Janice M. Morse

                            12

QUALITATIVE STRATEGIES FOR CONCEPT DEVELOPMENT

                Concepts shape how we think about the patients, families, and communities with whom we work. They direct our observations and our actions based on those observations. Important in their own right …, they merit our careful attention and nurturance.

—Knafl and Deatrick (2000, p. 365)

Qualitative inquiry does not take place in a vacuum, with the investigator blindly using inductive strategies and hoping that a concept or theory will eventually emerge from the data. Neither do qualitative findings emerge from a base of ignorance. Social scientists usually have a fair idea of “what is going on” in a context or with participant behavior, and that is why they ask certain questions about a topic area when developing a proposal. These researchers have developed research programs and conducted previous studies in their program of research; they are familiar with the literature, the state of the art, and in which direction the work of other investigators is moving. Researchers never start from scratch, with a “blank slate.” Indeed, to obtain funding for their research, they must review their previous work, situate it in the publications of others, and make a case for the significance of the proposed study within this context.

Some qualitative researchers argue that it is important to go into a setting naive, without preconceptions, to prevent bias—a view now largely discarded (Hegelund, 2005). But there are different kinds of bias, and all of these should be considered. Value-laden bias is problematic in qualitative inquiry: One does not do qualitative research to prove a point, in particular, a sensitive point, such as the worth (or lack of worth) of some behavior, opinion, and so forth. And, in this light, I always advise students that doing qualitative research is not a way to resolve their own problems. You cannot be both immersed in grief for the death of your father, or upset about the injustices of your divorce, and at the same time explore the responses of participants to the death of their parents or their divorces. Having experienced the experience you are studying increases the pain, gives you reason to continue with your own suffering or indignation or whatever—but it does not necessarily add insight. In fact, your own experience is a bias that may overwhelm your data and the experiences of your participants, introducing a source of invalidity.

On the other hand, bias, for the purpose of maximizing the behavior of interest, is essential for good qualitative inquiry, and it is this bias that undergirds the principles of qualitative sampling. If we are interested in pain behaviors, we must sample from a context in which we would see participants responding to maximal pain—not slight pain or average pain; if we are interested in uncertainty, then we sample from a setting in which uncertainty is going to be evident and frequent—otherwise we will be a long time collecting data, waiting for instances of uncertainty to occur so it can be observed, or for participants to think of uncertain situations to tell us about in an interview.

TARGETING CONCEPTS USING QUALITATIVE INQUIRY

When asking questions, qualitative researchers usually are interested, and frame their questions about a lay concept. Look at any issue of the qualitative journal on your desk and read the titles. The issue of Qualitative Health Research (December 2015) on my desk has the following concepts addressed in the titles:

In short, we cannot do qualitative inquiry without some level of understanding of concepts (what they are) and concept analysis (how to recognize them and their components) or concept development (how to build them). The trick is to know what type of concept you are working with, and its level of development, what will be the most appropriate methods and strategies to delineate and to develop the concept. Because concepts do not exist in a vacuum, you must also consider how to link the concept you are working with other established concepts. From there you can move to theory development.

To begin, you must be able to recognize if the concept is a lay concept or a scientific concept—or something in between. If the concept is a scientific concept, it has been developed, defined and the attributes assigned in the process of development. There is little use in using concept analysis techniques or qualitative inquiry to try to determine the meaning of a scientific concept. This information—the definition, the attributes, the boundaries, and so forth—was developed by the originator of the concept when the concept was first introduced. Usually there is no confusion about the meaning of such concepts: think of such scientific concepts as health promotion, public health, nursing, health, and so forth.

But remember that I said something about concepts that were “in between” lay and scientific—you can use the analytic strategies or techniques presented in this chapter or the structured techniques presented in Chapter 15, if you have some idea that concept drift has occurred, and if the concept is no longer being used as originally intended or as defined. Alternatively, you may use analytic techniques presented in this chapter if you find published descriptive studies of lay concepts that are partially developed. But once more: These qualitative techniques for concept analysis are useful only for lay concepts, and are not helpful for scientific concepts—concepts that have been initially defined and operationalized when they were created.

In this chapter, I examine the ways in which qualitative inquiry contributes to the development of established lay concepts, and qualitative strategies—analytic techniques—that may be used in defining the attributes or delineating the boundaries of developing data-derived concepts.

