Grounded Description: No No

Barney G. Glaser, PhD, Hon PhD

Grounded description is on the increase with the increase of grounded theory throughout the world. Much grounded description is jargonized as GT and it is not GT. Grounded description is trying to describe the population studied, like a QDA study requires, by describing all the interchangeable indicators that grounded the concept. In contrast, GT is not to describe the population. GT is the relation between concepts  which emerged from the population by constant comparing and then are related to each other by a theoretical code. The GT theory then becomes general and abstract of the study population by time, place and people. The GT stands on its own, and the data on which it is grounded is often forgotten for the grab and general implications of the theory. For example, rooting from here to there is based on BMRs (eg, planning basic mobility resources for a trip).

Why is this happening? There are several reasons. Description is  frequently demanded by an academic department perspective and professor demands since research is traditionally descriptive nearly 100% of the time. Being abstract of time, place and people is not the normal quest of academic research. Accurate description is the quest.  The generated concepts used for normal description can be very discreet and its indicators also. They prove relevance and fit for population data. The abstract nature and its general implications of a GT core concept and its potential for generating a GT are ignored or lightly referred to while describing takes over.

Grounded description is a step toward discovery of a GT, not a GT. But many new to the GT methodology do not realize this. They become thrilled with generating a grabby concept and think that it is the final step to generating a GT and then describe it at length to prove its accurate existence. This approach, of course, suites the descriptive perspective of most academic research.

Grounded description is further supported by the simple fact that most people, including academics, cannot conceptualize. It is best, if at all, if they can come up with one generated concept. Or if not they can use a conjectured—usually ungrounded—concept, taken from their field literature. That is fine, since description proven accurate, runs the world, with conjectured concepts based on no facts a close second. A distant third is conceptualizing concepts, however powerful they may actually be.

There are many books written now on GT that are actually about grounded description. They confuse the definition and methodology of GT with lofty academic scholarship backed by high-level positions in a university department. They bring GT back to description. Mixed writing on QDA and GT often tend to regress GT to description procedures, such as taping preformed interviews.

In this paper I will discuss many of the differences between GT and conceptual description, so the reader can spot the latter being termed the former. Conceptual description easily results in data overload in quest of full coverage.  For a GT concept why keep collecting interchangeable indicators as required by full accurate description once you a have a concept? Five or six indicators can be enough, as apposed to dissertation requirements to get data from all respondents on a concept for descriptive generality that will soon become stale dated anyway. Full coverage of data collection is a waste of research resources for generating a GT once a core concept pattern is discovered and theoretically sampled for.

Since concepts are abstract of time, place and people, accuracy is not an issue. Concepts are ranges varying from none to a strong presence as told by other conceptual properties of the core concept that give the core concept a value on the range. And the properties can come from other data. For example rooting (getting from here to there) is a core concept with many general implications depending on the BMRs (basic mobility resources) involved. I will not tell you the data that this theory was generated from. It is not necessary as its generality is so clear and applies so easily. We all root every day to work, to meetings, etc. with planning and budgets.

Coopting GT’s conceptual power, popularity etc. for certifying a paper, a meeting or a program occurs frequently by authors and professors. Certifying by reference to GT core variables is easily described at length since core variable general implications are so rich in description.   Then GT is lost to its being treated as grounded description. It is called the reversibility of interchangeable indicators. It is hard to stop the flow of indictors for a concept.

GT’s goal is to provide conceptual explanation of general patterns of behavior. It is not for verifying hypotheses like descriptive data is. Thus if GT is used to verify facts it easily becomes conceptual description and the GT power is lost to its descriptive use. And a GT has many indicators that can vary the GT conceptually. Some indicators may support the hypothesis and some disprove it. The loss of the GT goal and its power is great when a GT is reduced to descriptive verification.

Many novices and supervisors wish a double goal which is a grabby concept for a GT with lots of illustration. Conceptual theory suffers usually and description dominates. Lots of description easily crowds out conceptual theory with its rich empirical story talk about the general implications of a single rich core concept like super normalizing. Wanting credit for the rich full description wins, by giving many interchangeable indicators. Wanting full descriptive coverage is typical in rich, qualitative research that many colleagues want. Good grounded theory, concepts related to concepts, takes a minimum of illustration if at all. Often a good core concept is self-illustrating. No matter how rich conceptual description may emerge, it is not grounded theory. GT does not require full coverage. Indeed, often the data is forgotten for the emergent GT, which is abstract of the data and applies to many different substantive data. GT is not description, especially not full coverage.

In fact many a novice is never told about conceptual description not being GT and they pursue the former as the latter. They may discover later in the dissertation process that it is only conceptual description. Or they never discover it and firmly believer they did a GT thesis. Supervisors have a similar problem when they have spent a long career doing conceptual description, do not know GT and simply call conceptual description GT. They feel the jargon fits. They give up nothing methodologically, since a highly illustrated paper, whether conceptual, or just description without concepts, is the way to do it. In short combining concept with heavy description is the way to sucess. And it is a perspective that is often departmentally required for the dissertation as the OK perspective. Indeed many students who have done a GT (interrelated concepts) have been forced by their committee or supervisor to add much description to the theory to get their dissertation approved.

Conceptual description is forced as descriptive coverage of a topic or group of participants. It is ignored or forgotten or not known that with GT the data from which the GT was generated can easily be forgotten for its general implications for many other data. That GT is abstract of time, place and people is forgotten in favor of descriptive coverage.

