Janice M. Morse

Determining Sample Size

L month I attended a conference presentation given by a senior qualitative researcher. His project used a longitudinal design with multiple interviews conducted over a period of months. In response to a question regarding sample size, he explained that he obtained the number of participants necessary for his study by looking at a table that Morse had published (see Morse, 1994, p. 225). He had used this number in his proposal for estimating the number of participants without considering the number of interviews. Because his design used many more interviews than the studies in Morse’s table, he was clearly going to drown in data in a very short time. Evidently, it is time to clarify the issues in sample size once and for all. By clarifying the assumptions underlying sample size recommendations, I will not feel quite so responsible when someone takes my work at face value. Estimating the number of participants in a study required to reach saturation depends on a number of factors, including the quality of data, the scope of the study, the nature of the topic, the amount of useful information obtained from each participant, the number of interviews per participant, the use of shadowed data, and the qualitative method and study design used. Once all of these factors are considered, you may not be much further ahead in predicting the exact number, but you will be able to defend the estimated range presented in your proposal. Because the actual number of participants is still an unknown, should data collection not proceed smoothly when writing the proposal, it is wise to overestimate the sample size rather than to underestimate so that funds are available to collect all the necessary data.

Qualitative Nursing Research: A Contemporary Dialogue

Is qualitative research an end in itself or the beginning of a process? qualitative nursing research - a free-for-all, Janice M.Morse on bracketing the phenomenological perspective,Joan M.Anderson on developing theory inductively ethnography and epistemology - generating nursing theory, Agnes M.Aamodt on ethics and validity being a phenomenologicl researcher, Vangie Bergum on fieldwork in your own setting the use of self in ethnographic research, Julienne G.Lipson on nursing phenomena doing fieldwork in your own culture, Peggy Ann Field on the evolving nature of qualitative methods in nursing qualitative clinical nursing research when a community is the client, Judith A.Strasser on terminology strategies for sampling, Janice M.Morse on replicability are counting and coding acappella appropriate in qualitative research?, Phylis Noerager Stern on issues about reliability and validity issues of reliability and validity,Pamela J.Brink on interviewing interview techniques in qualitative research - concerns and challenges, Katharyn Antle May on the relationship between the researcher and the subject conducting qualitative studies with children and adolescents, Janet A.Deatrick and Sandra A.Faux on triangulation triangulation in qualitative research - issues in conceptual clarity and purpose, Kathleen A.Knalf and Bonnie J.Breitmayer,the granting game funding strategies for qualitative research, Toni Tripp-Reimer and Marlene Zichi Cohen on muddling methods institutional review of qualitative research proposals - a task of no small consequence, Patricia L.Munhall on the team approach feild research - a collaborative model for practice and research, Joyceen S.Boyle on teaching qualitative methods teaching qualitative research - perennial problems and possible solutions, Sally A.Hutchinson and Rodman B.Webb.

The Unstructured Interactive Interview: Issues of Reciprocity and Risks when Dealing with Sensitive Topics

Qualitative research using unstructured interviews is frequently reviewed by institutional review boards using criteria developed for biomedical research. Unlike biomedical studies, unstructured interactive interviews provide participants considerable control over the interview process, thereby creating a different risk profile. This article examines the interview process and literature for evidence of benefit and harm. Although there is evidence that qualitative interviews may cause some emotional distress, there is no indication that this distress is any greater than in everyday life or that it requires follow-up counseling, although the authors acknowledge distress is always a possibility. Essential to preventing participant distress is the researchers interviewing skills and a code of ethics. When research is conducted with sensitivity and guided by ethics, it becomes a process with benefits to both participants and researchers. The authors conclude that qualitative research using unstructured interviews poses no greater risk than everyday life and expedited reviews are sufficient.

Concepts of caring and caring as a concept.

If caring is to be retained as the “essence” of nursing, and if research in this area is to advance, then the various perspectives of caring must be clarified, the strengths and the limitations of these conceptualizations examined, and the applicability of caring as a concept and theory to the practice of nursing identified. Examination of the concept of caring resulted in the identification of five epistemological perspectives: caring as a human state, caring as a moral imperative or ideal, caring as an affect, caring as an interpersonal relationship, and caring as a nursing intervention. Two outcomes of caring were identified: caring as the subjective experience and as the physiologic responses in patients. The authors concluded that knowledge development related to caring in nursing is limited by the lack of refinement of caring theory, the lack of definitions of caring attributes, the neglect to examine caring from the dialectic perspective, and the focus of theorists and researchers on the nurse to the exclusion of the patient

Exploring the theoretical basis of nursing using advanced techniques of concept analysis.

In this article, the traditional methods of concept development are critiqued, and alternative methods that use qualitative methods of inquiry are presented. Variations of concept development techniques appropriate to the maturity of the concept being explored are then described, including methods for concept delineation, concept comparison, concept clarification, concept correction, and concept identification. To illustrate the application of concept development methods to nursing theory, a research program to delineate the construct of comfort is described.

Development of a Scale to Identify the Fall-Prone Patient

Patient falls are a serious problem, contributing to the morbidity and mortality of the elderly patient. This study reports on the development of the Morse Fall Scale. The scale consists of six scored items and discriminant analysis correctly classifies 80.5% of the patients. Validation of the scale by computer modeling was conducted. Data were randomly split and that analysis procedure repeated. Variables were obtained and weighted using half of these data, and these weights were tested on the remaining data. Similar results were obtained. Sensitivity of the scale was 78% and the positive predictive value, 10.3%. Conversely, specificity was 83% and the negative predictive value, 99.3%. Interrater reliability scores were r=.96. A prospective study in three clinical areas showed that the scale is sensitive to different patient conditions and to length of stay. Thus, the scale permits identification of the patient at risk of falling so that prevention strategies may be targeted to those individuals.

