| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
|
|
|
|||||||
|
|||||||||
a University of Kansas School of Nursing, Kansas City
Sarah Forbes, University of Kansas School of Nursing, 3901 Rainbow Blvd., Kansas City, KS 66160-7502 E-mail: sforbes{at}kumc.edu.
 |
Abstract |
|---|
 Top Abstract Types and General Features...Qualitative Research Design and...References |
|---|
THE growing challenges of promoting health and managing illness in an ever-changing health care system require an arsenal of research approaches. Qualitative methods are being used with greater frequency as interdisciplinary teams attempt to understand and explain complex problems. For example, when one of us (SF) analyzed information from the Minimum Data Set (MDS) collected on nursing home residents, quantitative methods provided specific information such as the functional and cognitive status of residents, use of advance directives by residents, and the number and type of medications residents were taking. Yet, these data were insufficient to answer the question of why a cohort of high-functioning elders remained in nursing homes 6 months postadmission. Only a qualitative approach, using interviews and participant observation, could reveal the complex decision making and social processes affecting length of stay for this group of residents (1)(2).
The literature is replete with similar examples of the value of a qualitative approach to gerontologists. As in the example provided above, qualitative research offers the gerontologist an in-depth view into complex social phenomena associated with aging, phenomena that potentially would be incompletely captured by using only quantitative methods. Gerontologists can quantify the number of elders who complete advance directives and qualitatively describe factors that facilitate or impede the actual process of completing the advance directive; gerontologists, through participant observation and interviews, can gain insights into why interventions to improve care in nursing homes, such as pain management or continence care, are not sustained; and gerontologists, through in-depth interviews, participant observation, and document review, can evaluate social programs and have detailed information from which to advocate for health policy. The purpose of this article, then, is to define and describe the main features of qualitative research and to examine ways in which this methodology is relevant and useful in gerontological studies.
|
Types and General Features of Qualitative Research |
|---|
|
TopAbstractTypes and General Features...Qualitative Research Design and...References |
|---|
A comparison of the major features of quantitative and qualitative research is summarized in Table 1 . We can see that these features are quite different from each other. Quantitative research requires the investigator to stand apart from research subjects, use a deductive approach, and use designs and instrumentation that test hypotheses through statistical analyses. It is the paradigmatic assumptions underlying these two approaches to research that account for the differences between them. Quantitative, or positivist, research assumes that there is an objective, measurable reality or "truth" to be discovered, that the results of the study must be generalizable to other populations across time and various contexts, that cause-effect relationships can be identified, and that scientific inquiry is value-free. In contrast, qualitative research assumes that there are multiple realities and perspectives to be discovered, that the researcher and participants are mutually interactive and simultaneously influence one another, that cause-effect relationships are impossible to determine, and that all inquiry is value-based (3)(4)(5)(6).
|
When To Use a Qualitative Approach
Qualitative research is not appropriate for every kind of investigation. There must be a match between the purpose or research question and the approach (11). Qualitative studies are particularly useful when the topic under investigation is insufficiently understood or when it is best communicated through detailed examples and rich narratives. They are useful as well for the identification of theoretical relationships and the development of quantitative tools. They are appropriate when the focus of inquiry concerns complex human emotions and interactions in a social context incompletely explained with quantitative measures. For example, in an intervention study proposed by Boyle and colleagues (12) on end-of-life communication and decision making in intensive care units (ICUs), quantitative measures regarding provider and family stress, communication, and satisfaction surrounding withholding and withdrawing of life-sustaining treatment will be obtained. However, given the complexity of human interactions involved in these issues, qualitative methods will be used to investigate provider-to-provider and provider-to-family interactions, including observations of both verbal and nonverbal communication. These observations will allow the investigators to move beyond participants' selective perceptions to an understanding of implicit assumptions and the interactive dynamics guiding communication and decision making.
Types of Qualitative Research
Qualitative research is an umbrella term often used interchangeably with naturalistic inquiry, interpretivist research, or constructivist research. It encompasses many different strategies of inquiry. A decade ago, Tesch (13) identified at least 26 different qualitative research strategies across a variety of disciplines. Since then, qualitative research has proliferated to become a field of inquiry in its own right (5)(6), with a stream of excellent publications explicating a wide range of strategies.
