In which type of qualitative research do the researchers intend to generate a theory that is based on the data systematically gathered and analyzed?

ABSTRACT

Surgical attire guidelines (SAGs) assist perioperative nurses with minimizing the risk of patients developing surgical site infections. However, some intraoperative team members fail to fully adhere to SAGs, which may put patients at risk. Because there is a lack of published literature on the reasons for intraoperative team members’ SAG nonadherence, I undertook a grounded theory study to explore the decision making of intraoperative team members related to the AORN SAG. The resulting model identifies the factors that affect decision making regarding SAG adherence, including personal identity needs, adherence anchoring activities, SAG awareness, direction from authority figures and organizations, guideline evidence strength, and resource availability. After weighing the influence of each factor, team members decide to what extent they will adhere to the SAG. The theory also provides guidance for enhancing adherence decisions.

Future Directions

Grounded theory offers educational researchers a method that complements varied forms of qualitative data collection and that will expedite their work. Adopting more grounded theory strategies will enable educational researchers to further the theoretical reach of their studies and to make tacit meanings and processes explicit. Constructivists have not only re-envisioned grounded theory, but also revised it in ways that make the method more flexible and widely adoptable than its earlier versions.

In the past, grounded theory has often been viewed as separate from other methods. Now, the constructivist version makes the usefulness of combining grounded theory with other approaches more apparent, as is evident in grounded theory studies in education. Grounded theory can make ethnography more analytic, interview research more in-depth, and content analysis more focused. Several computer-assisted qualitative data analysis programs are built on grounded theory, and this method can add innovation to mixed methods research. Grounded theory emphasizes focusing data collection and checking and developing analytic ideas. Hence, grounded theory offers the tools for building strong evidence within the analysis and for explicating processes. Consequently, grounded theorists in education have a bright future for making powerful arguments in areas such as curricular studies, educational leadership, and educational policy.

Grounded theory is introduced as an inductive, comparative methodology that provides systematic guidelines for gathering, synthesizing, analyzing, and conceptualizing qualitative data for the purpose of theory construction. The founders of grounded theory, Barney G. Glaser and Anselm L. Strauss, offered the first explicit, codified statement of how to analyze qualitative data. The intellectual traditions of each of its founders are discussed. Strauss brought Chicago School pragmatism, symbolic interactionism, and field research to grounded theory and Glaser's training in survey research gave the method its systematic approach, positivist proclivities, and procedural language. Debates between Glaser and Strauss and Corbin are noted and the distinction between objectivist and constructivist grounded theory is introduced. Objectivist grounded theory assumes the reality of an external world, takes for granted a neutral observer, views categories as derived from data, and sees representation of data and subjects as non-problematic. Constructivist grounded theory places priority on the studied phenomenon over the methods of studying it, uses grounded theory strategies as tools, not as prescriptions, and acknowledges the researcher's role in interpreting data and creating categories. Techniques involved in three major grounded theory strategies are described: coding, memomaking, and theoretical sampling. Last, a statement of current emphases and future directions is provided.

Grounded Theory

Grounded theory is a method in naturalistic research that is used primarily to generate theory.13 The researcher begins with a broad query in a particular topic area and then collects relevant information about the topic. As the action processes of data collection continue, each piece of information is reviewed, compared, and contrasted with other information. From this constant comparison process, commonalities and dissimilarities among categories of information become clear, and ultimately a theory that explains observations is inductively developed. Thus, queries that will be answered through grounded theory do not relate to specific domains but rather to the structure of how the researcher wants to organize the findings (Box 8-4).

As you can see, each query indicates that the research aim is to reveal theoretical principles about the phenomenon under study. Grounded theory can also be used to modify existing theory or to expand on or uncover differences from what is already known. In the two queries in Box 8-5, grounded theory is structured to address current theory from a new and inductive perspective.

Grounded Theory

Grounded theory (GT) is probably the most widely known methodological perspective on how to conduct qualitative research in the social sciences. Originally introduced by sociologists Barney Glaser and Anselm Strauss (Glaser and Strauss, 1967), GT is used extensively in education and related fields.

GT comprises a distinctive methodology, a particular view of scientific method, and a set of specific procedures for analyzing qualitative data and constructing theories from those data. The methodology provides a justification for regarding qualitative research as a legitimate – indeed, rigorous – form of inquiry. The view of scientific method adopted by GT is generally taken to be inductive in nature, although this is a contested matter. GT researchers gather non-numeric data from a variety of sources, including interviews and field observations. Once gathered, the data are analyzed using coding and theoretical sampling procedures. A set of interpretative procedure are then used to assist in the construction of theory that emerges from, and is grounded in, the data.

