No Preconception: The Dictum

Barney Glaser, PhD, Hon. PhD

I would like to begin and introduce this book on “no preconceptions” when doing grounded theory (GT) with a short trip of 45 years into the past by quoting the reasoning source of the no preconceptions dictum as first laid out in 1967 in the Discovery of Grounded Theory, by Barney Glaser and Anselm Strauss. The sources where (1) the zeal for verification of conjectured hypotheses research and (2) to explain the findings with theoretical capitalists demanding and commanding conjecture seldom if ever tapping the reality of what was really going on. Grounding induced theory in research data was what was needed.

Our first paragraph in Discovery reads as follows: “Most writing on sociological method has been concerned with how accurate facts can be obtained and how theory can thereby be more rigorously tested. In this book we address ourselves to the equally important enterprise of how the discovery of theory from data – systematically obtained and analyzed in social research – can be furthered, We believe that the discovery of theory from data – which we shall call grounded theory – is a major task confronting sociology today, because as we shall try to show, such a theory “fits” empirical situations and is relevant with understanding to sociologist and layman alike. Most important, it (GT) works by providing us with relevant predictions, explanations, interpretations and applications.

To achieve this goal we generated a methodology which we called grounded theory methodology which had, and still does have, many rigorous steps to achieve grounding. One aspect of GT was to stop hypothesis testing that was irrelevant and drew on conjectural theory explanations, by grand theorists – theoretical capitalists. These irrelevant preconceived tests yielded the dictum that No preconceptions were allowed. This dictum applies to the general research problem, the specific participant problem, what pre research conjectured theoretical categories and their connections would apply, and thus will provide the preformed explanations and in what theoretical shape. And preconceptions get even more subtle based on theoretical perspective assumptions and remodeled GT methods. I will lay out many of these utilities in this book. I saw many a research fail in those days because preconceived research and theory yielded no theory and findings of fit and relevance and workability.

As the reader knows, this position taken 45 years ago has flowered and boomed. Grounded theory today is used all over the world, principally for PhD theses and then in subsequent research of those GT PhD’s. We were sufficiently correct to open up a whole new world of theory generation no matter what the latent theoretical perspective of GT researchers have as academics in health, management, social work, political science, business and sociology. No preconceived research works as GT. But the world wide use of GT or supposed GT versions has increased our knowledge of the subtlties of requiring no preconception or giving the arguments for preconceiving research aspects in some ways. I hope to detail many of these subtleties in this book so the reader can be aware of what it means to suspend preconceptions in service of emergent generating of theory.

As we said in Discovery of GT, part of the trend (in 1960’s) toward emphasizing verification was the assumption by many sociologists that our “great men” and theorist forefathers (Weber, Durkheim, Simmel, Marx, Veblen, Cooley, Mead, Park etc) had generated a sufficient number of outstanding theories on enough areas of social life to last for a long while. Current great men such as Merton, Parsons, Homans, Blumer, and Goffman, to mention a few, continued their “think up” theories. Of course, GT will not replace these theories but the shedding of their claim to preconceive research and theory writing will, and has significantly occurred, in the research world of today. The GT researcher may not become a great man, but at least his/her GT theory will be done with autonomy and originality and will be a contribution he/she is known for in the literature. There are hundreds of substantive grounded theories now as of 2012. No preconceptions clearly work for the emergent discovery of GT.

Reviewing the Dictum

In the remainder of this chapter I will review the no preconceptions dictum in some detail. I have said over and over in my many writings that the researcher should not preconceive in doing GT research: 1. the general problem, 2. the specific participants problem, 3. what received concepts will explain the current behavior, 4. what theoretical code will integrate the theory, and 5. what theoretical perspective applies. The rule is to let these areas emerge. Discover them. The researcher cannot preconceive what he will discover by staying open to the emergent. What is allowed is a general area of interest coupled with a humble lack of knowledge of what problems may exist in the area.

I have emphatically cautioned against using extant concepts of a field by reading the literature in a field of study before the emergence of a substantive theory. Indeed, the researcher will likely not know what literature applies before his/her theory emerges. This stance is important so the researcher is not likely to be tempted or feel required to use preconceived literature concepts for coding. And especially to not use these “received before emergence concepts” to solve the initial confusion that usually arises when starting conceptual coding of the collected research data.

Keep in mind that preconceived concepts do not have to be forgotten. They are just to be suspended for the GT research so the researcher is open to the emergent. Why let them get in the way? Sure, they may have legitimate power as sanctified by the literature, but this power must be ignored or resisted. Otherwise it will take over and stop the generation and subsequent power of a classical substantive GT with fit and relevance that works in explaining what is going on. Many advanced GT researchers have said in response to the dictum of no preconceptions how realistic it is for the “getting out of the data” a genuine substantive GT theory.