The three most commonly used qualitative approaches are ethnography, which originated in anthropology; grounded theory, which was developed in sociology; and phenomenology, descended from philosophy and psychology. These three approaches share some features in common, but have different research purposes and terminology. Ethnography has as its purpose the description of a culture and the meaning of human behaviors within the cultural context. The goal of a grounded theory study is to inductively develop a theory "grounded" in data obtained through direct observation, interviewing, and field work. Phenomenological studies rely primarily on in-depth interviews with a small number of people who share a common experience, often one that is difficult to measure, such as an emotion like suffering or courage. The goal is to identify and describe the essence of the experience as it is lived by those who have the experience. If you choose to use one of these research strategies, it is important to read some of the seminal writings about them and to accurately use the terms specific to that strategy. The same is true for other specific qualitative approaches.
Many excellent qualitative studies, however, do not ascribe to any of the above strategies, but use more generic terminology while adhering to standards for scientific rigor (see section on Trustworthiness). There are a number of excellent guides demonstrating this usage (e.g., (4)(14)(15)). Sandelowski (16) states that qualitative description is the methodology of choice when one's goal is a straightforward description of the phenomena of interest. See Table 2 for a comparison of the more common types of qualitative strategies (17).
|
|
Qualitative Research Design and Methods of Data Collection |
|---|
|
TopAbstractTypes and General Features...Qualitative Research Design and...References |
|---|
Setting
Qualitative studies take place in natural settings, that is, where participants work and live. Given that researchers are interested in the interplay of human behavior in social settings, gaining entrée to the setting in the most unobtrusive manner possible is essential. Not all qualitative studies have a specific geographic site such as nursing home, hospital ward, or patient's home, though many do. Phenomenological studies, for example, rely primarily on interviews, which are usually conducted in settings chosen by the participants.
Sample
In qualitative research, people, events, activities, times, and documents may be sampled, and sampling is purposeful rather than random (11). Patton (15) identifies a wide range of types of purposeful sampling, including maximum variation, extreme or deviant cases, information-rich cases, critical cases, confirming cases, and politically important cases. Snowball or network sampling is another participant selection strategy in which the initial participants identify others whose perspective is important to the research question. For example, when doing a community study, one may begin by interviewing official community leaders, but in the process learn there are informal leaders in the community who can provide important data as well.
The exploratory nature of qualitative research typically requires investigators not to prespecify a study population in strict terms, lest an important person, variable, or unit of analysis be overlooked... Sampling aims to cover a range of potentially relevant social phenomena and perspectives from an appropriate array of data sources. Selection criteria often evolve over the course of analysis, and investigators return repeatedly to the data to explore new cases or new angles (19). (p. 359)
The investigator, then, may have some initial sample selection criteria in mind, but modifies these as the study unfolds. Similarly, a strict sample size is not predetermined. The criterion for "how many" and when to quit is when no new data are being obtained, sometimes referred to as redundancy (3).
Researcher's Role
In qualitative inquiry, the investigator is the research instrument. This is in keeping with the epistemological assumptions of the inquiry paradigm, which specifies mutual and simultaneous shaping between the investigator and the participants. Because the goal is to understand the social world from the participants' point of view (the emic perspective), it is necessary to be able to experience the subjective world of the performer, or to take the role of the other, to the extent possible. This requires sensitivity, flexibility, and openness. The investigator must be aware of, and guard against, personal prejudices and preconceptions in data gathering and interpretation. He or she must also have a high tolerance for ambiguity, be an astute observer and listener, and have the tenacity to continue probing a social phenomenon until it is thoroughly understood.
Data Collection
The primary methods of data collection in qualitative inquiry are observation, participant observation, interview, and documents analysis—usually a combination of most or all of these. Many qualitative studies in the health care field focus primarily on formal interviews. A great deal can be learned from this systematic questioning of participants about their world; however, limiting the study to interviews alone does not allow for validation through observation of what is said. We know from cultural studies that interview respondents often provide answers that correspond with a cultural ideal. It's only through observation and participant observation that one learns the range and tolerance for deviation from the behavioral ideal. For example, from data generated through interviews, nurse and physician providers perceive they provide good pain management for nursing home residents (10). However, data generated through interviews with residents, observations of staff responses to residents' reports of pain, and medical record review refute the providers' perceptions. Thus, relying solely on interview data would result in a false interpretation.