Toward a praxis theory of suffering.

This article revises and summarizes the major findings from a research program exploring the behavioral-experiential nature of suffering. Suffering is perceived as comprising two major behavioral states: enduring (in which emotions are suppressed; it is manifested as an emotionless state) and emotional suffering (an overt state of distress in which emotions are released). Individuals who are suffering move back and forth between these two states according to their own needs, their recognition/acknowledgment/acceptance of events, the context, and the needs and responses of others. Implications for the provision of comfort during suffering states are presented.

Aggregating Qualitative Findings: An Approach to Theory Development

The authors suggest that findings of independent, similar research articles may be aggregated into a cohesive study. Such a procedure greatly enhances the generalizability of the original studies and produces a relatively solid mid-range theory. In this article, the criteria for selecting studies, possible problems inherent in the aggregation approach, and potential areas for application are discussed.

Critical Analysis of Strategies for Determining Rigor in Qualitative Inquiry

Criteria for determining the trustworthiness of qualitative research were introduced by Guba and Lincoln in the 1980s when they replaced terminology for achieving rigor, reliability, validity, and generalizability with dependability, credibility, and transferability. Strategies for achieving trustworthiness were also introduced. This landmark contribution to qualitative research remains in use today, with only minor modifications in format. Despite the significance of this contribution over the past four decades, the strategies recommended to achieve trustworthiness have not been critically examined. Recommendations for where, why, and how to use these strategies have not been developed, and how well they achieve their intended goal has not been examined. We do not know, for example, what impact these strategies have on the completed research. In this article, I critique these strategies. I recommend that qualitative researchers return to the terminology of social sciences, using rigor, reliability, validity, and generalizability. I then make recommendations for the appropriate use of the strategies recommended to achieve rigor: prolonged engagement, persistent observation, and thick, rich description; inter-rater reliability, negative case analysis; peer review or debriefing; clarifying researcher bias; member checking; external audits; and triangulation.

Confusing Categories and Themes

Despite the fact that categories and themes are different—they are used for different purposes in the research process, are developed from data using different strategies, and capture different forms of knowledge—I see that the terms are sometimes used almost interchangeably in completed research. This matters for the conceptual and theoretical structure of the completed research, and if there is a lack of cohesion between the methods and the results, the study becomes less comprehensible. Basically, a category is a collection of similar data sorted into the same place, and this arrangement enables the researchers to identify and describe the characteristics of the category. This, in turn, enables the category itself to be defined, and then compared and contrasted with other categories, or if broad in scope, to be divided into smaller categories, and its parts identified and described. A theme, on the other hand, is a meaningful “essence” that runs through the data. Just as a theme in opera occurs over and over again, sometimes in the foreground, sometimes in the background, and sometimes co-occurring with other tunes, so does the theme in our research. It is the basic topic that the narrative is about, overall. This comparison of categories and themes becomes clearer if we carry our opera metaphor one step further. Once I heard on the radio a content analysis of an opera. The writer had sorted all the trills and all the “Ah-has” into separate categories, and the result was ludicrous. But this example makes the difference between a category and a theme immediately obvious. Now, categories are important for determining what is in the data (the “what”). So they are used in ethnography and in the initial analytic phase of grounded theory. The ultimate use of categories is in the development of a taxonomy, in which the researcher identifies relationships between categories and smaller units, or subcategories. In some models (and again grounded theory is a good example), categories are sorted in trajectories over time, with some categories (or forms of the category) preceding or following others. Pushing the analysis further, the researcher can then determine what triggers the change, and move into theory development. But themes are used in the later phase of grounded theory to tie it all together (as the core variable or Basic Social Process), and they are the basic strategy of analysis in phenomenology, where the purpose is to elicit meaning or the essence of the experience for the participant. Analytic strategies for categorizing and “themeing” differ. Categories are developed using content analysis, in which similar chunks of text are ordered or placed proximally. They are separated from the interview or document itself; they are positioned so that example after example of the same thing may be examined, and the major commonalties may be identified, coded, explicated, and described. If a category becomes large (i.e., it contains a lot of examples), it may be separated into smaller units or subcategories. To identify a theme, the researcher reads the interview or document paragraph by paragraph, asking, “What is this about?”, and thinking interpretively. Analytic strategies may ease the process. Using the computer program’s highlighting features, emphasize key words and phrases. Then, using the footnote feature of the word processing program, make footnotes about the major (and sometimes the minor) emphasis of each section of text. Then the “convert footnotes to endnotes” feature places all of the footnotes together, and enables the footnotes themselves to be examined as a whole. This eases the process of making the themes identifiable. If we compare these two strategies for creating a category and identifying a theme, a category may appear at one part of a process (or appear in different forms in different stages), while a theme should go right through the data. Can a theme also be a category? Let’s consider love as an example. Examining it as a category, we would have categories of instances of love, targets of love, and perhaps even instances of love scorned. We may have the data to identify the characteristics of love of money (greed), puppy love, being in love, love of work, or love of play. We may even be able to differentiate between affection and passion. Creating such definitions decontextualizes, strips the link to a particular individual, but