Barney G. Glaser
Barney G. Glaser, PhD, Hon PhD, The Grounded Theory Institute.
This article explores the use of grounded theory to generate conceptualizations of emergent social patterns in research data. The naming of patterns and their abstraction across time, place and people, are discussed. The constant comparative method employed in grounded data analysis is offered as a developmental tool for enhancing researchers abilities to conceptualize and form emergent theories. Conceptual levels, descriptions, power and flawed approaches to analysis are explored at length.
Keywords: data analysis; data coding; theory development;
Glaser, B. (2002). Conceptualization: On theory and theorizing using grounded theory. International Journal of Qualitative Methods, 1 (2). Article 3. Retrieved DATE from http://www.ualberta.ca/~ijqm/
Conceptualization is the core category of Grounded Theory (GT). We all know or have an idea what conceptualization is in general. In this article, I will detail those properties of conceptualization which are essential for generating GT.
I have discussed at length, in Doing Grounded Theory (Glaser, 1998), the conceptual license that GT offers. The researcher can use his or her own concepts generated from the data instead of using, and probably forcing, the received concepts of others, especially those concepts of unduly respected theoretical capitalists. Actually generating a concept is very exciting and it is where many an effort at GT stops. This stopping is far short of doing GT through all steps to the end product. The GT perspective in this article will hopefully move more researchers further toward doing a complete GT.
In Doing GT, I endeavoured to emphasize the complexity of the world and therefore the freedom, autonomy, and license required to write generated theory that explains what is going on in this world, starting with substantive areas.
All that GT is, is the generation of emergent conceptualizations into integrated patterns, which are denoted by categories and their properties. This is accomplished by the many rigorous steps of GT woven together by the constant comparison process, which is designed to generate concepts from all data. Most frequently, qualitative data incidents are used.
Through conceptualization, GT is a general method that cuts across research methods (experiment, survey, content analysis, and all qualitative methods) and uses all data resulting therefrom. Because of conceptualization, GT transcends all descriptive methods and their associated problems, especially what is an accurate fact, what is an interpretation, and how is the data constructed. It transcends by its conceptual level and its 3rd and 4th level perceptions.
By transcending, I do not say implicitly that description is "bad", "wrong", or "unfavourable". Description is just different with different properties than conceptualization, yet these different properties are confused in the qualitative research literature. Actually, descriptions run the world, however vague or precise (and mostly the former). Precious little conceptualization affects the way the world is run. We have many immensely funded description-producing agencies such as newspapers, police, FBI and so forth, as well as an immense qualitative data analysis (QDA) research movement.
It is sociologists, psychologists, social psychologists, and other social researchers who are mandated to conceptualize in the social sciences. GT provides a systematic way to conceptualize carefully and its audience, though small, is growing. Yet at this point in time, 33 years after Discovery of GT (Glaser & Strauss, 1967b) was written, many social-psychological researchers still have little or no awareness of conceptualization, conceptual levels and, therefore, the integration of conceptual hypotheses.
The two most important properties of conceptualization for generating GT are that concepts are abstract of time, place, and people, and that concepts have enduring grab. The appeal of these two properties can literally go on forever as an applied way of seeing events. In this article, I start by explicating what a concept is for GT, then explain its abstraction from time, place, and people, followed by detailed discussions of conceptual ability required, conceptual levels, conceptual grab, conceptual description, conceptual conjecture, conceptual foppery and vagary, and conceptual power. Of course, much of this discussion overlaps.
For GT, a concept is the naming of an emergent social pattern grounded in research data. For GT, a concept (category) denotes a pattern that is carefully discovered by constant comparing of theoretically sampled data until conceptual saturation of interchangeable indices. It is discovered by comparing many incidents, and incidents to generated concept, which shows the pattern named by the category and the subpatterns which are the properties of the category. A GT concept is not achieved by impressioning out over one incident, nor by preconceived forcing of a received concept on a pattern of incidents. GT is a form of latent structure analysis, which reveals the fundamental patterns in a substantive area or a formal area.
The pattern is named by constantly trying to fit words to it to best capture its imageric meaning. This constant fitting leads to a best fit name of a pattern, to wit a category or a property of a category. Validity is achieved, after much fitting of words, when the chosen one best represents the pattern. It is as valid as it is grounded. Some reification cannot be avoided.
In Theoretical Sensitivity (Glaser, 1978), I said that many concepts are "in vivo" concepts; that is, they come from the words of the participants in the substantive area. Let me be clear: standard QDA emphasizes getting the "voice" of the participants. In vivo concepts are not such "voice," in the sense that what phenomenon they attribute meaning to with a concept is only taken as a GT concept, not taken as a description. The participants usually just give impressionary concepts based on one incident or even a groundless idea. They do not carefully generate their concepts from data with the GT methodology and try to fit many names to an established pattern. They are not establishing a parsimonious theory. They may have many concepts that do not fit or work. GT discovers which "in vivo" concepts do fit, work, and are relevant.
Inviting participants to review the theory for whether or not it is their voice is wrong as a "check" or "test" on validity. They may or may not understand the theory, or even like the theory if they do understand it. Many do not understand the summary benefit of concepts that go beyond description to a transcending bigger picture. GT is generated from much data, of which many participants may be empirically unaware. GT is applicable to the participants as an explanation of the preponderance of their ongoing behaviour which is how they are resolving their main concern, which they may not be aware of conceptually, if at all. It is just what they do! GT is not their voice: it is a generated abstraction from their doings and their meanings that are taken as data for the conceptual generation.
