Printable PDF Version

Toward Holism: The Significance of Methodological Pluralism

Janice M. Morse and Seung Eun Chung

Janice M. Morse, PhD (Nurs), PhD (Anthro), D.Nurs, Professor, Faculty of Nursing and Director, International Institute for Qualitative Methodology, University of Alberta, Edmonton, Alberta

Seung Eun Chung, PhD (Nurs), Associate Professor, Department of Nursing, Chongju National College of Science & Technology, Korea

Abstract: In this article, we make a case for achieving comprehensive understanding by using multiple methods in a research program. We argue that used alone, qualitative methods are not holistic, but actively provide segmented perspective on a research problem. We illustrate this concern using three types of data. The limitations of individual methods used alone can be systematically overcome with investigators increasing methodology skill and developing research program.

Keywords: multi-method research, triangulation, rigour.

Citation information:
Morse, J. M. & Chung, S. E. (2003). Toward holism: The significance of methodological pluralism. International Journal of Qualitative Methods, 2 (3). Article 2. Retrieved [INSERT DATE] from http://www.ualberta.ca/~iiqm/backissues/ 2_3final/ html/morsechung.html


Qualitative research has long been considered to be holistic, contextualized and comprehensive, to the extent that researchers have accepted this advantage without question. Clearly when compared with quantitative methods–in particular experimental design, where the control of extraneous variables is the goalqualitative methods are holistic. However, in this article we examine this premise of holism, to ascertain if, indeed, such a claim is valid or perhaps if it should be made less boldly.

The nature of the problem

Qualitative methods have been considered particularly appropriate for nursing research. Nursing considers itself a discipline that uses the holistic perspective, encompassing the patient and family, the context, the patient’s past and future goals, the caregiver, and everyday life in nursing’s domain. Similarly, qualitative methods have the ability to record and include multiple dimensions within each method. In addition, despite tremendous advances in nursing research in the past two decades, nursing knowledge and theory remains relatively thin and underdeveloped (Meleis, 1997), so that qualitative inquiry, with its inductive approaches, is frequently justified by researchers arguing for their studies on the grounds that little is known about the topic of investigation. These factors have resulted in a preponderance of qualitative inquiry in nursing research.1

However, because they are considered holistic, we argue that reliance on qualitative methods is providing us with false confidence and holism is only partly accomplished. Qualitative methods are never completely holistic, containing assumptions and perspectives that partition reality and provide subtle biases, often silently excluding as they focus and as inquiry becomes directed as it proceeds2. In the next section we will explore the myth of holism associated with qualitative methods.

Ramifications of methodological perspectives

In qualitative inquiry, as in all research, the researcher’s question is driven by the identified problem. Research texts stress the benefits of such problems and questions being identified within the clinical setting. The obvious advantage of such an agenda is that research products become useful, implementing change and improving care. An obvious disadvantage is that such questions by and large serve the caregiver rather than the patient. For example, if one surveys a nursing research journal one may instantly note that the articles provide information on the nurse-patient relationship, rather than the patient-nurse relationship, on appropriate procedures with much less information on the patient’s perspective of those procedures. This value extends even to the configuration of patient rooms in hospitals that have been designed for caregiving convenience over patient safety3. Hence, research problems and questions whether they use qualitative or quantitative inquiry are frequently developed from the nursing perspective rather than the patient’s perspective or a more balanced combined interactive approach. Qualitative inquiry is therefore as guilty as quantitative inquiry for focusing on particular actors within the setting according to the researcher’s agenda.

Albeit, there’s a growing body of literature deliberately seeking the patient’s perspective, often even including family members. The first may be classified as seeking to understand the patient’s experience, so that nursing care may be modified to be more successful. An example of such research–which still seeks to benefit the nursing agenda–would be to understand the patient’s experience of pain so that analgesics may be more appropriately administered. This agenda does not seek, for instance, to empower the patient, by simultaneously imparting strategies of enduring to decrease the dependence of the patient on the nurse. Focusing on particular persons therefore delimits perspective and the holistic scope of the project.

A second mode of delimiting holism comes from the researcher’s agenda, which may be overt or covert. An example of an overt agenda forming the theoretical basis for the study would be the deliberate utilization of feminist or critical theory as a basis of the research project. For example, if our researcher noted a clinical problem perhaps related to the use of restraints with the elderly, the use of critical theory imparts a particular tone and expected outcome to the research project which dictates what is considered data and how data are coded, analyzed, and interpreted. Of course this may have the advantage of expediting the researcher’s agenda, just as it may be a disadvantage by resulting in a problematic bias. Our point is these advantages and disadvantages must be made explicit. The use of such frameworks is an increased risk to validity when they are used covertly in qualitative inquiry, as the researcher’s agenda may be concealed.