Alvita Nathaniel related the idea of no preconceptions well and succinctly. She writes,

Generating good codes also require that analyst to be for her coding a non citizen for the moment so she can come closer to letting the data speak for itself. And speak for itself further from the issue orientation implicit in the academic field’s view of the researcher’s data which view can dictate a preconceived biased view of the data that is hard to give up as it structures up the confusion. At first the researcher may feel that his non-preconceived field work and coding yelds only scattered uncodable observations. But as soon as he starts to comparatively analyze data – preferable as soon as possible with the beginning research …codes will emerge yielding theoretical leads. Then conceptual coding is off to a start without preconception. Description is left behind. A ‘new truth’ emerges. It is highly motivating.

Thus, as Alvita says, the initial suspension of preconceived ideas is soon replaced firmly by the joys of emerging discovery. To foster this transition to discovery the researcher should start the constant comparative coding with the initial interview or interviews that day or latest that night. The sooner discovery starts, the sooner preconceptions have less bother or claim on the research. Why drag it out with “waiting” schedules for typing tapes. Field notes speed up the generating process.  Evert Gummesson, a professor of marketing and management, firmly supports no preconceptions. He says:

Simply put, inductive GT research lets reality tell its story on its own terms and not on terms of received theory of academically accepted concepts. There is growing encouragement among customer centered companies with the prime goal of satisfying customer needs to stop coding with preconceived concepts and let the needs emerge conceptually.

This position of dropping preconception and taking on of open GT research applies to many fields of practice and service orientation such as nursing, medicine, education, management, social work, psychotherapy etc. Gummesson continues strongly:

Thus, code for what is there, not for what is preconceived to be there even though it appears not to support preconceived practices of marketing, bureaucracy, textbook or academic theory or the services of practicing professions and their short term practices or long term goals or facts for quick fixes. To start generating a theory or a research project for generating a theory by first designing preconceived clearcut categories and criteria for them will kill or mutilate chances for generating an emergent GT. As long as GT research is directed to an area of interest – we have seen it a multitude of times – patterns will emerge with the gentle assistance of the researcher using GT methodology. They will not be patterns brought on by forcing received concepts on data, nor on paying homage to the legacy of extant theory in any discipline. The GT researcher has to train himself momentarily to disregard or suspend existing knowledge while breathing in new real world data.

Gummesson’s influence in the world of marketing research has been fundamental and wide spread. Dr. Naomi Elliott writes me about her dissertation experience,

An important GT maxim is that the researcher enter the field with open questions to allow the participants own story to unfold without the direction of preconceived questions. Therefore, the guiding questions used throughout interviews focused on eliciting emergence what were the clinical practitioners; main concern and how they continually resolved it.

Anna Sandgren, PhD, wrote the following about her PhD research about “living on hope,” the same as Naomi:

Interviewing with open questions to allow the respondent’s answers to unfold without the direction from preconceived questions. Open conversations allow the respondent to keep talking about his main concern, which allows the yield of emergent latent patterns. Furthermore, coding becomes easier with open data than from data obtained from preconceived questions which are likely not to tap relevant latent patterns.

In short, open questions lead steadily to open coding for discovering the main concern and related categories. As Odis Simmons would say, “use grand tour questions and coding should start as soon after data collection as possible, which forestalls preconceptions on what the main concern and related categories may be…,” Coding should start the night of the first field note interview. Taped interviews and typing tapes take too long a delay to start the analysis, which foster preconceptions during the delay of systematic constant comparative coding.

Thus, the no preconception dictum applies also to the data gathering questions as well as to forcing preconceived concepts. Preconceived interview guides and questionnaires block emergence with pre-framed thought about the way it should be, not the discovery of what is going on. The researcher finds that emergent questions lead to emergent coded patterns as coding feeds on the emergent of codes from emergent interview questions. Coding feeds on itself when using emergent interviewed questions spawned by coding. Thus the researcher needs to start coding right away as he starts data collection. They go on simultaneously.

And also, since there is a great accumulation of GT dissertations at this time, the researcher can thumb through lots of GT articles, outside his/her area of interest to become what non preconceived codes look and sound like. This type of literature reading increases the researchers sensitivity to possible codes without forcing preconception concepts. It helps suspend professional problems and concepts.

“No preconceptions” is a dictum subject to growing in its procedural clarity of coding with constant comparisons for generating a GT. But adaption of no preconceptions to the multi version view of GT brings preconception back in many ways since the multi version do not use the constant comparative method to really discover emergent concepts and open questions. Existing concepts then are compared by incidents which simply test them or forces them. Then classic GT is remodeled to a QDA method of conceptual description.

The researcher can trust to the constant comparative method to discover what the participants view as the general problem and their specific problem. The researcher then starts to treat the emergent concerns as conceptually problematic. In the bargain, as I have said, preconceptions are very soon being replaced by the emergent concepts about what is really going on, and preconceptions are forgotten. Academic preexisting categories fade away in the wake of the grab of emergent categories with fit and relevance that emerge from interchangeable indicators. The data will produce categories that could never have been anticipated.

Preconceptions can surely rescue the initial confusion that comes with constant comparison of indicators and they can reduce the fear of never coming up with an emergent category. This confusion is quite real, but the researcher should be patient, as the constant comparative method will start revealing patterns to be named as concepts. Patterns are always there and will emerge, usually faster than expected, especially if the researcher starts with field notes and then coding the data immediately and then uses emergent questions from the coding to see if the codes work with relevance and fit. Theoretical sampling soon sets into questions of respondents, giving rise to questions about emerging patterns that could never have been preconceived.