When naming concepts, GT does not try to give a "concern to understand the world of the research participants as they construct it" (Glaser, 1998). GT is not "an enquiry that makes sense of and is true to the understanding of ordinary actors in the everyday world" (Glaser, 1998). GT uncovers many patterns the participant does not understand or is not aware of, especially the social fictions that may be involved.
Time, Place, People
The most important property of conceptualization for GT is that it is abstract of time, place, and people. This transcendence also, by consequence, makes GT abstract of any one substantive field, routine perceptions or perceptions of others, since there is always a perception of a perception, and an abstraction from any type of data whether qualitative or quantitative. Hence, GT is a general method.
Thus GT conceptualization transcends. Conceptualization solves and resolves many QDA difficulties which are not abstract of time and place. QDA focuses on description of time, place and people, so is confronted with the problem of accuracy, context, interpretation, construction and so forth in trying to produce what "is". GT generates conceptual hypotheses that get applied to any relevant time, place, and people with emergent fit and then is modified by constant comparison with new data as it explains what behavior obtains in a substantive area.
Most writers on methodology DO NOT have a theoretical clue of what it means to be abstract of time, place, and people. The result is that GT is down-abstracted to just another QDA with some concepts. Strauss and Corbin (1998) do this in the following:
Grounded Theory procedures force us to ask, for example: What power is in this situation and under specified conditions? How is it manifested, by whom, when, where, how, with what consequences (and for whom or what)? Not to remain open to such a range of questions is to obstruct the discovery of important features of power in situ and to preclude developing further conceptualization. Knowledge is after all linked loosely with time and place. We carefully and specifically build conditions into our theories.
Thus, Strauss and Corbin force descriptions, irrespective of emergence, on the theory to locate its conditions, to contextualize it and to make it "appear" accurately pinned down, thereby losing its true abstraction and, hence, generalizability.
Personal distance for accuracy is supposed to be an "attitude" of the QDA researcher. The GT researcher, in contrast, does not need this attitude to get a description accurate, which is not his goal. The GT method automatically puts him or her on a conceptual level, which transcends the descriptive data.
In fact, it is hard to give up time, place, and people for many researchers as it is most natural, taught in QDA method classes, and requires an ability they may not have fully, if at all, developed.
GT in abstracting allows the researcher to develop a theory on a core variable, such as cutting back, supernormalizing, credentializing, cultivating, pluralistic dialoging, atmosphering, toning, abusing and so forth which can be applied to any relevant time or place. This delimiting by conceptualization stands in stark contrast to QDAs lengthy descriptions which try to cognitively map an area in a nontranscending way. There are no QDA rules for delimiting but arbitrary cut off points in the face of data volume and lengthy descriptive complexity. The QDA researchers need to be different, and hence appear original, is subverted by an inability to assimilate and extend a method devoted to description. GT, of course, simply generates original concepts and hypotheses as a result of the method.
Conceptualization is the medium of grounded theory for a simple reason: without the abstraction from time, place, and people, there can be no multivariate, integrated theory based on conceptual, hypothetical relationships. Concepts can be related to concepts as hypotheses. Descriptions cannot be related to each other as hypotheses since there is no conceptual handle. Concepts can relate to a description, but that single hypothesis is as far as it can go. While concepts are "everything" in GT, many researchers find it hard to stay on that level to relate them to each other. They relate the concept to a description and go on and on with description as QDA would have it.
Because GT operates on a conceptual level, relating concept to concept, it can tap the latent structure which is always there and drives and organizes behavior and its social psychological aspects, all of which are abstract of objective fact. For example, we have theories now of cautionary control, supernormalizing, default remodeling, desisting residual selves, atmosphering, toning clients, competitive knowing and many more (for examples, see my readers Glaser, 1993; 1995a; 1996), none of which could have been generated without conceptual abstraction.
While a GT may have been generated from a unit, or many units, if adequate theoretical sampling was used, it does not describe the unit, as I have said many times. The GT gets applied to the unit to explain behavior or any other unit in which the process has emergent fit. Thus, GT does not generalize from a researched unit to a larger unit or a similar unit, as a description might. GT generalizes to a transcending process or other form of core variable, and in the bargain may relate seemingly disparate units to each other by an underlying process. For example a 4 year medical school may be seen as the same as a 6 week contractor's course, from the point of view of credentializing, one property of which is to insure quality work. I have written at length on this in Theoretical Sensitivity (Glaser, 1978, pp. 109-113).
The delimiting aspect of abstraction is clear. A GT need not describe the whole unit, just a core process within it. Yet many a researcher doing GT finds it hard to give up unit wholeness or full, accurate description no matter how good they are at generating abstraction. Thus, they throw in many unit variables, e.g. face sheet data (sex, age, marital status, etc.) and context, that have not proven themselves as emergently needed in the generated theory, but are treated as if necessary to understand the GT. But they are not necessary unless earned into it by making a conceptual difference. Some authors even generate a good theory, such as host tolerance, but keep going on since, although it resolves the main concern, it does not fully describe the unit. Conceptual abstraction limits these "overdos."