Researchers are frequently constrained by the nature of the setting or participants. For instance, Morse’s research videotaping in the trauma room continued despite the fact that the camera was located along a wall and the view of the patient was frequently obstructed by caregivers, thereby limiting the usefulness of data (Proctor, Morse & Khonsari, 1996). Further ethical considerations required that a mosaic patch be placed over the patient’s face, excluding facial expression from analysis. Even further limitations can extend from characteristics of patients themselves; for instance, neither patients with advanced Alzheimer’s disease nor those who are pre-verbal infants can be interviewed, or some patients may not consent to videotaping.

Of greatest concern is that qualitative researchers are unwittingly and unknowingly restricting their holistic perspective by subscribing to a single qualitative method. This may be for several reasons. First, much research in nursing is conducted by doctoral students, and these projects are by necessity small and circumscribed. They usually use a single method, rather than a mixed-method or multi-method design. While the scope of the project can be increased by using a mixed-method or multi-method design, it also increases the student’s workload and is therefore not feasible. However, the limitations of perspective from using a single method is not acknowledged in the research reports. It is unfortunate that advisors are often knowledgeable in only a single method, such as phenomenology or grounded thory, so that entire emerging research groups do not expose their students to other qualitative methods.

A second reason that holism is limited relates to the immaturity of research within the nursing discipline. Research programs with a single researcher or research group conducting multiple studies to increase understanding and scope, eventually leading to intervention within a single topic, are rare. If the researchers understood that the initial studies were restricted and that further research either extending in scope or vertically to the macro or micro levels would provide valuable information, this problem would be of less concern. As the number of nursing researchers increases, so must substantial funding be available to support emerging research teams. As nursing research is not generally disease-focused, expecting researchers to fit into medical funding agencies’ requirements is not generally successful.

Constraints from using a single method

A third mode of delimiting holism extends from the qualitative method selected. Qualitative methods have been developed using disciplinary theories to develop that particular discipline’s knowledge (Morse, 1994). Ethnography was developed in anthropology using culture as its theoretical foundation; grounded theory emerged from sociology based on symbolic interaction (Morse, 1994). Though there have been some attempts in nursing to develop methods based on disciplinary assumptions (such as Leininger’s [1987] ethnonursing), nurses primarily select established qualitative methods, with phenomenology, ethnography and grounded theory most frequently used. As with theoretical frameworks with quantitative inquiry, so do qualitative methods provide perspectives that dictate what is considered to be legitimate data, foregrounding some aspects in a setting and excluding or silencing others. Some methods enable the inclusion of particular types of data, omitting others. This point is of critical importance and will be elaborated on in the next section.

Ramifications of using a single method

Mature research methods4 provide implicit instructions for what is considered as data, and how such data are collected and analyzed. While quantitative research conceals raw and emotive experiences with numbers, so do particular qualitative methods conceal by excluding or ignoring certain aspects. This is clearly shown on Table 1. We have created data from the description from Betty Rollins’ First You Cry (1976) in which she describes an interaction where a physician informs her and her husband that she has breast cancer. We have converted this narrative into three data sets: data that would be obtained using unstructured interactive interviews, as would be collected for grounded theory; a recording of the dialogue, as would be used in conversational analysis; and participant observation field notes as would be recorded for ethnography.

Table 1. Comparison of types of analyses

Unstructured Interview data

Conversational analysis transcript

Participant observation notes

I don’t remember exactly how he (physician) put it, because as soon as I got the gist of what he was saying, my head seemed filled with air my eyes got hot. He was telling me that I had something there, and needed surgery.

Once he stopped talking, I realized I was supposed to talk. I thought he wanted me to say if I whether I wanted just to have my breast cut off, or whether I wanted my breast cut off and some other things too.

I looked at Arthur. He told me later he was unable to get the look on my face out of his mind. I didn’t want to be rude, so I asked if I had cancer… It was strange–he hadn’t used that word. I soon learned that cancer is a word doctors almost never use. I heard myself speak. I asked him how likely it was–it was weird… I asked him if it was likely… I think I made him uncomfortable. (long pause). He said people always want to know, but he could not be sure. And he stood up.

Suddenly, I fell down. I didn’t faint, exactly, because I didn’t entirely lose consciousness. Nor did I fall far, or get hurt. The office was so small that when I fell, Arthur caught me. There was a small sofa in the room, and I remember being placed on it. It was too short for my legs, so Arthur hung them over the arm of the sofa, like wet towels. I said I would be alright, but as soon as I said it, I began to cry, the bad, loud gasping kind. I wanted to hold something, so I held my face. I held it hard with both hands–as if it were someone else’s.