Because an important aspect of qualitative research is the context and how it influences behaviors, it must be fully and accurately described in field notes. This includes persons, activities, events, and interactions. The results of Kayser-Jones' studies on undernutrition in nursing homes provide a superb illustration of how the context, that is persons (lack of staff, lack of staff supervision), activities and events (lack of time for meals, loud and crowded dining rooms), and interactions (lack of social support, inability to speak English), results in poor dietary intake (7)(8)(9). Table 3 contains a guideline we use in our qualitative research courses to assist students with what to focus on in describing the setting.
|
Some of the qualitative approaches identified above have specific terminology and procedures for data analysis. Qualitative analysis in general, however, consists of five fundamental steps. These are (i) creating written text from the interview and field data; (ii) identifying units of analysis, e.g., paragraphs, sentences, or meaning units; (iii) reducing the data through a system of coding or classification; (iv) clustering the codes into categories, themes, or patterns; and (v) providing interpretations of the relationships that compose the descriptive or explanatory framework. Qualitative analysis is essentially a creative process that includes decontextualization of the data into units of analysis and recontextualization into a new whole that allows for new insights and interpretations. A number of texts provide detailed guidelines for ways to approach the analysis of qualitative data (e.g., (13)(15)(20)(21)(22)(23)).
Trustworthiness or Validity
As in all research, a critical question in qualitative inquiry is, "How do we know this is a good study?" Specific to this approach is whether the interpretations accurately depict the phenomena of interest. The most widely used set of criteria for evaluating qualitative studies is one developed by Lincoln and Guba (3), although there is currently a proliferation of articles on this topic (19)(24)(25)(26)(27)(28)(29)(30)(31)(32). Lincoln and Guba's criteria are credibility, transferability, dependability, and confirmability. These parallel, in order, the criteria of the positivist paradigm, i.e., internal validity, external validity or generalizability, reliability, and objectivity.
Credibility is manifested through five activities: prolonged engagement with the participants, observation over time, triangulation, peer debriefing, and member checks (3). Triangulation refers to use of multiple data sources and more than one investigator. Peer debriefing is a process of conferring throughout the study with a colleague who is not involved in the study but who has relevant expertise. And member checking (33) means taking one's interpretations back to selected participants to confirm the accuracy of the final framework.
Transferability means that in the final report, the investigator provides sufficient description of the context of the study that the reader can make a judgment as to whether the findings may be applicable or transferable to another, similar context. The criteria of dependability and confirmability are both met through the use of an audit trail, which means that an outside reviewer is able to examine the raw data and the decision process to determine that appropriate logical inferences were made. Use of qualitative software programs such as QSR NU*DIST make following the decision trail much simpler, because this capacity is built into the program. Confirmability is additionally enhanced through triangulation and through the investigator keeping a journal in which are recorded feelings, reactions to phenomena observed, and self-reflection on changing preconceptions and managing biases.
Giacomini and Cook (19) reduce evaluative criteria to two critical questions: (i) Was the data collection comprehensive enough to support rich and robust descriptions of the observed events; and (ii) Were the data appropriately analyzed and the findings adequately corroborated? For each of these questions, they give numerous examples of how the criteria can be met.
Summary and Conclusions
Given the complexities of our health care delivery system and the development of long-term, chronic illnesses from an interaction between biological forces, human behavior, and social environments, researchers need to combine the strengths of multiple disciplines. The development of interdisciplinary research teams requires an appreciation of various methodological approaches. We have attempted to provide the audience with a beginning understanding of qualitative research methods. One limitation of an article-length presentation is the necessity of oversimplification of the topic. Remember that simply adding a few open-ended questions to an otherwise quantitative study does not constitute qualitative research (34). An understanding of the underlying assumptions and methodological procedures of qualitative research is imperative. We encourage researchers contemplating the use of qualitative methods to access the expertise of a qualitative consultant and to refer to many of the references cited in this article.
Received June 25, 2001
Accepted August 31, 2001
|
References |
|---|
|
TopAbstractTypes and General Features...Qualitative Research Design and...References |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Cohen-Mansfield and S. Lipson Medical Staff's Decision-Making Process in the Nursing Home J. Gerontol. A Biol. Sci. Med. Sci., March 1, 2003; 58(3): M271 - 278. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| All GSA journals | The Gerontologist |
| Journals of Gerontology Series B: Psychological Sciences and Social Sciences | |