Researchers not clear on the distinction between conceptual and description become easily confused on whether the theory describes a unit or conceptualizes a process within it. Listen to this confusion in the writing of a well-known QDA researcher, Jan Morse (1997b):
Conversely, qualitative researchers develop theory that as accurately as possible represents the empirical world. Data analysis consists of organizing reality with inferences that are subsequently systematically confirmed in the process of inquiry. Qualitative data theory, as a product, is abstract but consists on minimal conjecture. These theories are rich in description, and the theoretical boundaries have been derived from the context and not from the researcher's arbitrary goals for delimiting the scope.
Recall that theory is theory, not fact, but as a theory is confirmed, it is moved into the realm of fact and is no longer theory. Because of the abstract nature of qualitative data theory, and because of the conceptual nature of knowledge, QDA, by and large, remains at the level of theory, not fact. Theory is not reality, but our perception or organization of reality, perhaps closely resembling reality, but not reality per se. It remains a representation of reality, malleable and modifiable.
In Morse's work, as the reader can see in this telling example, there is no end to the confusion and mix between abstraction and concrete phenomenon or between conceptualization and description.
It is hard for many a GT researcher to only relate concepts and not relate concepts to people. People become labelled or actioned by a concept like it is their whole being. In GT, behaviour is a pattern that a person engages in, it is not the person. For example, he is a cultivator of people for profit; this is not GT. In GT, a person engages in cultivating behavior and cultivating is a basic social process. "Labelling", a well known sociological theory, has no descriptive place in GT. People are not categorized, behaviour is. Or labelling as a basic social process does occur: people are labelled by laymen and processed accordingly. But the GT theorist only describes the labelling of people pattern as a social control process.
Perhaps the most important aspect of conceptualization is that concepts last forever, while descriptions are soon stale dated. Concepts are timeless in their applicability. For example, awareness context theory can still be applied 33 years later (see Awareness of Dying, Glaser & Strauss, 1967a). One author (King, Keohane, & Verba, 1994) makes the point that concepts even last longer than any of the hypotheses in which they were originally generated:
Max Weber has suggested that the essence of social theory is in the creation of clear concepts. Indeed, concepts such as charisma or division of labour or reference groups have been longer lasting than any valid claims about the causal effects of these concepts. Many such concepts guide our thinking and theorizing today!
Since GT concepts are rigorously generated and since they can make a continuing mark on us compared to soon outdated QDA descriptions, the GT researcher has been set up to become enduringly famous or achieve reknown for his or her best concept. The longevity and history of the concept is continually associated with his or her name (see, for example, the concept of emotional work by Hochschild, 1983). GT does offer fame or reknown to its researcher as opposed to QDA. It offers, also, freedom from received concepts and theoretical capitalism, because, however enduring received concepts may be, often these concepts are pure conjecture of little relevance.
Just think, the GT researcher can generate a theory, such as pluralistic dialoguing, that can be applied over and over for a 100 years or more. What significance for his or her GT research!
GT is an advanced graduate level methodology used for MA and PhD theses. This statement assumes that there has been an educational institutional sifting that brings people who have conceptual talent to this level. This institutional sifting of ability puts by the wayside those who have too much difficulty in conceptualizing.
But at this level ability still varies between the few who got through and have no ability to conceptualize, those who conceptualize sufficiently well, and those who are very capable. The former drift into QDA, whatever the methodology, and use a few received, preconceived concepts for a forcing framework that yields lots of description. The QDA approach is just waiting to be used by those unable, or with little ability, to conceptualize.
At this level, many who can may still have some skill difficulty in conceptualizing clearly with its meaning. GT provides the method (constant comparisons) which develops skill to generate the researcher's own concepts and at the same time gives him or her the legitimacy to not jump to using the received concepts that would force the data. So, though some researchers cannot conceptualize, many more researchers can, than have, heretofore, generated their own categories for a study. In the beginning they no doubt will exhibit their unskilled, untrained but strong ability to conceptualize using GT techniques. But ability grows as these techniques are developmental.
Some people have a natural ability to conceptualize based on data. One external examiner (Anonymous, personal communication, 1999) of a PhD dissertation wrote me of her candidate:
Her grasp and skill in the use of grounded theory methodology is high for her career level. Whatever problems she may have can and should be solved in her next study. Her generation of substantive categories and their properties achieve parsimony and scope and fit and worked well. Her theoretical coding was very weak, but no one at her level is strong on this.
I have known many other candidates who are similar in conceptual skill (see, for example, Brooks, 1997; Gibson, 1997). They all show that GT taps methodologically what many people do normally: conceptualize what is going on in their everyday life, as it now goes on in their research.
Those researchers with little or no conceptual ability, but who have made it to the graduate level, also have no notion of what the constant comparative method, interchangeability of indices, theoretical saturation, theoretical sampling, sorting, memos, delimiting, and so forth are, since one needs conceptual ability to understand these methods based on conceptualization. Their research results, again, are QDA, based on received concepts, preconceived problems, models forcing the data collection and analysis, and positivistic data collection techniques.
If the researcher can conceptualize, then he or she will trust to emergence of a theory. It's part of their vision and realization that concepts will emerge. Emergence of concepts often happens fast, even too fast, and the research must be slowed a little to check out best fit concepts and their saturation.
