Qualitative Analysis with a Spreadsheet

After spending the last five years of your doctoral program interviewing 200 back-woods still operators around the world, you must now analyze the open-ended responses that you have from your interviews.  How do you analyze the qualitative data from interviews? 

First, browse through the responses to try to get a mental idea of the different categories of responses and the different factors that make up the response to each question.  You probably will find, for example, that respondents have mentioned issues associated with the different types of tubing and pipe that is used in the construction of a still – PVC and iron pipe, copper and aluminum tubing, or rubber inner tubes.  "Tubing and Pipe" might be a category of responses, while the type of material might be factors about which different issues are discussed. 

To give yourself a better visual picture of what your data looks like, you can put these factors on a spreadsheet.  List each respondent's answer on the left column, and then create columns across the top using the initial factors that you observed in your first browse through the answers (materials of PVC, iron, copper, aluminum, and rubber). 

Now consider a response that says, "Well, copper has a problem with lead in the system, aluminum corrodes and wants to kink when bent, PVC is cheap and easy to assemble, and inner tubes are useless in the production process but are great fun for tubing in the crick at the end of a long day."  For this respondent, you have a comment to put under each column of material. 

Once you have entered these responses in the cells of a spreadsheet grid, you can now more easily sit back and mentally focus on the qualitative comments down through individual columns.  When you do this, additional factors might now come to mind. The comments of just this one respondent seem to suggest additional factors of contamination, durability, ease of assembly, price, production value, and enjoyment.  If you massage your spreadsheet to break out these additional factors, you can then do a quantitative sort of analysis/report that discusses factors that were not (could not be) considered a priori

An objective discussion of your findings might then say, "Twenty, or ten percent, of the respondents indicated that corrosion and kinking were problems with aluminum, while none mentioned these issues with PVC or iron pipe, copper tubing, or inner tubes.  One hundred sixty, or eighty percent, made mention of the absolute joy of working with rubber inner tubes.  However, the twenty percent who didn't mention rubber inner tubes as being associated with fun made comments to suggest that they do not know what “tubing” is, using this term as a verb rather than as a noun. Our conjecture is that rubber inner tubes do not replace the tubing used in a still, but rather that inner tubes are somehow being used in addition to operations of a back-woods still.  We are not certain as to how these types of tubing are being used in operating a still, but believe that they are being used after a production run has been completed and is ready for sampling and bottling.  Our quantitative follow-up questionnaire will be sure to include questions about this issue." 

Although this method might seem time consuming (it is!), note that by NOT automating analysis at this exploratory stage, you are able to tease out factors that might not otherwise become apparent.  E.g., by using a word processor or other program to count words, you will not pick up on the issues above. 

Also important to note is that you don't necessarily have to know very much about the topic being studied in order to see issues begin to be grouped if you use this method.   Indeed, I would argue that being naive to the topic can help you to make unbiased classifications and to maintain better organization in analysis than someone who has the pre-conceived notions of what s/he should be seeing.  Nonetheless, only an expert could take a final look at your analysis and be able to explain or conjecture why, say, PVC would be reported as the most favorite type of pipe even though it is a material that is specifically not used in many parts of a still.