In efforts to identify empirical social phenomena, and construct theories that are constrained by those phenomena, almost all accounts of GT adopt the three major strategies of data coding, memo writing, and theoretical sampling.

In GT, data gathering and data analysis are interactive. From the time data collection begins, grounded theorists engage in data analysis, which leads to further data collection, subsequent data analysis, and so on.

The first data analytic phase of GT begins with the coding of the data. This is undertaken to conceptualize the data by discovering categories into which they fit. The coding process has three phases: open coding, axial coding, and selective coding. In open coding, researchers describe the data by looking at it line-by-line. This strategy of focusing on small units of data, and their interpretation, encourages the development of a theoretical sensitivity to new ideas with regard to the data, and helps prevent the forcing of data into existing categories. Strauss (1987) maintains that when a full array of categories have been identified, one should undertake axial coding – whereby one puts the data back together again in new ways by making connections between the numerous categories. Following that, a selective coding step is implemented in which the researcher looks to systematically identify those categories that relate closely to the core category. The core category lies at the heart of the emerging theory and is central to its integration.

Although memo writing can occur at any stage of the research process, it frequently takes place between the coding of data and the writing of the initial draft of the research report. Memos are written to identify, develop, and keep track of theoretical ideas. Where relevant, they are recorded, recalled, and reworked to produce new theoretical memos. Memo writing becomes more systematic, focused, and intense as theory of greater density and coherence is produced.

Memos written with regard to data codes and theoretical ideas enable the researcher to identify gaps that require the collection of further data. For this, theoretical sampling is undertaken. With theoretical sampling – in contrast with traditional representative sampling – decisions concerning what data to collect, code, analyze, and interpret are directed by the emerging GT. Theoretically relevant events, activities, and populations are all sampled, and the comparisons between these are aimed at increasing the conceptual density and integration of the emerging theory. Thinking effectively with regard to data in theoretical terms requires an adequate degree of theoretical sensitivity. When the additional gathering and analysis of data no longer contribute to the understanding of a concept or category, a point of theoretical saturation is reached. At this point, one stops collecting data in respect of a category and moves to consider another category or concept.

Consistent with the pragmatist influences on GT methodology, Strauss (1987) characterizes scientific method as a sequence of induction, deduction, and induction: grounded theories emerge inductively from the data, test predictions are then deduced from the theories, and, finally, the theories are inductively confirmed or disconfirmed.

Despite the considerable attention given to the exposition of data analysis in GT, it is difficult to fathom just how – and in what sense – GT is said to inductively emerge from, and be grounded in, the data. This is because the nature of the inductive reasoning involved is not described.

Glaser and Strauss have been criticized on the grounds that they advocate a return to a simple and unacceptable Baconian inductivism. On this interpretation, GT is depicted as a tabula rasa view of inquiry which maintains that observations are not theory or concept dependent. However, this is not Glaser and Strauss’s position. In The Discovery of Grounded Theory (Glaser and Strauss, 1967), they explicitly disavow this view of inquiry – noting that the researcher requires a theoretical perspective in order to see and abstract from data. It is in the interest of obtaining emergent, diverse categories at different levels of abstraction that Glaser and Strauss would have the researcher hold all potentially relevant facts and theories in the background for some time. Clearly, this is a form of bracketing, not a tabula rasa conception of inquiry.

Although it is clear that Glaser and Strauss are not naive inductivists, the actual nature of the inductive relation that – for them – grounds emergent theories in their data is difficult to fathom. For Glaser and Strauss, GT is said to emerge inductively from its data source in accordance with the method of constant comparison. As a method of discovery, the constant comparative method is an amalgam of systematic coding, data analysis, and theoretical sampling procedures which enables the researcher to make interpretive sense of much of the diverse patterning in the data by developing theoretical ideas at a higher level of abstraction than the initial data descriptions. However, the notion of constant comparison is of little help in figuring out whether the inductive inference in question is enumerative, eliminative, or of some other form.

Given the pragmatist influence on GT methodology, it is not surprising that Strauss (1987) mentions the notion of abduction in his brief discussion of induction. Unfortunately, however, he refrains from including it in his discussion of the inductive generation of theory. In his mature writing, the American pragmatist Charles Sanders Peirce clearly distinguished between these two forms of inference. Both inductive and abductive arguments are ampliative, or content-increasing – in that their conclusions contain more information than is contained in their premises. However, the type of ampliation is different for each. Inductive arguments are descriptive in character because they reach conclusions concerning the same type of manifest attributes mentioned in their premises. By contrast, abductive arguments reason from factual premises to explanatory conclusions, as when we reason from presumed effects to underlying causes.