  1. phys: We have the results:: I am afraid to tell you
  2. that they are positive. There is definitely
  3. something there:: A mass, I’m afraid::° There is
  4. good chance of malignancy.° (0.2)
  5. We:: have to do surgery as soon as possible.
  6. (0.2) There are different kinds of mastectomy, as
  7. you probably know. (0.3) Some women say they
  8. want a separate procedure for reconstruction.
  9. Studies show that it makes no difference in the
  10. outcome, but in my own experience it should be
  11. done in one procedure. But, - of course, it’s up to
  12. you>
  13. (0.54) ((looks at husband, then back to physician))
  14. Pt: Are you saying that you think I have cancer?
  15. (0.10) I mean, –I know you can’t know for sure,:
  16. but what are the odds?::: what percentage?
  17. :: What is the likelihood:::?
  18. (0.4)
  19. Phys: Everyone wants numbers. It’s very hard to
  20. say::maybe:°seventy-thirty:: sixty-forty:: I don’t
  21. know °–"
  22. Pt: Are you saying, <do you mean it’s> sixty or
  23. seventy percent likely. You mean it’s
  24. LIKELY?=
  25. Phys: Look, percentages are just percentages.
  26. People want numbers (.hhh), but these numbers
  27. are always unreliable. You don’t really know
  28. until-:::but–"
  29. (1.2)
  30. ((Patient faints. Husband helps her onto a sofa))
  31. Pt:I’ll be a(c)ll right .hhh
  32. ((Patient begins to sob))

Physician enters the room and as he enters both the patient–a smartly dressed woman about 45 years–and her husband stand. The physician stands one foot and then the other looking at the chart he is holding and does not look directly at the patient or her husband. He tells them he has the results of the tests and that there is "something" there, and that surgery must be scheduled as soon as possible. He outlines several options for surgery, and tells the patient that she must decide if the reconstruction is done with the mastectomy or as a separate procedure.

There was stunned silence. The patient and her husband looked at one another. The patient paled, place her hand on her husband’s arm, and began asking questions about the likelihood of cancer in an odd-sounding voice. As the physician began to give more information, she suddenly swayed and sank to the floor. Her husband tired to catch her. He picked her up, and carried her to the sofa, and put her down with her legs trailing over the end. The physician went out to get the nurse.

The woman then spoke in a high pitched voice, saying she would be alright, and suddenly began sobbing, loudly, holding her face.

The unstructured interview data (Table 1, Column 1) provides us with an account of the incident from the perspective of the patient. The account provides us with exceedingly rich information about how she felt, how she was thinking at the time, what she understood and did not understand and how she behaved. Her reports of the physicians telling and of her husband’s response is tangential to her main story and is an incomplete description. Her description does not include any of the physician’s or her husband’s emotions, and limited interpretation of their experience. However, this type of data providing participant’s perceptions is essential for the development of grounded theory. While participant observational data may also be incorporated into grounded theory (Benoliel, 1996), it is this narrative that forms the basis of the developing process.

Conversational analysis (Table 1, Column 2) is a method of recording dialogue as it occurs and as it is spoken. The notations allow for documentation of the pacing and intonation, and the turn taking of speakers. In other words, while we have an excellent record of what was said when and how, we do not have any data about what was heard, the response and the understanding of the participants. Neither do we have any information about behaviours of participants. Conversational analysis is balanced, not providing precedence of the data obtained from one speaker over another, but at the same time it is partitioning experiential and behavioural dimensions.

On the other hand, notes made from observations (Table 1, Column 3) provide us with a reasonable and balanced account of the actions of all the participants, and a report of some of the dialogue. The detail and accuracy of such field notes depends upon the skill of the recorder, and may be improved if data are recorded initially on videotapes. If events are not recorded using videotapes, then we do not have a completely accurate report of the dialogue or behaviours. Observational data provides no information on the meaning of the events to the participants and this must be inferred by the researcher. Thus, participant observation provides an account that is balanced across participants and includes some of the behavioral and conversational nature of the events.

Attaining holism

At first glance, the obvious way to overcome the limitations of partitioning perspective would be to add additional data collection strategies or multiple methods to a single research project. While adding data collection strategies clearly increases the scope of the project by for instance including observational data in the grounded theory project or additional data sources to the ethnography, it does not overcome the problems of focus contributed by theoretical frameworks or the focus provided by the problem and question statements discussed earlier. Therefore, increasing strategies is only a partial solution to resolving the holism problem.