If the researcher CANNOT conceptualize, then he or she will not trust to emergence of theory. How could the researcher trust the idea of emergence, when confronted with volumes of unwieldy data, which he or she cannot get out of, and more specifically, cannot transcend conceptually. For the unable, there are few or no patterns, no delimiting, just endless description, uncontained except barely by a preconceived conceptual framework or a preformed model. They have no conceptual pick-up and its associated problems as they try to analyze the data. They miss many rich concepts and subproblems associated with them as they endeavor to describe. They even miss the participant's main concern and its resolving. I get astounded at this lack of conceptual pickup as they tend to believe as normal that which is really a problem. For example, in an organization designed to enhance creativity, people are subject to confrontation sessions to make them think of creative solutions to the challenge. The researcher did not realize that this technique can easily scare, stultify, and stun a creative mind.
This is why researchers who cannot conceptualize reach out for, and even need, the theory of the participants, however particularistic and low level it may be. They forget that the participants are the data, NOT the theorists. The participants, while having great involvement in resolving their main concern, seldom have a conceptual perception of it as a GT theorist does. And once "some" participants with conceptual scope hear a grounded theory of their abiding concern, they use it as power to help resolve it. Relevant concepts have this enlightening effect.
This is why, also, nonconceptualizers turn to, and need computer software to engage in rote sorting based on forced, received categories. With no ability to conceptualize, the researcher defaults to accuracy problems of description, particularly the popular ones of social construction of data by interviewers and the interpretation distortions by the interviewer. They have no way of transcending the data to get at the core pattern and subpatterns and indeed finds it hard to see patterns much less name them. They are left to total description, not conceptualization.
A researcher who can adequately conceptualize can use GT and have a good career. But remember, GT is merely a research option. If conceptualization is difficult for the research, he or she can easily have a good career in one of the many very worthy QDA methods. I know many a very intelligent researcher who cannot conceptualize. They are linear describers, good at empirical generalizations, and have successful careers. Doing GT or not is usually a self-selection phenomenon. Although being able to conceptualize, or lack of ability, may have to, or should, be pointed out to the researcher by a friend, colleague or supervisor, though it seldom is.
It works both ways. Good conceptualizers should be made aware of their ability if they do not know, so they can move faster into conceptual work and skill development. I was never advised by a teacher how good I was at conceptualizing which slowed my career 15 years. By the same token poor conceptualizers should be advised to "not fight it", and put their efforts into another QDA methodology. This other-reflexivity is vital in the formative years. It is a very sensitive topic for most researchers to hear about themselves, so be careful.
I have written at length on conceptual leverls in Doing Grounded Theory (Glaser, 1998, pp. 135 - 139). So here I wish only to make a few additional observations. Bear in mind, categories are generated from the data and properties are generated concepts about categories. There are many examples of this in my readers. Once discovered, concepts leave the level of people. They become the focus of the research, to be later applied to people's social psychological behavior. As the theory generates, and integrates through memoing and sorting, the conceptual level goes up, often from substantive to formal theory (see my reader, More Grounded Theory Methodology, Glaser, 1995).
The goal of GT is to arrive at minimally least the third level of conceptual analysis. First is collecting the data, then generating categories, then discovering a core category which organizes the other categories by continually resolving the main concern. From substantive theory one can go to a higher level, called formal theory. For example, becoming a nurse, a substantive theory, can be generalized to becoming a professional, a formal theory, and even raised to a higher formal level of becoming in general, a theory of socialization. This is done by theoretical sampling and constant comparisons (see More Grounded Theory Methodology, 1995).
Many researchers do not understand or realize the nature, use, and power of categories and their properties. Most fields of study are just properties of a substantiated category or two followed by extensive description. If you listen carefully, one hears of many properties of a category as an academic or professional talks with erudition about their field. The "best" in a field knows the most properties of the category and their most intricate integration. For example, students of non-political social movements just talk of the various properties of this category and a few sub-categories. Properties vary categories. Systems are complex ways properties of categories relate to each other and vary one category in relation to other categories. Once the research is on a conceptual level, dealing with categories and their properties becomes the mode, description fades and abstraction of time, place and people take over. The concepts do not go away: EVER!
Enduring Grab: Theory Bits
Concepts in general, whether conjectured, impressionistic or carefully generated by GT, have instant "grab". They can instantly sensitize people, rightly or wrongly, to seeing a pattern in an event or happening that makes them feel they understand with "know how". In a word, the person feels like he or she can explain what they see.
It is impossible to stop the grab of theory bits. The person talking with them can show his or her skill and power. They can be applied "on the fly", applied intuitively with no data with the feeling of knowing. They ring true with credibility. Use of concepts empower people, they compete in conversation well and in theoretical discussions. They are exciting handles of explanations, running fast ahead of the constraints of research. They can become stereotypical and routine. As enlightening as a concept is, it can cognitively stun others in the presence of users, as they stop thinking since all appears understood.
GT emphasizes the productive use of conceptual grab by generating relevant concepts that work and are integrated into a theory. It keeps grab under control and makes its endurance secured by working it in application. Of course, many will do and use GT with no idea of this property of conceptualization, but no matter. Even the concept "grounded theory" is used freely to orient and legitimate QDA research to its consumers and colleagues. It has enduring grab. QDA researchers muzzle in on the grab of GT concepts to perk up their descriptions. This bit use works well in QDA.