Conceptual theory (a GT) can be confused as another type of description and then be related to conceptual description, thereby weakening the GT. A supervisor wrote me after reading this paper,

I just read a paper and this is exactly what the authors were doing. Even when they come up with concepts they just do not get the difference between empirically describing the findings from their study for a GT. A theory to them is not abstract of their full descriptive coverage that they considered the basis of their full data coverage that was the basis of their study.

It is hard for novices and supervisors to give up traditional QDA. And of course many cannot conceptualize. The individualized autonomy, provided by GT methodology is required of researchers for doing GT fully conceptualized plus the ability to conceptualize new concepts. This confusion is usually and often supported by departmental social structures, especially for the PhD process.

Both GT and description are products of different methodologies that get confused to conceptual description. This occurs often during a collaboration of two or more researchers trained in different methodologies, GT methodology being one and a QDA descriptive  methodology another. It is hard to give up one’s training and its vision and a departmental structure  of which there are many structures supporting QDA methods. The researcher sees the way of his methodology, his department perspective and his colleagues. QDA perspective, as conceptual description, wins for its simplicity compared to GT and for its simple methodology that has been in use many years before GT. The same applies to journals: papers are often returned with the request for more description.

Both products are generated from a methodology, conceptual description being far simpler and requiring less ability to conceptualize. And most people, including researchers, live on descriptive accuracy with little or no abstraction. Few individuals and researchers live conceptually. Thus learning to conceptualize and then using the concepts is highly individualized and requires autonomy. Not many researchers can hold the conceptual level according to GT methodology. Slipping down to conceptual description occurs easily and automatically as normal.

To be sure GT and description with a concept or two are both generated by a different methodology. Slipping to description is a slipping from the GT methodology to a descriptive methodology that to most researchers is known. The difference between the particularistic, routine normative data we all garner in our everyday lives and scientific data is that the latter is generated using a methodology. The method makes the product scientific. This may sound trite, but it is just the beginning of many complex research methodology issues. Whatever methodology may be chosen to make a research scientific has many implicit and explicit varying types of data collections.  That is what respondents, what pacing and timing for data collection, what type of analysis, etc. and what type of product (book, paper and/or lecture) is the goal. In the case of description or conceptual description, the explicit research goal must be decided.  Is full coverage wanted, how to achieve worrisome accuracy, how to interpret the findings etc., etc., how to give the actual generality of the data, what procedures to use? These issues and many more are debated at length in the QDA literature. Thus slipping from GT into conceptual description puts a non-GT claim of many issues on the ensuing analysis. GT is lost to another method with its own issues and problems. GT procedures are passed over and regaining them is doubtful.

There is doubtful correcting of the loss and of slipping away of conceptualization theory (GT) and its procedures. It is a growing problem for GT with still little recognition of the problem in the literature. Given the natural, automatic aspects of conceptual description and its integration into departmental structure and journal requirements, corrections back to GT are apt to come slow and hard to explain, however necessary.

GT is itself a grounded theory with conceptualization being the core category. We all know or have an idea what conceptualization is in general. It comes naturally to most of us, yet GT has procedures for generating emergent, discovered GT concepts. They must be followed for secured grounding. I discussed at length in Doing GT (1998), the conceptual license offered by GT methodology to the researcher. In exchange the GT researcher is supposed to use his procedurally generated, emergent, discovered concepts from his data, as opposed to using the reified concepts of conjecture offered by theoretical capitalists. Discovering one’s own concept from the research data with fit and relevance is usually very exciting for the researcher. Only 4 or 5 concepts are usually necessary for a GT. New concepts are one of the original contributions of GT.

Unfortunately this is where generating GT often stops by slipping into extensive description of the general implications of the core concept that was discovered. Further steps of generating a GT for the core concern are ignored for and in favor of description procedures. The GT perspective is lost. The freedom, autonomy and license required to generate and write conceptual theory that explains the continual resolving of a main concern is lost. GT methodology is very different from conceptual description which uses a QDA methodology. The GT methodology goal is the multivariate integration of concepts that is of patterns named as concepts and their properties, generated by the constant comparative method. Conceptual description is full  accurate descriptive coverage. Interchangeability of conceptual indicators is ignored and used as description. GT methods abstractly transcend all description methods and are never stale dated as descriptions soon are. Worrisome accuracy for GT is not an issue as with conceptual description. Most GT is generated from qualitative data since it is easier to obtain and less expensive than other data, but it can be generated from. And qualitative data easily becomes descriptive.

These are but a few of the vast differences in the GT and conceptual descriptive methodologies. The reader who knows GT methodology can probably think of many more differences. The reader who does not know GT methods will just have to believe me. He/she will pursue descriptive methods to generate conceptual description with no conflict and not realizing the great lose to GT. AS GT spreads slowly throughout the world, the methodological conflict between concepts and description grows and is being resolved in favor of GT conceptualization. It is social and psychological scientists who are mandated to conceptualize a theory and thus learn the rigorous GT procedures of concept generation so they could generate a GT.

The two most important properties of conceptualization that attract researchers are that they are abstract of time, people and place, and the naming of the concept usually has much grab. Thus concepts can live with use forever. One grounded concept can attract much delightful description, like supernormalizing.

The GT concepts must be generated from data: from a series of interchangeable indicators (see my book Getting Out of the Data). Concepts based on a single impression do not work, they must be based on the constant comparative method. That is comparing interchangeable indicators showing a pattern of behavior. Forcing a pattern on behavior to surmise a pattern does not work as grounded GT is a form of discovered latent structure analysis grounded in systematically collected data. Conceptual description is accurate description.

In sum the researcher should guard against letting a GT research slip into routine extensive description of one grounded concept.