All entries for November 2017
November 27, 2017
Initial Stage of Grounded Theory Coding
Coding the data using the initial stage of the Grounded Theory process, known as Open Coding or Initial Coding, has progressed substantially since the previous update. In fact, I’ve actually completed the task of coding through the first set of data during the previous week, which I had not expected but has put me ahead of schedule!
Just as a brief reminder, Open Coding or Initial Coding refers to identifying concepts within the data and the use of codes that summaries or describes the meaning or characteristics of that particular data segment, and therefore identifies these concepts. You could say that coding gives data segments an identity that you can refer to time and time again as you progress through your coding, depending on the characteristics that you identify and interpret within each data segment. You are essentially making practical, empirical observations of the data, and interpreting that data to mean something that is of value or in some way contributes towards characterising the phenomenon of interest that you are exploring. Whilst all your codes and code-data segment matching is an interpretive process, it is also objective as all codes are grounded in the data, particularly with the process of comparisons between data segments for similarities of characteristics. I shall be talking more about this in my future short blog series of Grounded Theory from next week.
At the time of writing this blog post, I think I have about twenty or more different codes that I have used across the whole data set, and this is actually a reduction on the amount produced during previous coding sessions. What I am increasingly discovering within the grounded theory approach is the influential impact and role of context on my interpretations, and perhaps the way that I should be interpreting and coding the data, and identifying the appropriateness of code-data segment matching. What is assigned a particular type of code in one context would be coded as something completely different in another context. This appears to be the nature of exploring learning processes and phenomena using grounded theory: the understanding and acquirement of knowledge regarding the development and process of learning differs between contexts. With collaborative learning for example, the collaborative activities, processes and communication shifts and moulds what is happening within the data as time progresses, and can illuminate different patterns at different times depending on the context; depending on what is being dealt with at the time. It is simply not a case of observing a particular process and thinking that it’s always universally understood because learning processes and phenomena have a nuanced existence that is shaped and moulded by events, happenings, actions and others within collaborative situations.
Therefore, as a grounded theory researcher, when you are exploring learning phenomena, the context that envelopes or provides the basis for the learning process is able to mould and shape this learning process over time, yet grounded theory enables you to identify the nuanced existence and subtle differences between the characteristics of similar concepts. Beyond reading the textbooks on grounded theory, the biggest learning curve and learning experience of my application of grounded theory has been trying to understand the importance of context and the way in which this really impacts my interpretations and observations of the data. I’m still learning now. I’m still wondering and questioning if I have really coded everything correctly even though I have checked through things several times during the past week and have altered the coding where I feel necessary.
Along with coding, I’ve also been writing plenty of memos. Memos is a technique of grounded theory that helps you to build your theory by capturing all of your thoughts about the development of your codes, what you have observed, the similarities and differences that you find between coded segments, and the comparisons between different coded segments e.g., their characteristics and contrasts between similar and different concepts and what makes those data segments really what they are.
Additionally, all this information contributes valuable insights and input into your theoretical sensitivity and theoretical awareness of the data, as well as developing theoretical sampling. Theoretical sampling is a qualitative sampling method that determines what to sample next (e.g., what information or data you need next) based on the emerging theory: the observations and questions derived from the data and the codes all guiding and directing the next set of data to pick up and analyse. I shall be talking about this more either during the short blog series of grounded theory or at some point early next year.
Focussing on rewriting the memos shall be the focus of the rest of the week in an attempt to communicate my ideas more clearly, to tidy them up a bit, and to reduce their number and organise them into something that makes a bit more sense. The set of memo writing sessions just completed involved writing a memo page (in some cases several pages) per code, within which each data segment coded with that respective code was explained and compared to previous segments in order to identify and locate subtle differences between each segment, leading in some cases to identification of potential categories (which are basically a combination of various codes and provides the core of the theory) and categorical properties and dimensions.
Writing a memo per code worked fine for a while, and the potential categories identified so far are suitable although these obviously need to be re-examined continuously (shall talk more about categories next month) but what I have realised is I have been taking these data segments out of their context and trying to explain them as standalone entities. As I went deeper into the data I began to realise that data segments can be logically connected, therefore trying to explain them independently of each other was becoming an increasingly difficult task. I found myself referring to these logically connected data segments in order to provide a contextual explanation for the data segments and their difference between other similarly coded data segments.
What I shall do next is rewrite the memos and add more details about the context. Instead of writing about each data segment as stand alone entities, I shall now write about complete units of logically connected data segments. This way, I can break the unit down into constituent segments and attempt to explain them individually and then discuss their relationship to each other as part of that unit. Doing it this way, I think I can then explain the meaning of individual segments without losing its contextual meaning and relationship with other segments. And, I can compare data segment to data segment, and data unit (a series of logically connected data segments) to data unit. It makes sense, um, well, currently in theory……..
What’s The Aim Then?