While interviewing a buddhist swami who studied philosophy of religion at University of California, Berkeley, he told me he did not get much from most of the course, but what he did get of a lasting nature were two concepts from a famous sociology professor of religion, Robert Bellah. That's all he really remembered. He said Bellah was a genius, but was referring to the two concepts.
Thus we have the enduring grab which carries with it the great respect of the concept generator. Some of these generators, whether grounded or conjectured, produce concepts that turn them into idols and theoretical capitalists, virtual owners of their concepts. GT allows many to claim this power modestly and with the integrity that comes from grounding, while at the same time it reduces the magic of a few conjectured concepts coming from a learned man or woman. GT puts morality into this power as it brings it into reach of many researchers, and takes it out of the hands of a few.
Listen to this description of a famous theoretician, Robert K. Merton, which lauds the "grab" of his concepts that organize the patterns of much public opinion data. The grab of received concepts is used in fields having little or no conceptualization. A conceptualizer, like Robert K. Merton, has accumulated vast conceptual capital and acts like a theoretical capitalist. He controls the capital of production by requiring that his students use his received, conjectured concepts.
Robert K. Merton has shown us how to relate opinion findings to the history of ideas and to general social theory. He has demonstrated the use of many conceptual tools, for example, "reference groups" and "local and cosmopolitan" in interpreting the puzzling variation of public judgements. His concept of status-set has given theoretical relevance to the demographics included in every poll. His concept of status sequence helped us understand the opinions of socially mobile people.
As a theorist, Robert K. Merton, has identified and clarified a large number of the social mechanisms that constitute the present state of sociological knowledge. Several of them are highly relevant to the field of public opinion, for example, the self-fulfilling prophecy. His much quoted paradigmatic essays, Social Theory and Social Structure, are brilliant in content and in style; many of their passages qualify as aphorisms.
In appreciation of these many contributions, the world association for Public Opinion Research is pleased to present the 2000 Helen S. Dinerman Award to Robert K. Merton. (Speech recounted in personal communication from H. Zetterberg).
None of these concepts are truly grounded and they pale in the flood of grounded theory concepts coming from dissertations the world over. It is only time before the enduring grab of grounded concepts yield awards to their author! The awards can easily make him a great man in some quarters. Students can be required to use these concepts to honour a "great man". Rightly or wrongly, the enduring grab of these grounded, as opposed to conjectured, concepts organizes the patterns in a volume of otherwise disparate data. GT makes sure the concepts are right and organized right as they are grounded in the data. GT can generate with ease many concepts like Merton has conjectured with seemingly great thought. The magic and mystique of these conjectured or impressionistic conceptualizations is over.
The next sections build on these notions of enduring conceptual "grab".
This article is about grounding concepts in the data to which it will be applied. This ensures fit, relevance and workability. In contrast, conjectured concepts, from wherever, appear to be useful given their grab, hence apparent but probably with forced fit and relevance. They are all over the literature, and many are woven into grand theory, based on logical deduction, to wit, conjecture. Once applied they often do not work either on the conceptual level or data level. But they spawn many theory bits, to make the author or talker appear learned.
Conjectured, logically deduced concepts became more and more viewed as very "airy", too abstract, too reified, irrelevant and not workable in the 1960s. Lofland (1976) wrote about "undisciplined abstraction leading to concepts which bear little relation to the social world that they are supposed to refer to, either because they are not apparently based in any empirical research or are wondrous elaborate edifices of theory based on very little empirical research." Lofland calls for an empirical science "that gives rise to concepts, yet contains and constrains them by a context of concrete empirical materials."
Ten years before this statement Discovery of Grounded Theory (Glaser & Strauss, 1967) was written in response to this often empty, yet grabbing, conjectured conceptualization. It also gave impetus to QDA research in order to make research "get real" or "grounded", with little or no conceptualization. Barry Gibson put it well: "GT as a method was developed in opposition to the grand theories of the time and argued for all theory to be grounded in observations of data. The researcher should approach the data with as few preconceptions as possible in order to see what is going on" (personal communication, 1997). Preconceptions often mean little or nothing when derived from received grand theory concepts!
At the same time that Discovery of GT was written, Robert K. Merton (1967) wrote " theory, is logically interconnected sets of propositions from which empirical uniformities can be derived." This statement from a famous theorist such as Merton was virtually a license to conjecture theory. The researcher at the time experienced great cross pressure between license for logically deducing theory and our demand that it can only be grounded to be relevant.
Reification was out, grounding was in, and the problem arose as to just how many researchers had conceptual ability to change to grounding. Our answer was more researchers than were heretofore allowed to. Anselm and I gave researchers the license to generate concepts on ones own and ignore the theoretical capitalists, idols, and great men. They were always there for those who could not conceptualize on their own.
The research goal changes from theory based on the unobserved to explain more unobserved, to GT theory generated from observation to explain the observed. In the standard appeal to future research of a GT article, the author could conjecture a bit based on conceptual deduction from grounded concepts as he suggests next research steps; the only conjecture allowed which is a form of theoretical sampling.
Put simply, in putting a constraint on conjectural theory, based on the unobserved and thereby distrusted, GT gained theoretical leverage for social psychology in generating theory that could be trusted as fitting and relevant to the data. Social theory was rescued by GT from its growing disaffection by laymen and from being discounted as too airy by them. GT made theory accountable to others as it was right on and worked.