A growing number of authors have characterized the creative inference involved in the generation of GT as abductive in nature (e.g., Haig, 1996; Reichertz, 2007) – that is, rather than viewing a GT as an inductive abstraction from data analysis, it is thought of as the result of explanatory inference to factors that transcend the data in a more fundamental way. On this view, the data analytic dimension of GT can reasonably be construed as inductive in nature. However, in order to explain the abstracted data patterns, the construction of GT needs to be though of as abductive in nature.

Haig has gone further and suggested that the entire process of theory construction in GT can be cast in an abductive light (Haig, 1996, 2005b). On his account, the abductive nature of GT extends beyond theory generation to include theory development and theory appraisal. A strategy of analogical modeling is used to develop GT. Because analogical modeling increases the content of explanatory theories, the reasoning it embodies is referred to as analogical abduction. This reconstruction of GT adopts inference to the best explanation as the preferred approach to the evaluation of mature theories. Specifically, TEC – outlined earlier – is adopted, and the better of competing GTs is judged to be the one that is more explanatorily coherent.

3.06.2.1 Questions

Grounded theory is designed to be usable with a very wide range of research questions and in the context of a variety of metatheoretical approaches. Rather like the statistical analyses that psychologists are more familiar with, it deals with patterns and relationships. However, these are not relationships between numbers but between ideas or categories of things, and the relationships can take a range of different forms. In some respects the procedures in grounded theory are like the operation of a sophisticated filing system where entries are cross-referenced and categorized in a range of different ways. Indeed, this is one qualitative approach that can be effectively helped by the use of computer packages such as NUDIST, which was itself developed to address grounded theory notions.

Grounded theory has proved particularly appropriate for studying people's understandings of the world and how these are related to their social context. For example, Turner (1994; Turner & Pidgeon, 1997) has used grounded theory to attempt to explain the origins of manmade disasters like fires and industrial accidents; Charmaz (1991) has studied the various facets that make up people's experience of chronic illness; Clegg, Standen, and Jones (1996) focused on the staff members' understanding of their relationship with adults with profound learning disabilities. In each case, a major concern was to incorporate the perspectives of the actors as they construct their particular social worlds. Grounded theory methods can help explicate the relation of actions to settings (how does the behavior of key personnel in the evolution of a major fire follow from their individual understanding of events and physical positioning?); it can be used for developing typologies of relevant phenomena (in what different ways do sufferers of chronic illness conceptualize their problem?); and it can help identify patterns in complex systems (how does the information flowing between social actors help explain the development of a laboratory smallpox outbreak?).

Like most of the qualitative approaches discussed here, grounded theory is not well suited to the kinds of hypothesis testing and outcome evaluation that have traditionally been grist to the mill of clinical psychology, because of its open-ended and inductive nature. Although the researcher is likely to come to a topic with a range of more or less explicit ideas, questions, and theories, it is not necessary for any or all of these to be formally stated before research gets under way. The approach can start with a specific problem or it may be more directed at making sense of an experience or setting.

Grounded theory can be applied to a range of different textual materials such as documents, interview transcripts and records of interaction, and this makes it particularly suitable for certain kinds of questions. It can deal with records which exist prior to the research and it can deal with materials specifically collected. The processes of coding allow quite large amounts of material to be dealt with. For example, while Turner studied a single (lengthy) official report of a major fire in a holiday complex, Charmaz studied 180 interviews with 90 different people with chronic illnesses. The requirement is only that the material can be coded.

3.06.3.1 Grounded Theory Approach

Grounded theory is a data-driven, bottom-up approach through which theory emerges directly from the data. In line with a constructivist stance, theory is said to be “grounded” in the data because it is the data (e.g., participants' narratives) that drives the analytic process and emerging theories and frameworks, not the other way around as in positivist approaches (Glaser and Strauss, 1967; Glaser, 1992; Charmaz, 2006). Grounded theory is a particularly appropriate approach when studying understudied populations or phenomena for which there are no pre-existing empirical theories. A research investigation using a grounded theory approach begins with the generation of sensitizing concepts consisting of general interests, thoughts, and ideas about a topic, not formal theories or preexisting assumptions (Glaser and Strauss, 1967; Glaser, 1992; Charmaz, 2006).

Once data has been collected (e.g., interviews, focus groups) transcripts are coded using a systematic procedure, starting with open coding, which consists in a line-by-line summary of several transcripts (typically 1 to 3). Also starting during open coding and continuing throughout the data analysis process, memoing consists in writing memos about preliminary analyses, ideas, and thoughts. Once transcripts have been coded line-by-line, the resulting open codes are used to proceed to thematic coding, also referred to as axial coding, which consists in collapsing open codes into thematic categories or concepts (e.g., depressive symptoms, trauma), referred to as thematic codes. All transcripts are then coded using the thematic codes that were generated through thematic coding. Last, theoretical coding consists in arranging and rearranging thematic codes and memos, while examining overlapping content and conceptualizing relations between intersecting thematic codes and memos (Glaser and Strauss, 1967; Glaser, 1992; Charmaz, 2006; Rasmussen et al., 2016).