BUILDING LAY CONCEPTS EPISTEMOLOGICALLY

Lay concepts are not entities floating without referents. Because of their abstractness and lack of direct connection to reality, they require rich descriptions and theory to “locate” them. If this is not available for the concept, the concept is considered immature, and the first task is to create rich descriptions of the concept in its context. Nevertheless, inspecting your concept as described in Chapter 11, and by searching for a reasonable amount of literature, you will be able to make a decision about how the concept that you are focused on meets the adequacy criteria and how you should proceed with developing the concept. If the concept definition lacks clarity in the articles you have pulled, if you have to hunt for the definition, or use implied meanings, then the concept is probably immature. On the other hand, if the concept definition is clear, consistent, and consensual, it is probably mature. I use the word “probably,” because the decision for the level of concept maturity is not made on the basis of one criterion, but by assessing all four criteria discussed in Chapter 11, and again summarized in Table 12.1.

Next, check the presence and quality of the attributes and ensure they are clearly described. If they are easily located and appear adequate, the concept is mature; if they are not identified, the concept is listed as immature.

Again, this is on a continuum, and one may locate some attributes (but not all), ranking the concepts as “partially mature” or emerging.

For a mature concept, the antecedents and outcomes must be clearly identified, fully described, and demonstrated. If they are not listed and described, then the concept is considered to be immature. Finally, the boundaries should be delineated, and it should be clear to the reader what cases are examples of the concepts, which are not, and why. Remember that qualitative methods, as shown in Figure 12.1, are used for identifying concepts or for developing immature lay concepts.

QUALITATIVE METHODS FOR DEVELOPING CONCEPTS

Qualitative methods provide researchers with different methods for describing and interpreting reality. These methods are designed to best answer particular research questions, for identifying concepts, developing the anatomy and the physiology of concepts, and in theory building. Each method offers a different perspective, and the methods should be chosen according to what the researcher wants to accomplish. In this light, an excellent methodologist is one with a large repertoire of available methods and research strategies, so that what he or she wants or needs to know from the data, or what he or she wants to accomplish as a research goal, is possible given the methods and strategies accessible to him or her.

The various methods that may be available for concept development in qualitative inquiry are listed on Table 12.2. At a broad level, some methods are more appropriate than others for certain tasks. For instance, phenomenology is appropriate for concept identification, and I will demonstrate this in Chapter 13. However, knowledge of the structure of concepts and of other research strategies is necessary if the researcher desires to continue with the development of the concept.

Ethnography and grounded theory are well suited for the development of concepts and subsequently, for the development of theory (see Table 12.2). In nursing, qualitative researchers are mainly using methods that develop concepts and theories from unstructured interviews, in particular, using ethnography and grounded theory. Less commonly, there are structured techniques for concept development, primarily from ethnoscience (a method of cognitive ethnography), which has many strategies for the delineation of attributes and identification of conceptual boundaries. These structured methods are denoted by an asterisk in Table 12.2, and will be described in Chapter 15.

Note that the strategies for concept development do not include the entire research method, but rather are techniques to clarify concepts or themes in the process of doing a grounded theory or ethnography. This is because these methods embody creating concepts as a part of the process when using the complete method. In fact, becoming adept in qualitative concept development will greatly assist you in the most difficult parts of qualitative analysis—theory development—when you are doing qualitative research in general.

QUALITATIVE STRATEGIES FOR IDENTIFYING CONCEPTS

This section consists of a compendium of qualitative strategies for delineating data to support concept development or analysis using qualitative data. As mentioned earlier, these strategies are a part of major qualitative methods, extracted from the methods to facilitate concept development, when you do not need to conduct an entire study.

Nevertheless, the principles of qualitative inquiry still hold.² If you are using qualitative inquiry for identifying a concept or developing an immature concept, data adequacy and appropriateness are essential. You must attend to your sampling techniques for saturation, scoping, and must attend to obtaining variation in data. Principles of induction and abduction remain important.

Unstructured (Narrative) Interviews

Although the form of interviews may vary according to method, qualitative interviews share the common characteristic of inductively allowing participants the freedom to tell their stories without interruption. Participants talk about the general topic of the research project, each telling their story as they wish, and taking as long as they wish.

Rich Description

Interview data or unstructured interviews must be in the form of narrative description, preferably without prompts, guidance, corrections, or queries by the interviewer. The participants must be permitted to tell their story their own way, with the interviewer listening intently. These interviews, when transcribed, produce rich description—descriptions that are in-depth, detailed, and use words arising from the interviewee’s own experiences. The researcher can check that the interview is indeed unadulterated by the interviewer’s interruptions—simply look at the transcribed interview. One should find that the text appears in large blocks on the page, without showing lines of interjection from the interviewer. In fact, Corbin and Morse (2003) note that when the participant is immersed in interview, the interviewer no longer exists for the interviewee. The interviewee becomes totally immersed in his or her own story including the time and circumstances in which it happened. The emotions experienced at that time are reflected in their present emotion as emotional reenactment (Morse, 2002).

