Getting Started

Barney G. Glaser

Editor’s note:  Especially helpful for those thinking of beginning new research projects, this paper addresses common questions about the particular way in which grounded theorists identify a research problem and craft a research question appropriate for classic grounded theory.  Getting Started was first published by Sociology Press in Glaser’s Emergence vs Forcing: Basics of Grounded Theory Analysis (1992).  This important chapter has been excerpted and lightly edited for clarity and context.

It may sometimes be said that one of the most difficult parts of doing research is to get started.  The making of choices and commitments to a research problem seem less secured and structured when doing descriptive research in quantitative or qualitative research.  This occurs because the research problem is chosen beforehand and therefore forces the data, thus the yield may be small or nothing since the problem, in fact, may not be relevant.  A “thought up” problem may sound juicy, but the preconception leads nowhere.

The underlying principal in grounded theory, which leads to a researchable problem with high yield and relevance, is that the research problem and its delimitation are discovered or emergent as the open coding begins on the first interviews and observations.  They soon become quite clear and structured as coding, collection, and analyzing begin and a core variable emerges, and saturation starts to occur.  In short, getting started in grounded theory research and analysis is as much a part of the methodological process as are the ensuing phases of the research.

The researcher should not worry.  The problem will emerge as well as the manner by which the subjects involved continually process it.  As a matter of fact, it emerges too fast most of the time and the researcher must restrain herself until sure if it is core and will account for most of the variation in the action in the substantive area under study.  As categories emerge in open coding, they all sound like juicy problems to research, but all are not core relevant.  Only one or at most two.  Remember and trust that the research problem is as much discovered as the process that continues to resolve it, and indeed the resolving process usually indicates the problem.  They are integrated.

Area vs Problem

There is a significant need to clarify the distinction between being interested in an area compared to a problem.  A researcher can have a sociological interest which yields a research problem and then look for a substantive area of population with which to study it.  But this is not [classic] grounded theory.  It is a preconceived forcing of the data.  It is okay and can produce good sociological description, but it usually misses what subjects in the substantive area under study consider, in their perspective, the true problems they face.  This kind of forcing with the support of advisor and colleagues can often derail the researcher forever from being sensitive to the grounded problems of the area and their resolutions.  A missed problem is a problem whether the researcher discovers and attends to it or not.  It does not go away.  We find, as grounded theorists, so often in preconceived research that the main problem stares us in the face as the researcher just attends elsewhere and misses it completely in an effort to describe what is going on.  Squelching it from focus does not remove its relevance.

In vital contrast, the grounded theory researcher whether in qualitative or quantitative data, moves into an area of interest with no [pre-identified] problem.  He moves in with the abstract wonderment of what is going on that is an issue and how it is handled.  Or, what is the core process that continually resolves the main concern of the subjects.  He discovers that truth is stranger than fiction.  If he moves into an area with an interest in studying people in pain, he will discover what problem pain produces and how it is resolved or processed.  The social structure of each substantive area can make this resolution quite different.  The grounded theorist keeps his mind open to the true problems in the area.  A forcing researcher may study risk taking in steeplejack work; a grounded theorist will probably discover that the main problem is negotiating the day’s voyeurism, with the risks involved as a minor consideration.

As mentioned in Theoretical Sensitivity (1978), it is most advisable to the grounded theorist, when at all possible, to choose an area with a life cycle interest to gain enough motivation to get her through the research to the end product.  But, even when a researcher has to study an area of lesser interest, it is likely that the conceptualization of it will still be of interest as a general sociological concern and process.  Thus, if one has [grant] money to study meatpacking, he may be able to study on an abstract level the style of eating patterns in diverse social classes.

Areas of interest are not hard to come by.  They abound, and with grounded theory, the research problem emerges easily.  Whereas a preconceived problem is hard to come by with the surety it will both yield findings and will be supported by enough data.  When a research problem is elusive or hard to come by, a lot of people tend to give advice.  However, the grounded theorist should be wary, since his approach to the research problem is both grounded and easier.  The researcher’s search for the preconceived problem is subject to the whims and wisdoms of advisors with much experience and of colleagues.  He should be careful as he may just end up studying his advisor’s pet problem with no yield for him and data for the advisor.  And, he will likely miss the relevance in the data.