At the conclusion of the week I aim to have a complete coded first set of data (shall be rechecking again), a full set of rewritten memos and an updated theoretical framework. This will then, as far as I am currently aware of, bring grounded theory work to a conclusion for the year. I shall send everything off to the supervisor for feedback and guidance, and up to the Christmas holiday I shall work on the first literature review chapter, and write the blog series on Grounded Theory!
Plenty to come; watch this space (or just read the blog!)
November 17, 2017
During the past week I have focussed on applying Grounded Theory to my data. The current task is to recode the data that had previously been coded in order to find or discover anything new that had not been previously observed. It is quite fascinating when you have reread the collected data several times, because with each reread you do observe interactions, events, happenings, and actions that were not previously perceived or observed. You begin to construct hypotheses and explanations that you had not previously constructed, thought about, or even were anywhere near being consciously aware of their importance and relevance to your research. This is the beauty of Grounded Theory! It’s not simply the case of trawling through all of your data and note every observation on your first reading, and that’s it. It takes several readings to really get to know the data you are going to be coding, and beyond that it takes several readings to observe everything that is going to be observed.
But even if we have reread all the data several times, is it really possible to observe every single event, happening, interaction, object, action and so on in relation to our aims and objectives? There are means and ways in which we can be sure that what we are observing or perceiving is as close to the data (or reality) as is possible through abductive reasoning and Hypotheses testing. But this does not enable us to become consciously aware of every event, action, interaction, happening, and objects that could possibly be observed in the data. I wonder if this is actually possible? Can this possibility or impossibility even be known? In what way can something be known or something be missed if we do not become consciously aware of or theoretically sensitised to its existence?
Earlier, I caught myself in a mode of thinking that I am coming to know is very much incompatible with Grounded Theory, and is something I need to slap out of myself. During lunchtime I was planning out the afternoon work when a sudden realisation came over me: you cannot plan Grounded Theory work. Yes! You read that right. You cannot plan Grounded Theory work. Ok, I said to myself that I intend on recoding ten separate sets of data but I was basing the quality of what I do on the quantity of what I was going to achieve. This is impossible because with grounded theory, what matters is not the quantity of data that you code within a particular session, but the detail, depth and breadth of your observations of what is happening in the data. This detail, depth and breadth of observations not only come from what you observe and code in the data, but also of the theoretical memos that you write. These memos capture your thoughts and ideas about what might be going on in the data as well as enabling you to compare between data sets, to compare data segments and codes, and to hypothesise and imagine beyond what you are observing in the data. Perhaps what you are observing beyond the data relates to what you had previously read about in existing published literature.
I am finding that coding the data is not taking too long but I do need to be careful not to rush anything, and to be careful that the codes that I construct closely relates to the reality of what is occurring in the data, whilst at the same time accepting that I might not be fully reflect reality because of my philosophical beliefs influencing the way I use Grounded Theory and indeed engage with the data.
What is taking the time is writing the memos. Heck, earlier I wrote a memo on what I was observing in the data within particular data points and it came to over three thousand words! Other memos have come to a few hundred words each. The most unpredictable aspect of grounded theory I find is when it’s appropriate to write a memo, because inspiration can occur at any point in your reading of the data.
The interesting point here is that you are not being guided fully by your prior knowledge and theoretical understanding of what you are observing, but you are being guided by the data itself shaped by your philosophical beliefs. The data itself is guiding when I write a memo, what the content might be, and the purpose of the memo. I cannot predict when or where I shall write a memo and therefore, this is the main reason why it’s difficult to quantify your grounded theory work plan of any single session you do grounded theory work. You have to simply let go of control and let the data and your philosophical beliefs shape and guide what you do, when, where, why, and in what way.
In general, as I recode the data I am observing events, objects, happenings and occurrences of phenomena that I had not previously observed, understood, perceived, or was aware of. I think because of my readings of Philosophy and the increased awareness of my own philosophical beliefs, as well as all the other readings I have carried out for the literature review so far, has helped me to become more theoretically sensitised and arguably more aware of what is going on in the data. This is not to say that I have a complete and full understanding because this understanding is forever in development, and I have to argue if I can really reach the ultimate reality of what is really going on in the data. Hence, the data and the use of the Grounded Theory methodology are shaped by my philosophical beliefs. Nevertheless, this might explain why I have been able to observe what I have not previously observed. I can view things beyond the data that I had not been able to view before, but I have to be careful here that I still ground abstract thoughts and concepts in the data itself.
It’s an exciting journey! Lots of ideas and observations going around, which I never thought were possible just a few weeks ago. This is the beauty of grounded theory and of unchaining yourself from dogmatic, restricted approaches to thinking and research. With grounded theory you have to think as broadly, as detailed, as comprehensively and as complete as you possibly can, whilst keeping everything grounded in the data itself. Hence the name, “grounded theory!”
‘till next time!