Findings of data analyzed using a grounded theory approach are typically presented in a report organized around the various interconnected themes that emerged through thematic and theoretical coding. Consistent with the idea that emerging theories and framework are grounded in the data, reports commonly include a significant amount of raw data, that is, direct quotes from participants' narratives. Also characteristic of grounded theory is theoretical sampling, a process through which the researcher simultaneously collects, codes, and analyses the data in order to decide what data to collect next (e.g., additional participants' narratives). Because grounded theory is based on the premise that both the analytical process and new theory emerge directly from the data, data collection must be flexible and may continue until the researcher has gathered enough data to form an emerging theory. Relatedly, precise criteria for both inclusion of participants and data collection cannot be planned in advance at the outset of the study, as these criteria emerge as the theory and data analysis process evolve (Glaser, 1978, 1992; Glaser and Strauss, 1967). Last, theoretical sampling allows the researcher to generate insights and conceptualizations by drawing on comparisons among samples of data (e.g., participants' narratives), and to then gather more data if more comparisons are warranted (Van den Hoonaard, 2008). Theoretical sampling usually ends when data collection has reached saturation (described in an earlier section).

Rasmussen et al. (2013) provide an illustrative example of how grounded theory can be used to explore novel or understudied topics. Using data collected through 18 focus groups and 8 individual interviews with a multiethnic group of West African immigrant adolescents and adults (n = 59), the researchers explored West African immigrants' resolution of family conflicts. Interest in this topic (i.e., sensitizing concepts) grew from the researchers' clinical work with asylum seekers and, in the case of one researcher, being from West Africa herself. Consistent with a theoretical sampling approach, data analysis began immediately after the first focus groups and continued throughout the data collection process. Theoretical coding led to an in-depth analysis of intersecting thematic codes such as “disciplining and monitoring,” “interpersonal conflict,” and “gender roles.” Results were organized using a social ecological approach (Bronfenbrenner, 1979) and indicating that, among West African immigrants, strategies to resolve family conflicts involve four levels of problem-solving resources: individual/dyadic, extended family, ethnocultural community leadership, and state authorities. Providing guidelines for those working with West African immigrants, Rasmussen et al. (2013) note that strategies involving extended family and ethnocultural community leadership are rarely emphasized in mainstream Clinical Psychology research and practice.

Kinds of Qualitative Analysis

An exhaustive review of all qualitative methodologies being used in anthropology is beyond the scope here, but the most common qualitative methodologies are grounded theory, ethnography, phenomenology, narrative inquiry, and case study. Each qualitative methodology, with its attendant analysis, focuses on the qualities of experience, although in different ways (for example, descriptive analysis, categorical analysis, thematic analysis, or narrative analysis).

1.06.6.2.5 Grounded Theory

Though grounded theory is also seen as a form of qualitative analysis, the fact that these studies are typically designed to fit this format at the outset qualifies them as a distinct type of qualitative study. Grounded theory studies, as pioneered by Glaser and Strauss (2000), aim to construct a formal theory for a specific phenomenon, maintaining a focus on the data and the information provided directly by participants. It is widely considered to be one of the most intensive and rigorously scientific qualitative approaches. The whole process, including but not limited to participant sampling, is informed by theory. The analytic procedure is iterative, meaning the researchers continually extrapolate findings from the data and then return to the data source in order to seek further information and support (or lack thereof) for their growing theory. In grounded theory studies, this process is typically referred to as open and axial coding. If new information from the data does not fit into the evolving theory, the theory is adjusted. In this way, the final theory has evolved from, is “grounded” in the data. Methods like this can be used to form theories about how recovery processes may look for particular mental health conditions or how clinical practitioners develop their therapeutic theoretical orientations.

Peripheral

Early theories grounded in the specificity theory of pain relied on peripheral mechanisms to explain phantom limb pain. Phantom limb pain is more frequent in patients with long-term stump pain.12,31 Neuromas at peripheral nerve endings have been implicated in the development of phantom pain. This was initially inferred from the alteration of phantom limb pain following stimulation of the presumed neuroma within the stump. Nystrom and colleagues demonstrated the increase in C fiber activity associated with tapping of transected nerves in amputees and correlated that with an increase in pain sensation.27 Sodium channel activation in amputees' stumps has been shown to increase pain sensation.32 Lidocaine, a sodium channel blocker, in contrast blocks phantom pain.32 Despite the clear association of neuroma to phantom limb pain, surgical removal of neuromas has not shown much promise in limiting phantom limb pain in amputees.33