Analysis, the difference between qual and quant

Earlier this month, NewMR held its first Explode-A-Myth session and my contribution was a discussion why there is no method that is a melange of qual and quant, because the underlying paradigms are different.

Through the Q&A session at that event, and in particular a question from Betsy Leichliter, I gained a clearer understanding of the core difference between qual and quant. Betsy asked “So should the ‘qual’ or ‘quant’ labels be driven by the method of analysis, not necessarily the method of “data collection”?”. I think this question from Betsy is the best answer to the question about what is the difference between qual and quant I have seen.

Within reason, any data can be assessed quantitatively or qualitatively. Of course, there are some limits to both approaches. A very small amount of data is likely to produce findings that are hard to generalise. We can count the sales of brand X, in one store, on one day, but it is hard to draw any inferences about the world from that. Similarly, ten-thousand open-ended responses could only be assessed qualitatively with a large team, or a large amount of time.

The quantitative approach is based on an assumption that there is a ‘real’ world, which we can measure objectively (or, at least, that we can get fairly close to that ideal). The underlying beliefs are a) it is the method that provides the results (different researchers should provide the same answer if they use the same method on the same data), and b) that the researcher is discovering and reporting something that exists.

The qualitative approach, as it has developed over the past thirty years, is based (for most researchers) on a constructionist paradigm (there are several different models, but they all tend to be constructionist). The researcher does not discover truths, the researcher creates a narrative that provides useful insight into what is happening. The researcher is part of the analysis, different researchers will provide different narratives, and the value of the narrative depends on the ability of the researcher to observe what is happening, to synthesise an analysis, and to create a narrative that conveys something useful to the end client.

The key difference between quant and qual is the difference between discovering and creating, overlaid with ritual of using numbers for quant and words for qual.