Qualitative Strategies That Facilitate Concept Development

Because the researcher is learning about the topic incrementally (as the study proceeds), the inductive approach to data collection is most unstructured during the first interviews, but it may become more targeted as the researcher begins to understand the area, to focus the study, and to strategically (or theoretically) sample for particular information.

Interpretative Coding

The nature of this particular information may come from the initial coding of categories (and sometimes themes). The category is given an emic label, that is, a label derived from the interviews themselves, and it may be a slang term, an already accepted and used concept label, or even simply a descriptive phrase used by some of the participants.

Building categories requires that interviewing continue until the category of interest has adequate data—a lot of data. There must be enough data to provide the researcher with many instances of all aspects of the category. Use techniques of content analysis to select the relevant blocks of text according to topic, and sort them by placing them into categories, which are initially broadly titled. Warning: Do not have many categories—at this time, up to 12 categories is ample for a data set. If you have too many categories, sorting becomes too difficult and you find yourself splitting hairs trying to decide into which category a piece of text belongs. Your analysis is slowed, and you tend to forget your category labels.

This art, the researcher’s skill of analyzing and developing concepts, lies in the ability to sensitively and critically appraise data and ask analytic questions of these data. Using your library knowledge of the concept, compare the library descriptions of the concepts with those from your data. Look at your data for what it is, look at it for what it implies and implicates, and look at it for what it represents. Ask analytic questions—the why, what, and how—questions that will enable variation in responses across author perspectives, between studies, and within different contexts and topics.

Interview techniques assist in processes of sifting and sorting; comparing and contrasting. Look for things that are the same, as well as those that are different. Look for differences in use, differences in form, differences in function, and similarities in use, form, and function. And many things that you need to see are hidden behind the lines of the transcript, and behind the text, concealed in metaphor, silence, contradictions, and riddles. The trick is to ask analytic question to obtain the answers, often by inference rather than expecting the answers to be directly in the body of the text. Working in a seminar is helpful. We look for what is there (is said), as well as what is not there (is not said). We look for reported behaviors that signify the presence of the concept. We look for metaphors and other linguistic indicators.

EXAMPLE OF INTERPRETATIVE CODING: GETTING BEHIND THE DATA

Janice M. Morse, Kim Martz, and Terrie Vann-Ward

I am sitting with dissertation students, Kim and Terrie in a dissertation seminar. Kim is conducting a study exploring the experiences of families during their elderly loved ones’ transition from assisted living to a nursing home; Terrie is attending to provide reflective questions and insights; and I am listening/questioning/providing poor jokes. A digital recorder is recording, in case we need to listen again to a particularly brilliant comment.

Kim has conducted about 15 interviews and is conceptually “stuck.” In this situation, Kim is asked to “tell her participants’ story” as a synthesized, single storyline (without notes); others listen and ask analytic questions. This process is different from coding, as the purpose is to open Kim’s data, give her study theoretical direction, and facilitate the building of concepts. The process also builds theoretical sensitivity, so the style of interviewing changes with the new insights, and theory building commences.

Let us look at the coding session. The dialogue is on the left, and the theoretical/methodological comments are on the right.