Do multiple or mixed method designs overcome the limitations afforded by a single method? Again, simultaneous or sequential triangulation of more than one qualitative method or combining qualitative and quantitative methods provides a more balanced perspective, moving toward holism. When used sequentially, the researcher has the prerogative of deliberately determining the direction of inquiry and the level of analysis (macro or micro), and of selecting the focus of inquiry to answer questions that emerge from the first phase or to add understanding to gaps, or fascinating or thin areas identified in the first study (Morse, 1994).

The use of multiple methods leans toward developing a systematic research program, with one study dictating the direction and nature of the next. In this way, the researcher may carefully identify and encompass the scope of the phenomena or project, with each study being complete in itself (Morse, 1994). With minimal overlap between these projects, but with each project validating and extending the previous, the results may be fit together to form an understanding of the concept. For example, an understanding of the nature role and processes of comfort was developed over a period of eight years and consisted of 56 publications5.

In North America some of the perceived limitations of research being conducted within a single discipline are being addressed by encouraging the development of interdisciplinary teams. Not only does such an approach increase the repertoire of research methodologies available as the researchers (and their disciplines and their knowledge) join together, but also the theoretical insights and combined practical experience adds to the richness of the study design, analysis, and development of implications and interventions. However, such team research may become unwieldy and extraordinarily expensive.

As research becomes more holistic, new problems arise. Obviously, it mocks the results to be developed holistically into a useful theory, only to be dismantled and published concept by concept in many journals. In North America, the publication of monographs and longer texts are becoming more difficult as the market for purchasing books declines, and as universities are less likely to award credit for tenure and promotion for non-refereed publications. In light of this, several research groups link publications developed from a single project in innovative ways. For example, research teams are beginning to use project acronyms in their list of authors, so that bibliographic retrieval services could locate all pieces of the project.

Summary

In this article, we have argued that qualitative methods, while claiming to be holistic, actually partition phenomena, settings and concepts according to researchers agenda, the methods used and the underlying theories and concepts. While we recognize that single projects may by necessity be focussed, the boundaries and limitations of such projects should be made explicit. However, a single method does not build a comprehensive and competent research program if the researcher’s goal is to understand a single concept holistically. It is our obligation to continue to become increasingly versatile as researchers, deliberately building our methodological toolboxes. When this is not possible we should be collaborating with researchers who have the methodological knowledge that complements our own. This strategy will have the benefit of developing more certainty in our qualitative research products, of enabling qualitative inquiry in nursing to be moved from describing experiences forward to the identification of interventions and to at last allow us to have much important research to contribute to our discipline. Qualitative inquiry has the capabilities of developing theories for praxis, but this effort has largely been truncated by a lack of concerted effort in inquiry. We look forward to the development of solid, comprehensive and pragmatic qualitatively derived theory impacting on and directing nursing.

Endnotes

1. The trend continues despite the fact that most nursing faculty’s primary preparation is in quantitative inquiry, resulting in a lack of mentors for graduate students. Back to text.

2. Qualitative inquiry may begin more holistically but as inquiry proceeds it becomes increasingly targeted on the phenomena of interest. Back to text.

3. For instance the gap between the bed and the bathroom may be directly attributed to patient morbidity and mortality due to patient falls. Despite this little effort is made to reconfigure the rooms to reduce the risk of patient mobilization. Back to text.

4. Mature research methods are those that are well developed with a theoretical foundation and clearly described techniques. Immature or emerging research methods, such as research that uses semi-structured interviews or focus groups, as sole-source data do not meet the above criteria of mature. Back to text.

5. A list of these publications is available upon request from either author. Back to text.

References

Benoliel, J. Q. (1996). Grounded theory and nursing knowledge. Qualitative Health Research, 6, 406-428.

Leininger, M. (1987). Importance and uses of ethnomethods: Ethnography and ethnonursing research. Recent Aadvances in Nursing, 17, 17-36.

Meleis, A. I. (1997). Theoretical nursing: development and progress (3rd ed.). Philadelphia, PA: Lippincott-Raven.

Morse, J. M. (1994). Designing funded qualitative research. In N. K. Denzin & Y. S. Lincoln (Eds.). Handbook of qualitative research. (pp. 220-235). Thousand Oaks, CA: Sage.

Morse, J. M., & Richards, L. (2002). Read me first for a user’s guide to qualitative methods. Thousand Oaks, CA: Sage.

Proctor, A., Morse, J. M., & Khonsari, E. K. (1996) Sounds of comfort in the trauma center: how nurses talk to patients in pain. Social Science & Medicine, 42, 1669-1680.

Rollin, B. (1976). First you cry. New York: Signet.


Editor’s Note: This article was previously published in the Korean Journal of Nursing Query, 12 (1). Reprint by permission.