Initial thoughts on the methodological issues of integrating quantitative and qualitative data

Regular readers will have probably noted the discussions I have made (or starting to make) about the Philosophical difficulties of integrating quantitative and qualitative data in a single research study, relating mostly to the fact that quantitative is usually associated with the Positivist perspective whilst qualitative is usually associated with the Interpretivist perspective. But what I have not really touched upon at all are the difficulties of the methodological perspective (yes: there are Philosophical difficulties AND methodological difficulties, and both appear to be related to each other: check earlier blog entries that discuss relationships between Philosophy and Methodology). The methodological perspective is beginning to gain more attention as I come to understand Grounded Theory, and a couple of questions that have come to me are: what methods are appropriate for data integration? Along with, which methods are suitable for my research?

With the data collection this is no longer a problem: a mixed data questionnaire shall collect both qualitative and quantitative data and an extra method or couple of methods shall be used to gather more qualitative data. Quantitative data shall be analysed using a series of different statistical methods (descriptive statistics and also methods to identify and analyse relationships between different identified variables), whilst qualitative data shall be analysed using a series of analytical methods inherent to Grounded Theory processes (though there are some debates about the usefulness of some methods depending on the context of the research). Essentially, Grounded Theory involves interpretation of the collected data, and to develop codes and categories using the coding methods in order to explain or describe what is actually going on within the data. These codes and categories are developed for each qualitative data set and then compared across each set using a method called “constant comparison.”

Describing the “constant comparison” technique is way beyond the purpose of this blog post, but it suffices to say that it is used as a mode of comparing codes and categories across data sets as part of the process of continuous and simultaneous data collection and analysis, in order to develop a theory or to theorise about what is going on within the data. It’s a bit more complicated than that but for now that’s the best way that it can be described rather briefly. The point I am trying to make here is there has to be a way to generate codes and categories from statistical, quantitative data in a way that is comparable and compatible with codes and categories generated from qualitative data, in order for constant comparison to be utalised across all data sets produced from all data collection methods. If this is possible within my own research, then the theory or theorisation that shall occur as a result of analysing and integrating different data sets shall increase its reliability, validity, and possibly even generalisability. But this is something that I shall need to work out and perhaps it might be related to the quantitative methods: could I create comparable and compatible codes and categories from descriptive statistics? Could comparable and compatible codes and categories be generated from relational descriptive data analysis such as, say, the likes of ANOVA? What about regression analysis? Are codes and categories even meant to be compatible and comparable across differing data sets? If not, then in what way can a theory or theorisation even begin to happen if these codes and categories cannot integrate? What, exactly, is required to develop a theory or theorisation from a complete and cohesive collection of data? In what way can a collection of data be considered complete and cohesive? Does any of that even matter?

There are many many issues and problems, debates and perspectives relating to Philosophy and Methodology of data integration that shall have to be considered, and as you can imagine I shall probably banter on about them on here as and when I come across them! Regardless I do have the belief that I can create comparable and compatibles codes and categories across all data sets. I have the belief that Positivism and Interpretivism in some way can complement each other rather than compete with each other. But I do not know this for sure at this time: this time next year I might have a completely different picture of the way that Grounded Theory works and the way in which integration of quantitative and qualitative data can happen and should be appropriate for the context of my research. But that’s the way research forms and develops!

Alex Darracott : 29 Nov 2015 16:10 |  Tags: Groundedtheory Methodology Qualitative Quantitative |  Comments (0) |  Report a problem