Ignoring Grounded Description

Barney G. Glaser, PhD, Hon PhD

Why is there so much grounded description? The simplest, direct answer is that to many a researcher this is GT. This view is supported by several factors. It is easy and natural to describe accurately. So slipping into grounded description comes naturally and is ok as GT. Also departmental support for description is strongly supported by perspective and academic rewards and history and routine QDA. Also many researchers and readers of research cannot conceptualize very well if at all. They want accurate description about the data in the study. They are not into taking a core category as a general category applicable to general implications applicable to much data elsewhere. Their study is about explaining processes the data, NOT in studying the implications of core and sub-core categories as they are integrated into an explanatory theory. I trust the reader can think of other sources of letting GT research slip into conceptual description.

Another major source of ignoring detailing no conceptual description when doing GT is the write up of the methodology for doing GT in the many books now written on doing GT and its procedures. The reader is not warned of the possibility of slipping from the prospect of doing good conceptual GT into the grab of doing extensive conceptual grounded descriptions. Conceptual description is assumed as GT. For example in Holton and Walsh’s new excellent book (Classic GT 2015, Sage) they have a chapter entitled “Discovering New Theory as the End Purpose of Classic GT.” They state immediately that “developing is what we are meant to do” doing GT. They then devote ten pages complete with charts and diagrams explaining different types of theory. It is too complex and abstract to follow for designing a theory for a GT. Not once do they warn the reader about the slipping of conceptual into extensive description of a grounded concept. They talk of grounding concept with no illustration of data source, which is the opposite of giving too much data. Mild illustration dosage stops excessive conceptual description in writing the final product. Having a mild illustration dosage design prevents excessive conceptual description take over.

In another chapter on analyzing data (chapter six) they again do not warn of excessive conceptual description. They talk quite correctly that GT depends on the conceptualization of data by coding and memoing. They refer to the several incidents used as interchangeable indicators when using the constant comparative method to generate and discover conceptual codes. But they do not warn of writing about all the interchangeable indicators yielding a concept. This, of course, results in excessive conceptual description to no benefit to generating a conceptual GT. The excessive writing of incidents just describes the grounding of the GT over and over. It slips the theory into description, and loses the conceptual level of a GT, while still calling it a GT. Telling one incident as an illustration of a concept/code is enough. Discovering a latent pattern is exciting and it is hard to not describe it at length and easy to miss not relating it to other concepts to generate a conceptual theory. Holton and Walsh do come close to citing the grounded conceptual description problem when they say “description captures a moment in time. But the essentially limited nature of descriptive writing hinders the theory’s ability to produce a complex yet parsimonious multivariate abstract theory”. Thus hinders theoretical explanations of the latent patterns in the data. Description takes over. The warning is but a step away and very important to highlight. Again the natural need to describe a length and the tremendous prevalence of description in academic research forces disattending the take over of conceptual description to the loss of GT. As Holton and Walsh say, “Concepts remain buried amid detailed empirical accounts and it is difficult if not impossible to see the simple elegance in the relationship between concepts that together integrate a GT and provide its explanatory power.”

 

In their current book (Rediscovering Grounded Theory, 2014) Gibson and Hartman devote 21 pages to an elegant chapter on the philosophy, perspective and use of codes, that is concepts. Yet due to their lack of experience in actually doing GT research, they do not mention slipping from concept into description thus losing the theory involved. They see codes or concepts as necessary for theory, but they do not warn about slipping from the concept level back to empirical description level by excessive illustration of the data that grounded the code. Their chapter 9 on rediscovering coding is 20 excellent pages relating data to codes, yet there is no mention of excessive description of the data used to ground a code for a theory. When extensive description is done it is considered GT without realization that the GT has slipped into lengthy description. When doing GT we all can make this mistake inadvertently still in 2016. Hopefully that paper will correct this slippage.

My book Theoretical Sensitivity (1978) deals extensively with substantive coding pages 55 to 82. I introduce the chapter by saying coding releases the abstract analysis from the empirical bond of the data. Coding allows the researcher to transcend the empirical nature of the data, which is so easy to get lost in, while at the same time trying to account for abstract patterns in the data. The abstract conceptual view of data was there, but lost in actualizing research procedures. Thus I saw the problem in 1978, but I neglected to say careful to not reinstitute the data bond back to the empirical level by excessive illustration of the code using a lot of interchangeable indicators. I do say later in the chapter to keep code illustrations down to a bare minimum. It is clear that I almost saw the problem of conceptual description but missed it and thus did not warn my readers.

For the researcher who finds himself doing conceptual description her/his decision or choice is clear. Does he stop and do GT methodology carefully and not use excessive description Or does he continue to do a conceptual description with a QDA methodology? Either is quite acceptable generally speaking. Of course academic department perspective plays a role in the decision, but logically neither is better than the other. A rhetorical wrestle between the two methodologies to favor one over the other is a waste of time. Both are quite respectable (see Choosing GT, by Barney G. Glaser, 2014). It is a simple choice depending on what the researcher wants and what his department will accept for a PhD thesis. A choice must be made to keep the methodology decided on clear. No choice and continuing conceptual description results in a jumbled report. Open substantive conceptual coding goes anywhere out of control when not in control by a GT procedure. Is the search for more concepts or more data indicating a concept becomes the question for the researcher and his readers.

Facebooktwitterredditpinterestlinkedinmail