Conceptual description is a frequent occurrence among researchers trying to do GT. They even call it GT, but it is really a form of QDA. One concept is generated and then the researcher spends the rest of his time describing it and describing it with incident after incident. There is little or no constant comparative work to generate conceptual properties of the category based on the interchangeability of indices and conceptual saturation. The researcher just "incident trips" with no more conceptual analysis. I have written about this in Doing GT (Glaser, 1998, pp. 162 - 165). The researcher just goes on and on overdescribing with no more conceptual analysis. Story after story is forced into the concept. It reads like story sharing at a party once a sharing topic emerges. It is conceptual grab by one concept gone wild.
There are two types of interrelated sources of conceptual description: the researchers skill and the concepts source. It is done by researchers with limited ability to conceptualize and to relate concepts theoretically. The thrill of generating one concept is all they need and can handle. Close colleagues, who are use to QDA, can unawaredly pressure the researcher into lots of description. Some researchers cannot handle the tedium of conceptualization by constant comparison while coding, collecting by theoretical sampling and analyzing, writing up theoretical memos and saturating the concept. In actuality a researcher may get to 2 or 3 concepts, which is even more strain on conceptual theorizing for the conceptually unable. Thus, we see many conceptual descriptions.
Mixed into these sources of conceptual description among researchers are the sources of the concept. One major source is to read through an interview 2 or 3 times for an overview of whats going on and conceptualizing it. This overview impression source of a concept can be so grabbing that incident tripping (called description by others) takes over and thats it. Since the rigor of line by line study of the interview and then careful comparing incident to incident and then to concept, and constant theoretical sampling, is abdicated, there is no control over the tendency to conceptual description. In QDA fashion it appears the way to explicate the concept. My readers (see Glaser, 1993; 1995a; 1995b; 1996) give many examples of this.
Another source of concept that leads to incident tripping is the one incident concept. It is usually way over analyzed and once the incident is conceptualized the tendency is to describe it ad infinitum. A one incident concept is not a careful generation of a pattern. It is not a pattern. It is a conjecture. Its name can have tremendous grab and since there is no more analysis, conceptual description ensues. This happens frequently with beginning researchers who may discuss at length what one incident means. They are not used to, or trained to, continual coding and analyzing and, therefore, conceptually saturating the concept, thereby discovering its properties and that it is a true pattern. The one incident may prove to be just that: ONE INCIDENT, and not relevant to much. In GT, the researcher must keep moving through the data to see the incident over and over and constantly be comparing and conceptualizing. This is not easy. Researchers easily default to QDA.
To be sure, conceptual description can also result from a carefully generated concept. The concept can have so much grab and the researcher be so thrilled, with no taste for the tedium of constant comparison, and so on, that describing ensues at great length.
A colleague also advised me that there are fields with an overload of received concepts, such as his: business management, where one starts with the received concepts and tries to force many incidents into the meaning of the chosen concept. Conceptual description ensues. It is preconception to the max! Bengt Gustavsson, teacher and author of a GT book in Swedish, says in a writing to me:
to repeat, my main problem as a management teacher is to deconstruct all the fancy, but ill-grounded, conceptions in the management field, e.g., knowledge management, customer relationship marketing, virtual organizations employability, and intellectual capital. It turns out that although they are in vogue, nobody really knows what they stand for (personal communication, 1998).
Fields like this may have many reified concepts (concepts with no empirical referents) so even conceptual description is hard. GT in these fields puts grounded concepts into direct competition with the field's conceptual jargon, which itself can cause conceptual description to undo the jargon.
The dosage mix between conceptual hypotheses and conceptual description is, of course, the prerogative of the GT researcher. It is his product.
Closely related to conceptual description but at the other end of the continuum is conceptual foppery. This occurs when every incident in sight is conceptualized with no theoretical meaning deriving from a clear focus analysis of a main concern and its resolving. Dozens of one incident concepts are offered with no problem, no theoretical integration, no constant comparison, no realization of prime mover of the people's predominant behavior, no proven patterns, etc. No parsimony and scope are sought. The license to generate has gone wild! Too many concepts go every which way with no relation to "whatever"! Conceptualization in GT must be done as a careful part of theory generating and emergence, with each concept earning its way with relevance into the theory.
Concept foppery is engaged in by researchers who can conceptualize one incident easily, but who cannot compare many incidents. They do not seem to see patterns. The sensitivity of the concept is enough to imply patterns. They also dip into reification easily without knowing it. There are academic subfields in business and psychology where concept volume, grounded or not, is the thing with little or no regard for data.
Close to concept foppery is concept vagary. The concept rings true and is sensitizing at first blush. But then it seems vague or sufficiently undetailed so the reader does not know what is being referred to or discussed. The grab starts to diminish. Theory bits are often quite vague upon examination. People use concept vagary like descriptive vagary, in order to not reveal sensitive information. It obfuscates. If vague concepts are submitted to conceptual description, the description becomes vague also.
GT, of course, does not produce and, perforce, does allow concept vagary if the category and its properties are carefully generated and saturated conceptually. By definition the concept fits with clarity. Properties of a category substruct it by showing its degrees, dimensions, aspects, types, etcetera. These properties become the categorys clear definition.
