All entries for Friday 06 April 2018

April 06, 2018

Some Current Thoughts on the Qualitative Strand and Open Coding

There has to be a sense of emergence where codes are derived from data and categories are derived from codes. This idea of emergence makes sense as I have not been able to identify an existing framework that is suitable and relevant for the type of data I am using and the way that I am now exploring the phenomenon of interest, and therefore, theoretical constructs, relationships and hypotheses must emerge from the data.

I read an interesting paper the other day where an author aligned the idea of an emergent approach with the realist ontology: truth emerges from the data after a continuous cycle of coding and recoding, but this brings about a couple of problems. First of all, what is defined as truth and of being true? How can it be measured? How can I know that something is true even after going through continuous cycles of coding and recoding? How can I know that this truth emerges from the data and not simply a reflection of my interpretations of the data? Is it absolute truth that emerges from the data or is it that with each coding and recoding I could come closer to the truth without completely attaining it? How do I know either way? Is there some set criteria for truth? If so, then would this criteria itself represent truth if it’s simply been constructed by another human being? Would it therefore be better to consider the set criteria is bringing one closer to the truth rather than mirroring the location of absolute truth?

One thing I do know is when I think about the qualitative strand, the purpose that it brings to the research, and what I currently would like to achieve with the strand, allowing the data to speak for itself; to enable this “voice” to emerge naturally and to code in accordance to what is believed to be occurring within the data makes sense. And this is where it is interesting because some authors suggest that enabling the data to speak for itself (please note that data do not literally speak!) and to therefore let understanding and meaning emerge from the data, but it is clear that there is an interpretation process happening. We as researchers interpret what we are observing in the data and attach to chunks of data what actions and events we believe are occurring within that data segment. Question here therefore is what is the relationship between truth and meaning? Is meaning objective and already exist within the “voice” of the data? Or, do we define meaning and apply it to what we perceive or interpret to be happening within the data? There are techniques within grounded theory such as theoretical sampling and constant comparisons that provide some answers to these questions but to what extent is truth realised by just grounded theory alone? Can ultimate truth really be attained?

What is the purpose of the qualitative strand within a mixed methods approach? From what I have been rereading, mixed methods can be used to build and test a theory, theoretical constructs, relationships and hypotheses. Their development occurs in the qualitative strand and then tested in the quantitative strand, and therefore adding an extra dimension of richness, integrity, authenticity, verifiability and validity to the research design.

A question I am working on at the moment given that Grounded Theory is part of the qualitative strand is to what extent do I use grounded theory? I have now more or less worked out the initial phase of qualitative data analysis, and this initial phase shall consist of Open Coding also known as Initial Coding. I think this is more or less a definite because it is through Open Coding or Initial Coding that meaningful data segments are labelled with suitable codes that describe what is happening; where a technique known as constant comparison is used to identify similarities, differences and variances, and where (in broad terms) these similarities, differences and variances contribute towards categorical development. I have recoded my data a few times and so far I have lots of codes, and some initial categories developing but shall now have to recode the data since developing new ideas about the research design and about the way I want to explore the phenomenon of interest. And also because I understand the data more now. This leads me to an interesting thought: not only do our theoretical understanding of what is occurring in the data develops over time along with the need for particular research design elements (assuming emergent research design), but also understanding of the data itself emerges from the way that we perceive and interpret what is going on. There is an interesting relationship going on here between our own perceptions and interpretations, the development of these perceptions, and the data itself. What role does the data play in this relationship?

I can begin to observe what I had not previously observed and I can understand the grounded theory techniques better than before. I have started to draw out the steps and phases of the new research design with the current focus on the qualitative strand. I understand more now about categorical development and have outlined more questions I want to ask about the data as I proceed with recoding the data and continue to develop categories.

Aligned with my philosophical beliefs, I believe that there is a truth out there behind the process of the phenomenon of investigation but whether or not this real truth can occur only from coding and recoding for the context of my research is doubtful. But a mixed methods design perhaps could lead me closer to that ontological truth without actually reaching absolute truth. Aligned with my epistemological beliefs, the logical process (abductive) that underlies my use of grounded theory (develop hypotheses inductively from the data and use deductive methods to test the hypotheses against the data) aligns with my beliefs that knowledge is not certain and absolute. We need to continuously think about the data, think about what is happening in the data, think about how we interpret the data and how we know what we know to be true or perceive to be truth (meta-Philosophy) as long as everything is grounded in the data. All hypotheses, ideas, observations, and thoughts must be grounded in the data. We need to question our own biases and acknowledge them. All this while we maintain our sanity long enough to do so!

A big question that I have next is: when I have all the codes, and have developed all the categories and identified relationships between each category and the relevant properties and dimensions, what then? Grounded theorists talk about bringing everything together to form a theory whilst other grounded theorists discuss the idea of linking categories together to identify relationships in a process known as Axial Coding. I think I am currently leaning towards axial coding or some sort of coding technique that enables me to relate categories, because it is through the relation of categories and really understanding the way that categories interact with each other could I then begin to understand the way that the particular learning phenomenon of interest progresses from start to conclusion. This is challenging and whilst I shall try to work it all out for the sake of the diagrams I am drawing out as plans, the only way I think I am going to know for sure what I shall do is to simply do the coding. But the way I am viewing this at the moment is whilst the categories in themselves explain what is happening with certain parts of the phenomenon, by themselves they do not explain the process. There needs to be that extra step that identifies the process and the relationships therefore between elements of this process in order to better explain the phenomenon.

Once I have developed the ideas of the way I am going to approach the qualitative strand I shall then deal with the quantitative strand, fit everything within a mixed methods scenario if proven to be the most appropriate strategy, as well as a case study methodology if necessary, and then actually test my ideas against the data and remodify accordingly after receiving feedback from the supervisor.

‘till next time!