November 05, 2017
Literature Review Update
The first literature review chapter drafting has officially begun, and during the week I have been able to write more structural and content details (the vision) of the chapter whilst exploring the relevant literature. The vision of the chapter is taking shape: it shall provide the social, pedagogical, and technological context of the phenomena of interest. The chapter shall detail the notion of a contemporary society, the relationship between Education and society, and the way in which technological advancements have assisted Educational aims and directions connected to the notion of contemporary societies. Debates about contemporary society, Education and technology provide the basis for explaining why there is a need to focus research upon the particular learning phenomena, and explain the importance of the focus.
Beyond detailing more of the chapter content and structure, and rewriting the first paragraph about a million times or so before I became somewhat happy with it (took a while to find a starting point: still not fully happy with it, but shall look at the starting point again soon), I found that Grounded Theory reading needed extra focus. Deciding this was the result of being generally happy with current progress of the literature review, and also because I felt at the time I had to decide what grounded theory procedure to adopt. Although I have been open to the possibility of having to combine different methods of grounded theory from different authors to suit the complexity of the research, deciding on the way in which grounded theory should be or can be applied, and which procedures to adopt (and possibly create), to my research is not quite straightforward.
Brief Update on current Grounded Theory thoughts
During the week I have read through selective chapters of “The Discovery of Grounded Theory” by Glaser and Strauss, “Basics of Qualitative Research: Techniques and Procedures for Developing Grounded Theory” by Strauss and Corbin, and “Constructing Grounded Theory” by Charmaz. Glaser, Strauss and Corbin are who I consider to be among the key Grounded Theory authors alongside Clarke and Bryant. At the time of writing this post however I have yet to reread the work of Clarke and Bryant and have yet to read through the other publications of Glaser, Strauss and Corbin (these are current tasks however) therefore , this section of the blog post refers only to the aforementioned texts. I therefore acknowledge and accept that views presented here are tentative and changeable. But, it suffices for the purposes of tracking the continuous development of my understanding and thinking about Grounded Theory.
The key authors do share common ground, upon which stands the most common and typical features of a grounded theory implementation: the importance of a coding procedure, of theoretical sampling, of theoretical saturation, of writing theoretical memos, of using a constant comparison method, and of developing theory. The key areas where the authors disagree are the nature and application of the coding procedure, and Philosophical perspectives.
From my readings of the texts, all authors appear to support different philosophical traditions. Glaser and Strauss (and perhaps which can also be found in Glaser’s later writings) lean towards pragmatism and realism; Strauss and Corbin appear to lean towards a post-positivism perspective, and Charmaz appears to lean towards a Constructivist approach with hints of post-modernism. Because of the complexity of my philosophical beliefs (a middle-range realist ontology accompanied by an interpretivist / relativist epistemology), there are philosophical and methodological claims and approaches of all key authors that I agree with, and some I do not.
Arguments for and against claims and approaches are driven by the nature of the research problem, the source of the data, and the nature of the relationship between myself and the research participants. Strauss, Corbin and Charmaz base their writings on interview data located within Ethnographic designs, and advocate a strong connection between researcher and research participants. This is, I find, more so with Charmaz as there is a section in her book pertaining to the use of Grounded Theory within Ethnographic research.
The design of my Grounded Theory research therefore differs to what Strauss and Corbin and Charmaz appear to be discussing, and appears to be closer to Glaser’s more realist grounded theory writings. But this is not actually that easy to suggest as it appears (at least tentatively) that ontologically I align with Glaser whilst epistemologically I align with Strauss, Corbin and Charmaz. This is why at the practical level, the application of grounded theory procedures is not straightforward.
Thankfully, at the practical level of procedures, the key authors do agree in general on the purpose, function, process and application of the beginning coding procedure (termed either “open coding” or “initial coding”), the general idea of merging codes and classifying similar codes into categories, and the development or “filling out” of these categories within the realm of theoretical saturation. What happens next with the filled out categories in order for these categories to be investigated and linked theoretically to form a theory is a matter of substantial debate and disagreement among all key authors (and, indeed, many critics and commentators of Grounded Theory).
Whilst I need to give some consideration to the later coding stages, I’m more concerned at present about the beginning coding stage. I have coded some of the data before, but for various reasons beyond the scope of this blog post I shall be reanalysing the data as I am viewing the data from a slightly different perspective than before. I am not utterly convinced that my codes that I have initially created make sense, so shall be combing my way through them soon before continuing with more coding of the data.
What I am assuming shall happen is that I shall be able to select (or create, as necessary) and apply the correct coding procedures as defined by either Glaser, Strauss, Corbin and Charmaz (and others) depending on the way that I perceive the nature, structure and the processes involved with constructing the data, and the way that I think the coding is progressing. The only way this is possible, in my view, is to code the data, examine and integrate codes to form categories, and really develop and explore these categories. I think it is only then will I be able to decide the direction of the next stage beyond initial coding and category development.
The problem I have is that the source of data for coding is different in nature to what is typically and commonly used by grounded theorists. Therefore, this is providing unique challenges. But these challenges, whilst formidable, will be overcome!
Or so I keep telling myself………
‘till next time!