Given our above discussion of conceptualization, especially that of conceptual grab, most people feel the power of conceptualization and its ability to transcend the descriptive, its ability to generate "wise" propositions that explain behavior in an area (especially its main concern), its ability to organize and make meaningful many seemingly disparate incidences, and its ability to be used or applied as a wise academic and/or consultant.
GT gives the researcher this power. He or she becomes an expert in the substantive area of their study. The researcher is asked to guest lecture and to consult on the subject. For example, one grounded theorist who did his dissertation on "enhancing creativity" in bureaucratic organizations gives lectures to many corporations on this subject as an expert.
Most GT researchers do not fully realize this power of abstractedness. They allude to the power, but cannot quite formulate it or be precise about it. Yet it is up to the GT researcher to use it and give it and show it to others not only in the substantive field from whence it was generated, but in other areas, and to laymen. The GT power gives control by its sensitizing, enduring grab, its generalizability and its being abstract from time, place and people!
Listen to Wendy Guthrie (2000) allude to this power while not quite being precise about it:
The veterinarian is undisputed in his ability to describe the way things work in his domain. He does not welcome an outsider telling him what he already knows. The explanatory power characteristic of grounded theory is where the difference lies. It gives practitioners a new conceptual understanding and control over actions like never before. Even when confronted by novel situations, conceptual theory yields powerful predictive ability enabling those with access to it to influence the direction ahead with recourse to appropriately selected strategies. Grounded theory elevates the theorist to an influential level. Description on the other hand, complicates and distracts.
Guthrie is very close to seeing the power and control of GT. Others are not so close, but feel and see it strongly. Listen to Jan Morse (1997b),
...the identification of theory or a model that enables the case study to generalize to other cases. It is important to note that this linkage is done conceptually -- in this case using the concepts of enduring and suffering -- and it is this abstraction from the descriptive data that makes the study more powerful.
So why not just go for the power of GT? Timidity is the answer Jan Morse (1997b) says:
Qualitative researchers are theoretically timid. Some researchers may be more comfortable staying within the safety zone of their data. They may be unwilling to take the risk inherent in interpretation and move their analysis beyond the descriptive level. Theorizing is also work: often researchers make the mistake of submitting their study for publication without making the effort to do the conceptual work necessary for the development of theory.
She is no doubt right about timidity, but not doing the conceptual work is not quite correct. They simply do not know nor have training in GT, so they are at loggerheads on how to do conceptualization. After learning GT, I have seen many a researcher lose the timidity that comes with knowing how to conceptualize and feeling its power.
Lastly, Jan Morse does not realize the full power of GT. So she sees a continuum with no power valence. She says, "Qualitative research as a product may be classified according to the level of theoretical abstraction. This ranges from the most descriptive research to the most abstract, generalizable research, and it is outlined in Table 9.2." (Morse, 1997b).1 The categories of her outline are neutral to the growing power of abstraction as QDA turns to GT. QDA and GT are confused, as always.
Barry Gibson, a well known GT researcher, extols the conceptualization of GT but does not clearly extol the power of it. He says:
As stated previously the procedures underlying grounded theory are not designed to yield themes within the data, but are aimed at developing a theory. By theory what is intended is a conceptual account of the 'main concerns' of those resolving a particular problem. Grounded theories are therefore conceptual communication concerning observation on how persons resolve particular problems in a particular area (such as health care or business organizations) (personal communication, 1998).
He is quite correct but he does not say that these conceptual communications on accounting for the behaviors in a substantive area are quite powerful. So near and yet so far.
King, Keohane, and Verba, political science researchers, actually turn to briefly bring out the power of concepts in the following: "Compelling concepts need not be part of a valid causal inference to be powerful, but to remain powerful, these concepts must be part of a research agenda that seeks to identify their systemic implications, revealing their link on a causal chain." To be sure the power of conceptual grab is there and it does endure, not fade. But endurance is enhanced by the concept being integrated into a grounded theory that endures also.
In my section on conceptual grab above, we have seen the power of Robert K. Mertons concepts to organize with meaning public opinion findings and to relate them to general social theory. Exercising this power brings reknown, even fame, since there are so few good conceptualizers. GT, a rigorous conceptualizing method, gives many an average researcher a chance at such renown.
This is a long paper, and obviously could be longer since there are many more properties of conceptualization. I have tried to discuss those most pertinent to GT.
1 The table can be found in Morse, 1997b, p. 174. The four types of qualitatively derived theory referred to are 1) descriptive, 2) interpretive, 3) disclosive, and 4) explanatory. Back to text
Andriopoulos, C. A. (2000). Mind stretching: A grounded theory for enhancing organizational creativity. Unpublished doctoral dissertation, University of Strathclyde, Glasgow.
Brooks, I. J. (1998). Weighing up change: A grounded theory explaining the response of middle managers to organisational change. Unpublished doctoral dissertation, University of Canterbury, Christchurch.
Burawoy, M. et al. (1991). Ethnography unbound: Power and resistance in the modern metropolis. Berkeley, CA: University of California.
Calvin, A. O. (2000) Hemodialysis patients and end-of-life medical treatment decisions : a theory of personal preservation. Unpublished doctoral dissertation, University of Texas, Austin.
