc
A free alternative might be to use RQDA as part of the R package. The main advantage is that it uses a relational database (SQLITE) so that you can restructure your data much more easily for analysis. I've used it for a variety of purposes and it's not too difficult to learn and use.
I've frequently made use of Excel for coding and organizing qualitative data - particularly when the responses are highly structured questions with relatively brief responses.
I have also used Excel to organize and analyze interviews. My interviews have been with open ended questions and long answers. I presume that software designed to handle qualitative data have a lot of advantages but I have managed with only Excel. The greatest disadvantegae using Excel for interviews is that Excel is not originally designed for texts.
Here is a blog-post with some "how to do it" content of using Excel
http://www.qualitative-researcher.com/qualitative-analysis/using-excel-for-qualitative-data-analysis/
Excel will help you note the similar and repeating words, phrases and texts but you may still have to read in between the lines to really grasp the contextual meaning of these repeating words.
