Choice Points:
Creating Clinical Qualitative Research Studies
Paul V. Maione

The Qualitative Report, Volume 3, Number 2, July, 1997


Today, more psychotherapists are seeing the utility of studying their own and others' therapeutic work. With the growing popularity and acceptance of qualitative methods, the research process takes on special significance for the clinician/researcher. Using qualitative methodologies, therapists can conduct studies that are immediately relevant to their therapeutic work. In this paper, I discuss eight decisions or "choice points" clinician/researchers face when conducting clinical qualitative research studies. The choices I discuss are not all inclusive, yet they are representative of the choices most clinical qualitative projects require.


Too often in the field of psychotherapy an artificial line is drawn between the activities of researchers and those of therapists. One finds that people oriented toward research are not always interested in therapy. Likewise, therapists are oftentimes reluctant to perform research. One reason researchers may be wary of therapy is because they believe that it can be potentially harmful to use clinical techniques that have not been statistically proven effective.

Therapists, on the other hand, might refrain from conducting research because traditional quantitative methods typically yield statistical data. The results and conclusions drawn from many quantitative studies can be difficult to apply in an immediate way. Although useful, statistics provide little information about the nuts and bolts of therapy, like what therapists and clients actually say to one another. The challenge for therapists is to find research methods that fit with their clinical theories and their goals as clinicians. This is where qualitative methods are useful. Qualitative methods allow therapists to design studies that are immediately relevant to their ongoing clinical work.

Clinical qualitative research can thus be defined as inquiry conducted from a qualitative perspective on the nature of the therapy process. Such studies are performed to learn more about what happens during therapy sessions. This is accomplished by including the perspectives of the researcher, the participants of therapy, and sometimes both. The remarkable fit between family therapy and qualitative research is well documented (Atkinson, Heath, & Chenail, 1991; Gale, 1992; Moon, Dillon, & Sprenkle, 1990) and these authors point out that therapists and researchers, particularly qualitative researchers, perform many of the same activities. Well designed clinical qualitative research helps clinicians learn about their work in ways that can be immediately beneficial to ongoing therapy and therefore to clients.

It should not be surprising that many family therapists are turning to qualitative methods when designing and conducting research projects. Therapists who embrace ideas like constructivism, nonintervention,and second-order cybernetics (for a review see Gergen, 1985; Golan, 1988; Goolishian & Anderson, 1992; Hoffman, 1991; Watzlawick, 1990) are attracted to qualitative methods because the underlying assumptions of such methods are compatible with their clinical theories. For other therapists, like those embracing narrative and conversational metaphors (see Anderson & Goolishian, 1988; Epston, 1989; Keeney, 1991; White, 1989) qualitative methods are valuable because they focus more on discourse, text, and conversation. As such, clinicians/researchers are able to get close to the actual talk of therapy sessions, and the research endeavor takes on special relevance.

Like therapy, clinical qualitative research requires that clinicians/researchers make a number of choices. On a daily basis, therapists must decide who will participate in therapy, which questions to ask, and what techniques and interventions to use, among others. Similarly, clinical researchers are faced with a number of decisions when creating qualitative studies. These decisions or "choice points" are important because they help shape and guide the research endeavor. In this paper, I describe some of the choice points researchers face when creating clinical qualitative projects.

My list of choice points is not all inclusive, but they are representative of the choices most projects require. While discussing these choices, I provide examples from a study I recently completed; this study involved analyzing the discourse of family therapy sessions with adolescents incarcerated in adult jail. The study took place over several years beginning in 1992 when I helped start a project to provide family therapy to inmates at the Broward County Jail in Fort Lauderdale, Florida. Along with another therapist, I conducted therapy for approximately a year and a half. The inmates were adolescents being held over in the adult court system because of the seriousness of their crimes. The research I conducted fulfilled the dissertation requirements for the degree of Doctor of Philosophy in Family Therapy at Nova Southeastern University.

Choice One: What are you interested in learning more about?