Example 1: Initial Interpretative Coding Seminar

---

DIALOGUE	RATIONALE
Kim: I am really stuck with the coding. I thought about uncertainty, I thought about grief; and I thought about powerlessness …	Jan realizes that Kim is thinking theoretically about relevant concepts—a good start when moving toward abstraction.
Jan: Tell me what is going on …
Kim: I see guilt. Lots of guilt. The women I interviewed yesterday—she cried the whole time …	Kim has chosen this concept to focus on. This is what emergence is all about.
Jan: Tell me about the guilt.
Kim: Lots of dysfunctional grieving.	Kim has immediately linked guilt with grieving—but Jan does not immediately explore dysfunctional grieving, rather builds on the relationship between the two concepts. For the time being, dysfunctional (a value) is sidelined.
Jan: Oh. Let’s think about that: Does guilt lead to grief? Or do they co-occur? That is, does the guilt remain, and the grief overlay it? Or, do you have to resolve the guilt in order to grieve?	First, slowly and carefully the two concepts are separated and their relationship explored.
How does one transition from guilt to grief?	These data should be in the story interviews, but may have to be teased out.
How do you distinguish between guilt and grief?
Are there different types of guilt?
Kim: Oh. [Mulling these thoughts over.]	Leave time in the discussion for “conceptual whirling”—a cognitive state, common in analysts.
Jan: How is guilt resolved? Don’t look in the literature; look in your data.	A common mistake is to be led by the literature, not one’s data. Your analysis is about analyzing the data—at this point stay there. This is the only way your study can eventually add to the literature.
Kim: (They say) talking about it is cathartic. They say, “I am sad, but I can move on.”   But others say:	Yes—this response to unstructured interviewing is well documented BUT here we are analyzing the interviews—take the student to that space.
“My parents had me and took care of me. But who says I have to take care of them when they are older? They had me by choice—I don’t have the choice.”
Jan: Great. Now name that kind of argument/justification.	Name the content—do not label the person.
Terrie: “Failed” or “rejected reciprocity?” Jan: Mmmm. Close.	Keep all suggestions on the table, but do not close the discussion.
Kim: “Unspoken,” “unfulfilled reciprocity?”	Critique all suggestions.
Jan: The problem is, it can’t be met. That is important. It is an “unspoken reciprocity that is not met.” It is the “not met” part that causes guilt. Think of a better label—then we will be doing real science.	Keep looking at all of the characteristics. The label MUST be inclusive.
Kim: The literature uses “family obligation.” How about “unfulfilled family obligation?”	This is exactly why qualitative researchers must be theoretically literature smart.
Jan: That still does not capture it head on—so our label is better.	Use data comparatively.
Go back and see what they all say, and use your conceptual scalpel.	You need a rich, dense data set.
Remember: you are analyzing the guilt; the feelings that underlie guilt; the causes of guilt.	Keep focused.
Look at their expectations of themselves; of their mother, their husband, their kids. Expectations of their friends; from work, church, community, and society at large.	Use different contexts. Different roles. Consider antecedents. Scope data.
Look in your data for different forms or types of guilt that accompany different situations. Look at the families reported behaviors and see if those provide clues about managing guilt. Make a lot of notes.	Sort data. Identify the types.
Kim: Remember the son who was in a different city, yet he micromanaged his mother’s care over the phone?	Discuss variation.
Terrie: Oh yes—too guilty to step away.	Identify characteristics that differentiate types
Kim: Jan, by “go back” do you want me to reinterview the ones I have done?	Yes—as new questions arise. You already have relations ships with those participants—do not close at the end of the first interview.
Jan: Look first in your data with these new eyes. When you do a new interview, continue the interview—do not interrupt in the middle. At the end of the interview, take the participant back to that point, and ask direct questions: Say: “I am really trying to get handle on how it feels to leave your mother in nursing home. You told me you felt terrible—tell me…”	Loop the interview back to the “rich points” to obtain additional data.
Ask: “You said you felt guilty. Tell me…”   Ask: “You talked with other families in the nursing home—was it like this for them also?” Use shadowed data (Morse, 2001). Sort. List the characteristics. At this point, try and get stories from all of the families. If you do not have the information in your interviews, call them back.	Shadowed data is asking your participant to report on others who felt or did not feel this way.       Compare/contrast.
Look at your nurses’ data and incorporate those perspectives.
Once you have started to identify the characteristics of guilt and its interaction with grief, the types of grief, and answered the questions listed earlier, then look in the literature, and see if you have something new. But at this point, build the concepts and analyze as much as possible, before you go back to the literature.	Once you have developed your concepts THEN go back to the literature.   Link to the literature.
The concept of guilt you develop may or may not be the central theme in your emerging theory—you are in the driver’s seat, so you decide on the direction our study takes. But today, guilt and its relation to the families’ behavior, and to grieving, seems to be an important piece of the puzzle.	Remember the AIM of your study—do not get lost, unless you have to—really MUST—reconceptualize your entire study.
Jan: Super. Build me a table that will enlighten us next week. And write! Write this piece now. And diagram. This will be a very important study, Kim.	Always write as you go. Tables clarify. Diagrams illustrate.

---

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Incorporating Theory Into Practice Research
2. The Theory of Becoming Resolute: Guiding Mental Health Practice With Survivors of Maltreatment
3. The Prototypical Method
4. Self-Transcendence and Self-Reformulation: One Concept or Two?

Stay updated, free articles. Join our Telegram channel

Join