Preconception using the technical literature can have a level of groundedness in it, especially at the end of a piece where the author “appeals to future research.”  This is, of course, a good lead and the grounded theorist should consider these issues but be careful that they are born out in his own emergence of problems in the area, as later date conditions may have changed relevancies.  Thus, for [some] women, women’s liberation may not be an issue when they would rather [quit work and stay at home while others] wish to get away from [home].  Personal experience and/or professional experience associated with it can produce strong, life cycle, substantive area interest.  But the grounded theorist must be careful not to force data with his or her own problem and keep an open mind to the emergence of the subjects’ problem.  The researcher’s personal problem may be idiosyncratic, but once the general concern emerges, it is almost sure to integrate as a varying property of it.  The life cycle interest will be taken care of and be enhanced with understanding coming from the emerging theory.

The Research Question

The need to preconceive is strong when there is no trust in discovery of a problem.  The researcher should fight this and learn not to know, when telling himself or others what he is studying.  Do not say anything until the core problem has emerged and proves to be a stable focus of the research.

In comparison to preconceived description, there is no dilemma when choosing the grounded theory methodology as to when the problem may become known, whether with quantitative or qualitative data.  There is no need to waste time on the debate as to whether or not the research question should dictate the method or the method the research question. The researcher need not be concerned whether or not the data should be collected quantitatively or qualitatively or in what combination, as required when studying the preconceived problem.

Once grounded theory methodology is chosen, this debate is moot.  The methodology process [uncovers] the emergent problem and all data of whatever type is grist for the mill of constant comparison….  The emergent research problem will core out and be delimited by diverse conditions such as the researcher’s training, the locale of subjects, funding, etc.  Boundaries to the problem will emerge and the one criteria of grounded theory, modifiability, says that is a good grounded theory should be readily modifiable to new conditions, new subjects, and perspectives on the same problem, provided that the same problem is relevant in the new area.

Remember that grounded theory research is the study of abstract problems and their processes, not units.  Unit analysis is for description. Thus, studying women managing pregnancy is not to focus on women, but to discover their emergent problems and their resolutions for managing the pregnancies. These problems will likely vary considerably with studies in different areas.  The problems of middle-class women trying to communicate with [physicians] who do not favor natural childbirth are far different than those of lesbians trying to communicate with macho, male [physicians].

The Specific Research Question

To repeat, the research question in a grounded theory study is not a statement that identifies the phenomenon to be studied.  The problem emerges and questions regarding the problem emerge by which to guide theoretical sampling.  Out of open coding, collection by theoretical sampling and analyzing by constant comparison emerge a focus for the research.

Even then, when specific questions can be asked without forcing the data or its collection, the researcher never, never asks the question directly in interviews as this would preconceive the emergence of data.  Interview questions have to relate directly to what the interview is about empirically, so the researcher maximizes the acquisition of non-forced data.  These specific questions are in the thoughts and the analysis of the researcher, to be reviewed later.  Think theory, talk everyday common [language].  And this method of qualitative analysis is the same for qualitative as for quantitative collection of data.  In grounded theory there is no preconception of being too broad or global or narrow at whatever stage; the grounded theory process steers the path to bounded focus.  And with grounded theory there is also no preconceived relevance as to whether questions to subjects are interactional, organizational, biographical, psychological, or whatever.  The emerging questions simply tap the variables that work, whatever the field.  Obviously, a researcher is trained in the sophisticated use of one or the other variable and will be more theoretically sensitive in his own area.  If a major variable occurs in an area outside his training, he may have to call in a consultant.  At minimum, he will have to report this grounded fact in his theory, not ignore it as having no relevance.  Thus, a sociologist may have to consult with an economist or psychologist or political scientist at times to better understand processes in their fields.

In sum, when a researcher flounders in getting started on a research project, it is quite often the result of forcing on the data a preconceived problem that ought to take the data apart and give yield, but does not, because of lack of relevancy.  The researcher is lost and sees the data as recalcitrant.  The grounded theory researcher bypasses this problem in getting started by simply studying what is to be studied with no preconception of what should be in advance of its emergence.  He has the patience and security and trust to wait for its emergence.  Also, he trusts himself not to know in advance and forces himself not to pontificate that he knows better than the subjects involved what is most relevant to them.


Glaser, B. G. (1978). Theoretical sensitivity: Advances in the methodology of grounded theory. Mill Valley, CA: Sociology Press.

Glaser, B. G. (1992). Emergence vs forcing: Basics of grounded theory analysis. Mill Valley, CA: Sociology Press.