In qualitative research, you begin with a curiosity. What do you want to learn more about? Maybe you are working in a shelter for abused women and would like to know more about how you and your colleagues work with this population. Or you might consider yourself a solution oriented therapist and want to know more about your specific use of the solution focused model. Some questions you might ask include: When do I typically ask the miracle question? How do I simplify the answers I get to the miracle question? How do I generally ask about exceptions to the problem? How can I categorize responses to questions about exceptions?

Another scenario might be that you have a case that you feel went particularly well and you want to know more about it. On the other hand, you might have a particularly difficult case and want more information about the source of your struggle. These are all legitimate starting points for the clinical qualitative researcher; the first step is finding out what you would like to learn more about.

In my study with juveniles in adult jail, my original curiosity was about the therapy process. I knew there was little research on family therapy in jails, and even fewer studies on what issues therapists and incarcerated people discuss in therapy. For me, this was a good place to start. Some of my initial questions were: How is talk organized in a jail context? What issues are important for families to discuss? Is the talk dominated by the jail context or does the family talk about events outside of their son being in jail? How can therapy provide ways for families to mobilize their resources and not be overwhelmed by the jail context? So again, the first choice you face is deciding what you are interested in learning more about.

Choice Two: Who will participate in your study?

This question is important for several reasons. First, you must have sufficient access to the population you are interested in studying. Secondly, they must be available to you at a reasonable cost. Some questions you will want to consider include: Is it ethical for me to study this population? Do I have to go through a review board? How will I secure peoples' consent to participate in the study? These are important questions because they underscore the researcher's ethical responsibility to treat people with dignity and respect. At most universities, for example, there are human subjects committees that must approve studies involving people before the research begins. In any case, researchers have an ethical responsibility to ensure the welfare of the participants.

Another issue when choosing participants is deciding how extensive the study will be. For example, will you be looking at only a single session of one family or will you be comparing first sessions over several cases? Lastly, you will want to consider how the research process will affect participants, recognizing that the line between therapy and research is often blurred. If you are conducting interviews with people about the therapy process, what steps will you take if the research elicits therapeutic issues that need to be addressed separately from the research endeavor? (For a review of this issue, see Gale, 1992; Shilts, Filippino, Chenail, & Rambo, 1995). In my study, participants included incarcerated adolescents, their families, and the therapists. All families signed a consent form prior to therapy stating that they were aware that sessions were being audiotaped and may be used for research purposes.

Choice 3: What research tradition will you follow?

There are many different research traditions available to the prospective clinical researcher. Another word for a research tradition is a paradigm. A research paradigm is a set of assumptions about the nature of reality, knowledge, and the goals and aims of the research process. In choosing a paradigm, what is often confusing is that similar constructs are stated by researchers from different fields. Oftentimes comparable ideas are made to sound different because researchers use terminology appropriate for their specific disciplines.

My advice when considering a research tradition is to first consider your interests and your original questions. Your choice of research tradition should be consistent with both of these as well as what you intend to study. Also, it can be helpful to find out what tradition other researchers have followed if their work seems particularly analogous to your own project.

In the jail project, I was interested in learning more about the talk of therapy sessions in a context like jail. My therapy was guided by a constructivist perspective, where the therapist is part of the therapy process. It made sense then to follow a research tradition that fit with my theoretical stance as a therapist. Constructivism is discussed in both the family therapy literature (For a review see Hoffman, 1991; Golan, 1988; Watzlawick, 1990) and the qualitative research literature (see Denzin & Lincoln, 1994; Schwandt, 1994). Though there is no single definition of a constructivist paradigm, according to Denzin and Lincoln (1994), "the constructivist paradigm assumes a relativist ontology (i.e., there are multiple realities), a subjectivist epistemology (knower and subject create understandings), and a naturalistic (in the natural world) set of methodological procedures" (pp. 13-14).

As a therapist operating from a constructivist perspective, I had to be sensitive to the variety of ways meaning was co-created by participants in conversation. I further presumed that meaning is embodied in language, and by studying conversation, I perceived and constructed meanings. The constructivist tradition I adopted also fit with my theoretical ideas about research. Instead of being the objective outsider, I adopted the stance that I was part of what I was researching, just as I am a significant part of the therapy process.