Coup, A. (1998). Being safe and taking risks: How nurses manage childrens pain. Unpublished masters thesis, Massey University.
Denzin, N. K. & Lincoln, Y. S. (Eds.). (1998a). Collecting and interpreting qualitative materials. Thousand Oaks, CA: Sage
Denzin, N. K. & Lincoln, Y. S. (Eds.). (1998b). Strategies of qualitative inquiry. Thousand Oaks, CA: Sage.
Denzin, N. K. & Lincoln, Y. S. (Eds.). (1998c). The landscape of qualitative research. Thousand Oaks, CA: Sage.
Gibson, B. (1997). Dangerous dentaling: A grounded theory of HIV and dentistry. Unpublished doctoral dissertation, Queens University of Belfast, Belfast.
Glaser, B. G. (1978). Theoretical sensitivity: Advances in the methodology of grounded theory. Mill Valley, CA: Sociology Press.
Glaser, B. G. (1992). Basics of grounded theory analysis: Emergence vs forcing. Mill Valley, CA: Sociology Press.
Glaser, B. G. (Ed.). (1993). Examples of grounded theory: A reader. Mill Valley, CA: Sociology Press.
Glaser, B. G. (Ed.). (1995a). Grounded theory1984-1994 (2 vols.). Mill Valley, CA: Sociology Press.
Glaser, B. G. (Ed.). (1995b). More grounded theory methodology: A reader. Mill Valley, CA: Sociology Press.
Glaser, B. G. (Ed.). (1996). Gerund Grounded Theory: The basic social process dissertation. Mill Valley, CA: Sociology Press.
Glaser, B. G. (1998). Doing grounded theory: Issues and discussions. Mill Valley, CA: Sociology Press.
Glaser, B. G. (2000). The discovery of grounded theory. Mill Valley, CA: Sociology Press.
Glaser, B. G., & Strauss, A. L. (1967a). Awareness of Dying. Chicago: Aldine.
Glaser, B. G. & Strauss, A. L. (1967b). Discovery of grounded theory: Strategies for qualitative research. Hawthorne, NY: Aldine de Gruyter.
Gummesson, E. (2000). Qualitative methods in management research (2nd. ed.). Thousand Oaks, CA: Sage.
Gummesson, E. (1999). Total relationship marketing : from the 4Ps-product, price, promotion, place- of traditional marketing management to the 30Rs-the thirty relationships- of the new marketing paradigm. Boston, MA : Butterworth-Heinemann.
Guthrie, W. (2000). Keeping Clients in Line: A Grounded Theory Explaining How Veterinary Surgeons Control Their Clients. Unpublished doctoral dissertation, University of Strathclyde, Glasgow.
Hartwig, F. & Dearing, B. E. (1979). Exploratory data analysis. Beverly Hills, CA: Sage.
Hochschild, A. R. (1983). The managed heart: Commercialization of human feeling. Berkeley, CA: University of California Press.
King, G., Keohane, R. O., & Verba, S. (1994). Designing social inquiry: Scientific inference in qualitative research. Princeton, NJ: Princeton University Press.
Lee, R. M. (1995). Dangerous fieldwork. Thousand Oaks, CA: Sage.
Lofland, J. (1976). Doing social life: The qualitative study of human interaction in natural settings. New York: Wiley.
McCallin, A. (1999). Pluralistic dialogue: A grounded theory of interdisciplinary practice. In Conference proceedings of the New Zealand Rehabilitation Association Biennial Conference (pp. 115- 122). Dunedin, NZ: New Zealand Rehabilitation Association.
Marshall, C.. & Rossman, G. B. (1999). Designing qualitative research (3rd ed.). Thousand Oaks, CA: Sage.
Merton, R. K. (1967). On theoretical sociology: Five essays, old and new. New York: Free Press.
Morse, J. M. (Ed.). (1997a) Completing a Qualitative Project: Details and Dialogue. Thousand Oaks: Sage.
Morse, J. M. (1997b). Considering theory derived from qualitative research. In J. M. Morse (Ed.), Completing a Qualitative Project: Details and Dialogue (pp. 163 - 188). Thousand Oaks: Sage.
Nathan, H. (1986). Critical choices in interviews: Conduct, use, and research role. Berkeley, CA: Institute of Governmental Studies.
Ragin, C. C. (1987). The comparative method: Moving beyond qualitative and quantitative strategies. Berkeley, CA: University of California.
Silverman, D. (2000). Doing qualitative research: A practical handbook. Thousand Oaks, CA: Sage.
Starrin, B., Dahlgren, L., Larsson, G., & Styrborn, S. (1997). Along the path of discovery: Qualitative methods and grounded theory. Lund, Sweden: Studentlitteratur.
Strauss, A., & Corbin, J. (1998). Grounded theory methodology: An overview. In N. Denzin & Y. Lincoln (Eds.), Strategies of qualitative inquiry. (pp. 158-183). Thousand Oaks, CA: Sage.
Zetterberg, H. L. (1997). Sociological endeavor: Selected writings. (R. Swedberg & E. Uddhammar, Eds.). Stockholm: City University Press.
IJQM Editor: Janice Morse
Editors note: A variation of this article has been previously published as chapter 2 of:
Glaser, B. G. (2001). The grounded theory perspective: Conceptualization contrasted with description. Mill Valley, CA: Sociology Press.