Suspending preconceptions apply to field domain of perceptive theory bits, cherished concepts, types of data, etc and applies to the personal domain of pet conjecture, system perspectives’, cultural predictions and social biases, religious dogma, issue bias and affiliated preconceptions. Keep in mind that I am saying suspending preconceptions for the duration of the research goal of generating a substantive theory. One does not have to give up what one has learned and believed, though in many cases the substantive GT will change thinking with confirmation. Correction or abandonments of preconception grow with the generating of the substantive theory.

Remaining open to what is really going on will soon transform the researcher to going where the data takes him. This applies to most GT researchers except for the most intransigent ideological and, field driven thinkers with an immutable reality to push on others. Most researchers will when coding and analyzing go through the eureka effect of discovery and from then on suspending preconceptions becomes routine.

Here is another helpful thought from Dr. Odis Simmons, well-known GT teacher: “It is common during coding to generate concepts that relate to the researcher’s particular professional practice or to ubiquitous popular psychological concepts like self esteem, separation anxiety or identity. These codings are usually large inferential leaps and are based on one indicator and neglect a series of interchangeable indicators. They can burden the emerging theory extant conceptual baggage and imported connotations. This can easily diminish the unique value and contribution of the emerging GT.” Odis is correct, that no matter how one uses them preconceptions diminish the grab, relevance, fit and workability of a generated GT.

All researchers using their own or others data when doing GT must learn the skill of tolerance, with ambiguity and “not knowing” before emergence. Preconception clears up confusion quickly, but they must suspend the professional and or personal preconceptions to frame up the confusion quickly. They must deepen their analysis to reach the promise of emergence by constant comparison of indicators and in the bargain deepen their knowledge of GT methodology.

The toughest is suspending   especially those types of professional preconceptions, reinforced by professional training, collegial input, academic social structure requirements, the best peer review journals and dissertation committees. Taking them all on is not easy. Remember one does not throw out everything they have learned. The researcher just suspends it when using GT methodology, especially when coding and theoretically coding. Ideologically driven researchers usually have difficulty suspending preconceptions since they overlay what is going on with jargonized biases as to what they believe ought to be going on. The researcher doing constant comparisons of interchangeable indicators and remaining open to what is going on in the data and coding its abstract patterns reduces the “what ought to be” to “what is.” The literature and library are always there. They do not disappear. The correct literature can always be related to the final substantive GT to bring its contribution into the main stream of current thought within the appropriate field.

Joy of Coding without Preconceptions

          There is a joy of coding without preconceptions. It gives the researcher energy that goes with autonomy and openness and it speeds up the theoretical sampling for selective coding. Astrid Gynnild, PhD, wrote me about one of her students, “she is now doing much better since she is allowed to go into collecting data without reading all about her area of interest first. She is very energized by this autonomy and doing all the preexisting theory stuff first was what she feared the most.”

I cannot tell you how many PhD researchers call me with the same response to being given their autonomy. They say, “I am supposed to study this problem and I cannot find it.” I reply, “When using GT, forget what you are supposed to find and just see what you are finding.” My reply frees them to discuss with me what they are finding that is there. They are energized to the max.

I further warn them to be wary of a supervisor’s need to stop their beginning confusion with demands to frame up the analysis with extant concepts from their field. I tell them that they do not have to know the area of interest problem nor the main concern of the participants regarding the general problem. Constant comparative coding will soon reveal it for the researcher. When doing GT, interest in an area of interest does not require a preconceived problem to legitimate the subsequent. In fact, the GT researcher should be prepared to have his emerging theory radically opposing mainstream theoretical thinking. In this case he/she must remain faithful to the emergent coded patterns since they came from data.

To say the least preconceived questions, problems and codes all block emergent coding, hence block classic GT. Preconceived field research is often flat or boring for its lack of grounded fit or relevance. There is no grab. Starting a GT research without knowing the participant’s problems or concepts explaining their resolution is highly motivating, because the researcher starts the path to autonomous discovery. This is the path of knowing nothing about participants main concerns to knowing an in-depth theory explaining how they resolve their main concerns.

For example, Amy Calvin in her award winning dissertation starts with the reconceived problem of how patients on dialysis plan for death and give body parts in advance directions. The participants would not talk with her about this field imposed problem. She phoned me and asked what she should do. I told her to go back and ask general questions that allow them to vent (instill a spill). She soon learned that their main concern was staying alive by beating the odds. They would not discuss advanced directives. Staying alive was continually resolved by helping each other with equipment AND by appealing to a higher force through religion: God.

Judith Holton’s comment on this chapter is poignant here. She writes:

One of the things that your chapter brings home so clearly – and that has been forgotten or dismissed by qualitative researchers – is that the motivation for generating GT comes not simply from generating theory from data but from generating theory free of preconceived frameworks of any kind. So we still read in papers that it is a GT but the authors still use preconceived frameworks and concepts to guide the study.