Choice 4: What literature will you research for your project?

In all types of research, a thorough and coherent review of the literature is imperative so readers can see what others have written thus far on the subject under investigation. It also helps place the study in context, which aids other researchers in understanding why the study is being conducted in the first place. An effective literature review requires striking a balance between being exhaustive and including only what is relevant. Most qualitative studies are narrow investigations focusing on a specific area. A good literature review should include only those studies that serve to highlight the present work and make it more understandable for the reader.

A common question people ask regarding literature reviews is, "Where do I start?" There are several ways of conceptualizing what areas you need to research for your project. Schooley (1995) discusses several conceptual maps that help organize a meaningful literature review. One such map is a Venn diagram. Borrowed from mathematics, a Venn diagram is a grouping of circles that are linked together to form a visual representation of where a study fits in with existing literature. Each circle represents a general area related to the study. Venn diagrams are valuable because they help you to see where the gaps in the literature are. They also help you build an argument for why your study is important, given the gaps in our current knowledge. A practical number of circles for a Venn diagram is three, but some studies may require four. A first step when constructing a Venn diagram is to consider some of the key words you might use to describe the area(s) your study addresses.

In my study, some key words were family therapy, crime, juvenile delinquency, punishment, adolescents, rehabilitation, qualitative research, and jail. After spending time in the library, taking a good look at my original research interests, and playing with different combinations of circles, I settled on three broad areas that I would need to include in my literature review. These areas included family therapy, adolescents in jail, and qualitative research. I used these broad areas to guide my initial foray into the literature. When I interlock these circles together, the space they overlap is where I see my study fitting in with the existing literature. It is important to remember that gaps in the literature do not exist independently, waiting for us to find them. Instead, gaps are created by the researcher based on what has already been studied and what their present research interests are.

Choice 5: How will you generate data for your study?

In clinical qualitative studies, usually the data is some form of written or spoken word. Data of this kind is collected by audio/video taping of therapy sessions, interviews, and/or field notes recorded by the researcher. Most qualitative researchers gather more data than will appear in their final study. There are several important questions related to data management. For example, how will you secure permission to gather data? How will the data be stored? Who will have access to the data? What type of transformation will the data undergo? In my study, I recorded therapy sessions at the jail on a micro-cassette recorder. I stored the tapes in a locked file cabinet and nobody had access to them except my co-therapist and myself. We accumulated over ninety hours of recordings. Of these ninety hours, I only chose to analyze and include approximately seven hours in my final study.

Choice 6: From what epistemological stance will you approach the data?

An epistemological stance refers to a set of assumptions about the world, knowledge, and human behavior. These presuppositions guide all types of research and they determine how a researcher interacts with his or her data. Epistemological stances vary, and at one extreme is the view that a researcher is separate from the data. From this perspective, a researcher tries to be as objective as possible, drawing a very distinct line between himself or herself and the data. In this approach, researchers value the idea of objectivity and are usually looking for absolute truths and causal relationships.

At the other extreme is the view that there is no such dividing line between researchers and their data. This perspective, sometimes referred to as a relativist perspective (or as I discussed previously as constuctivist), embraces the notion that meaning is constructed by an observer (researcher) and that it is context dependent. I adopted such a perspective throughout my study as I focused on the interaction between myself and the data. I looked for what I could discover about the data using myself as the research instrument. From this approach, I did not draw an arbitrary line between myself and the data but rather I considered myself part of an interaction that included me and the data. I was not interested in finding any absolute truths, but instead I wanted to see what I could discover about my data and ultimately my relationship with it.

Choice 7: What method of analysis will you use to study the data?

During data analysis the researcher takes a close look at the data they gathered and begins to make sense of it. In qualitative studies, this means spending a lot of time with your data. In the early stages of data analysis, you begin to play with different ways of organizing the data so you can start making sense of it. Sometimes qualitative researchers go through several organizing schemes before they hit on one that fits. It is helpful to remember that the purpose of data analysis is about creating meaning or sensemaking (Chenail & Maione, 1997).

There are many different analysis tools you can use to study qualitative data (Crabtree & Miller, 1992; Tesch, 1992). Analysis tools are simply ways of organizing data into meaningful units. They help you manage the data so that you can begin the process of meaning construction. Basically what you are doing with any analysis tool is drawing distinctions in the data. After drawing some initial distinctions, you will be in a better position to comment on what you are finding and whether or not you are moving in a productive direction. It is essential that your data analysis tool be consistent with your research interests, research questions, and epistemological stance.

In my discourse analysis of therapy sessions at the jail, I wanted to learn about therapy talk in a jail context. Because there was little research on therapy in jails, I was interested in creating some broad categories in the types of talk I was encountering. My choice of analysis tool was Recursive Frame Analysis (RFA). Developed by Brad Keeney (1987, 1991) and later refined by Chenail (1990/1991, 1991, 1995), RFA provides therapists, researchers, and theorists with a way of organizing and understanding talk in therapy sessions. As Keeney (1987) explains: "By enabling immediate access to the organization of a therapy session, it provides a general bridge for intersecting the intentions of the researcher, practitioner, and theorist" (p. 3).

According to Keeney (1987, 1991), the basic building block of RFA is a frame. He borrowed this term from Bateson (1972), who described a frame as a psychological concept for ways of understanding meaning in human and animal interaction. In studying therapeutic discourse we can create frames to show the contexts participants offer one another throughout a conversation. Keeney (1987) believes frames are best understood as embedded within other frames, with each frame contextualizing and, in turn, being contextualized by other frames. Of utmost importance for the user of RFA is to recall that an act of framing is an act performed by an observer.

A second and related notion to frames are what Keeney (1987) calls galleries. A gallery is the name given to a cluster of related frames that serves to further the organization of the conversation. Chenail (Rambo, Heath, & Chenail, 1993) explains that galleries are created "by chunking or collecting all the 'joining' frames and presenting them together in one configuration called a gallery, an RFA term for a collection of frames grouped together and named by a recursive frame analyst" (p. 166).

In my study, I began by noting different frames and galleries in the therapy talk. What I found initially was that this level of analysis was too narrow for my purpose. I was overloaded with frames and galleries to the point where creating some general categories was going to be difficult. I decided to look at a different conceptual level of analysis. RFA was a flexible enough method to enable me to do this. I switched my focus from analyzing frames and galleries to looking at galleries and wings of talk. A wing is one conceptual level higher than a gallery. With this distinction I was able to perform an analysis that was more consistent with my research questions. In short, what I found was that you could organize therapy talk in the jail into two different wings. The first wing I called Context Bound Wings. These wings contained talk that was directly rooted to the jail context. The second group I named Context Related Wings. Here the talk was further removed from the jail and more like therapeutic talk in other contexts.

The choice of analysis tool is dependent on what level of analysis you are aiming for. For example, researchers who are interested in the finer nuances of conversation would probably not use RFA as an analysis tool. They would be better served by a method that would give them access to the more subtle aspects of a conversation. In addition, they would probably study smaller chunks of discourse but pay greater attention to the many details contained in the talk.

Choice 8: How will you establish credibility in your study?

Credibility is an issue for all forms of research, yet it can be thought about and achieved in different ways (Ely, Anzul, Friedman, Garner, & McCormack Steinmetz, 1991; Maxwell, 1992). How a researcher establishes credibility is based on the epistemological assumptions guiding the research. Validity and reliability, terms commonly used in quantitative studies, are based on positivist assumptions that underlie quantitative and experimental research (Salner, 1989). In qualitative studies, researchers use several different terms to address issues of credibility with respect to how a study is conducted. Along with credibility are: transferability, dependability, confirmability (Lincoln & Guba, 1985; Zyzanski, McWhinney, Blake, Crabtree, & Miller, 1992), authenticity criteria (Guba & Lincoln, 1989), and trustworthiness (Atkinson, Heath, & Chenail, 1991).

Traditionally, the burden of proving the credibility of a study has been with the researcher. In clinical qualitative studies however, researchers are more likely to share the responsibility with the consumers of the research. (Atkinson, Heath, & Chenail, 1991). Oftentimes, qualitative researchers will build into their studies ways that the reader can assess for him or herself how credible the study and the findings are.

Another concern related to credibility is researcher bias. In qualitative studies, researchers view bias as unavoidable and they are more likely to state their biases openly. As Brody (1992) states,

Since the naturalistic investigator is him- or herself the research "instrument," naturalistic inquiry cannot avoid observer bias by using the instrument to insulate the experiment from the preconceptions of the investigator. Instead, open disclosure of preconceptions and assumptions that may have influenced data gathering and processing becomes an inherent part of the conduct of the inquiry. (p. 179)

Qualitative researchers are more likely to see bias not as something to avoid, but rather as a researcher's greatest asset. As Greene (1994) explains, "it is precisely the individual qualities of the human inquirer that are valued as indispensable to meaning construction" (p. 539).

In my study, I took several steps to establish credibility. As a first step, I examined some of my personal and therapeutic preconceptions regarding therapy, research, and people in jail. I included this discussion as part of my methodology section so the reader, knowing a bit more about me, could make better sense of the claims I made. Another way I addressed credibility was to look at the visibility of my data. Visibility refers to the extent others have access to the actual data of a study. I addressed visibility by providing transcripts from the therapy sessions I studied. By having access to the original data, readers could judge the accuracy of my claims and see how I drew distinctions in the talk. I accomplished this by providing readers with enough surrounding text that they could draw some conclusions of their own as they assessed what I was saying.


I have outlined eight method choice points that most clinical researchers will encounter when designing a qualitative study. As I mentioned earlier, this list is by no means exhaustive or representative of all the choices a researcher makes throughout the life of a study. Included, however, are some of the more pressing points. They are also the choices that face most researchers interested in studying conversation or therapy sessions. It is important to realize that these choices are not really separate from one another. Each choice is constrained and constrains the other choices. What you decide about one choice affects the other choices as well.

A unique advantage of qualitative methodology is the ability to rethink your method choices often throughout the course of a study. In fact, it is not uncommon to revisit a given choice several times during a study. It is vital, however, for the researcher to make the reader aware of the different choices he/she made; and they must let the reader know when and why if they decided to revisit a choice. This helps readers to understand the logic of the choices made while judging for themselves the value of these decisions.


      Anderson, H., & Goolishian, H. (1988). Human systems as linguistic systems: Preliminary and evolving ideas about the implications for clinical theory. Family Process, 27(4), 371-393.

      Atkinson, B., Heath, A., & Chenail, R. (1991). Qualitative research and the legitimization of knowledge. Journal of Marital and Family Therapy, 17(2), 175-180.

      Bateson, G. (1972). Steps to an ecology of mind . New York: Ballantine.

      Brody, H, (1992). Philosophic approaches. In B. Crabtree & W. Miller, (Eds.), Doing qualitative research (pp. 174-185). Newbury Park, CA: Sage.

      Chenail, R. (1990/1991). Bradford Keeney's cybernetic project and the creation of recursive frame analysis. The Qualitative Report [On-line serial], 1(2&3). Available:

      Chenail, R. (1991). Medical discourse and systemic frames of comprehension. Norwood, NJ: Ablex.

      Chenail, R. J. (1995). Recursive frame analysis. The Qualitative Report [On-line serial], 2(2). Available:

      Chenail, R. J., & Maione, P. (1997). Sensemaking in clinical qualitative research. The Qualitative Report [On-line serial], 3(1). Available:

      Denzin, N., & Lincoln, Y. (1994). Entering the field of qualitative research. In N. Denzin & Y. Lincoln (Eds.), Handbook of qualitative research (pp. 1-18). Thousand Oaks, CA: Sage.

      Ely, M., Anzul, M., Friedman, T., Garner, D., & McCormack Steinmetz, A. (1991). Doing qualitative research: Circles within circles. London: The Falmer Press.

      Epston, D. (1989). Collected Papers. Adelaide, South Australia: Dulwich Center Publications.

      Gale, J. (1992). When research interviews are more therapeutic than therapy interviews. The Qualitative Report, [On-line serial] 1(4). Available:

      Gergen, K. (1985). The social constructionist movement in modern psychology. American Psychologist, 40, 266-275.

      Golan, S. (1988). On second order family therapy. Family Process, 27(1), 51-65.

      Goolishian H., & Anderson, H. (1992). Strategy and intervention versus nonintervention: A matter of theory. Journal of Marital and Family Therapy, 18(1), 5-15.

      Greene, J. (1994). Qualitative program evaluation practice and promise. In N. Denzin & Y. Lincoln (Eds.), Handbook of qualitative research (pp. 530-544). Thousand Oaks, CA: Sage.

      Guba, E., & Lincoln, Y. (1989). Fourth generation evaluation. Newbury Park, CA: Sage.

      Hoffman, L. (1991). A reflexive stance for family therapy. Journal of Strategic and Systemic Therapy, 10(3 & 4), 4-17.

      Keeney, B. P. (1987). Recursive frame analysis. Unpublished manuscript, Texas Tech University, Department of Human Development and Family Studies, Lubbock, TX.

      Keeney, B. P. (1991). Improvisational therapy. New York: Guilford.

      Lincoln, Y. & Guba, E. (1985). Naturalistic inquiry. Beverly Hills, CA: Sage.

      Maxwell, J. A. (1992). Understanding and validity in qualitative research. Harvard Educational Review, 62(3), 279-300.

      Miller, W., & Crabtree B. (1992). Primary care research: A multimethod typology and qualitative road map. In B. Crabtree & W. Miller (Eds.), Doing qualitative research (pp. 3-28). Newbury Park, CA: Sage.

      Moon, S., Dillon, D., & Sprenkle, D. (1990). Family therapy and qualitative research. Journal of Marital and Family Therapy, 16(4), 357-373.

      Rambo, H. H., Heath, A., & Chenail, R. J. (1993). Practicing therapy: Exercises for growing therapists. New York: W. W. Norton.

      Salner, M. (1989). Validity in human science research. In S. Kvale (Ed.), Issues of validity in qualitative research (pp. 47-71). Lund Sweden: Studentlitteratur.

      Schooley, A. (1995). Playing with qualitative research: Designing a research project with diamonds and venns. The Qualitative Report [On-line serial], 2(3), Available:

      Schwandt, T. A. (1994). Constructivist, interpretist approaches to human inquiry. In N. Denzin & Y. Lincoln (Eds.), Handbook of qualitative research (pp. 118-137). Thousand Oaks, CA: Sage.

      Shilts, L., Filippino, C., Chenail, R., & Rambo, A. (1995). From solution-focused therapy to client informed research and back again. In I. Maso, P. A. Atkinson, S. Delamont, & J. C. Verhoeven (Eds.), Openness in research: The tension between self and other (pp. 133-146). Assen, The Netherlands: Van Gorcum.

      Tesch, R. (1992). Qualitative research: Analysis types and software tools. New York: The Falmer Press.

      Watzlawick, P. (1990). Munchausen's pigtail or psychotherapy and reality. New York: W. W. Norton.

      White, M. (1989). Collected Papers. Adelaide, South Australia: Dulwich Center Publications.

      Zyzanski, S., McWhinney, I., Blake, R., Crabtree, B., & Miller, W. (1992). Qualitative research: Perspectives on the future. In B. Crabtree & W. Miller, (Eds.), Doing qualitative research (pp. 231-248). Newbury Park, CA: Sage.

Author Note

+The author wishes to thank Dawn Shelton, Ron Chenail, and Jan Chenail for their helpful suggestions in the preparation of this paper.

Paul V. Maione, Ph.D. is the Clinical Director at the Center for Family and Relational Therapy in Fort Lauderdale, Florida and an adjunct faculty member in the School of Social and Systemic Studies at Nova Southeastern University, 3301 College Avenue, Fort Lauderdale, Florida 33314 USA. His e-mail address is

Article Citation

      Maione, P. V. (1997, July). Choice points: Creating clinical qualitative research studies [37 paragraphs]. The Qualitative Report [On-line serial], 3(2). Available:

Paul V. Maione
1997 copyright

Return to the top of the paper.

Return to the Table of Contents.