1 - Introduction
2 - Epistemological Bases of Quantitative and Qualitative Research
3 - Implications of adopting a Qualitative Framework for Research in the Human and Social Sciences
4 - How to Recognize a “Good” Qualitative Research Design
5 - Ethical Concerns in Qualitative Research
6 - Endnotes
7 - References
8 - Preparation and Acknowledgements
This paper has been prepared as a contribution to knowledge and understanding of qualitative research undertaken by some researchers in the disciplines of sociology, anthropology, women’s and gender studies, cultural studies, religious studies, history, education, psychology, nursing and traditional medicine, and alternative medicine. Qualitative research provides the opportunity to investigate and express the diversity and richness of social life in a manner that is different from the experimental, quasi-experimental, and survey approaches that have traditionally tended to be employed in quantitative research. Qualitative research is able to work with intricacies and differences of social meaning in ways that permit a nuanced and complex understanding of the reasons that social actors do the things that they do. Consequently, research conducted using a qualitative framework addresses different questions, uses different methods, analyses, and data, and attempts to draw different conclusions than research conducted from within a quantitative framework.
The aim of this paper is to promote an understanding of the epistemological bases, research methodologies, and concomitant ethical issues that arise from research conducted using a qualitative perspective. In order to achieve this aim, comparisons are drawn between quantitative and qualitative traditions in research. Two provisos to this comparison should be noted. First, it is not possible in a document of this format and length to outline all of the tensions, debates, and layers in the quantitative and qualitative research traditions. The intention of the paper, therefore, is to present a more general understanding of these two research traditions. Second, the comparison of quantitative and qualitative research traditions is in no way meant to be read as inferring that one research tradition is superior to the other. Rather, this paper would argue that both research traditions make important contributions to various bodies of knowledge, through their ability to pose and address different types of research questions.
The following issues are addressed in this paper: Epistemological bases of quantitative and qualitative research, implications of the adoption of a qualitative framework for research practice, how to recognize a “good” qualitative research proposal, and ethical concerns that could be addressed in a qualitative research proposal [1].
Quantitative research: A review
Many researchers in the various branches of the natural, physical, and human sciences have adopted positivism as the paradigm within which “good” research should be conducted [2]. The epistemological bases for positivism are objectivism and empiricism. The epistemology of empiricism holds that that there are in the world “real objects and events whose characters are entirely independent of any beliefs that might exist about them” (Buchler, 1955 in Rabinowitz & Sechzer, 1993, p. 29). The goal of conducting research within positivism is to produce an objective account of some behaviour or phenomenon, that is, an account that minimizes researcher bias as much as is possible. Within positivism, a scientific theory is defined as a set of concepts and propositions that describe and explain certain aspects of human experience. Scientific theories are public statements that indicate what a researcher believes to be true about his/her area of investigation, and must satisfy three criteria.
Firstly, a scientific theory must be parsimonious; that is, it must be able to explain a broad range of phenomena. Secondly, a scientific theory must be falsifiable. It must be able to make predictions about future events, so that the theory can be supported or discredited. Thirdly, a scientific theory must have heuristic value. It must build on existing knowledge by continuing to generate testable hypotheses that if confirmed, will lead to a deeper understanding of the behaviour or phenomena being investigated.
Quantitative research conducted within the epistemological framework provided by positivism follows the scientific method. When research adheres to the scientific method, it proceeds deductively, that is, following an hypothetico-deductive pathway [3]. That is, the researcher begins with a theoretical proposition, and from this proposition establishes hypotheses about what might be expected if the theoretical proposition is correct. These hypotheses attempt to predict causal relations between two or more phenomena. In order to test these hypotheses, researchers from different disciplines employ designs that involve the tight control of the phenomena being investigated. That is, all extraneous factors that might be affecting the phenomena under investigation must be virtually eliminated or controlled. Without this control, there is no way that the researcher can conclude that a causal relationship exists between the phenomena being investigated [4]. The aim of conducting research in such a manner is to be able to ensure objectivity and also to generalize the theoretical claims being tested. Generally, (although not always), the use of the scientific method means that research is quantitative. Within the social sciences, there is a sense among some researchers (e.g., Gergen, 2001; Phoenix, 1990; Rabinowitz & Szecher, 1993) that emphasis on the use of the scientific method as a framework for research has limited the scope of research activity, especially in terms of delineating a methodology to be adopted by research, and in minimizing the emphasis on the social, cultural, and historical context within which much of human behaviour occurs. These issues form the basis for the critique of empiricism, positivism, and the scientific method that has been offered by some researchers working within a qualitative framework.
Qualitative research: A review
Over the last decade, some researchers within the social sciences have expressed dissatisfaction with the scientific method as a means of both conducting research and generating knowledge(s). These researchers have argued that the aim of research practice should be to focus upon understanding the meaning that events have for the individuals being studied. The attempt to understand meaning is at the base of what has been termed qualitative research.
Challenges to mainstream empiricist approaches have come from several sources. Polkinghorne (1988, cited in Knapik, 2002, p. 2) argued that human science needed to focus on knowledge that “deepens and enlarges the understanding of human existence” [5], rather than focus on the goals of prediction and control of experimental models to remove all but one variable. More radical challenges to mainstream science came from within the feminist movement. For example, feminist standpoint theory challenged empiricism in that it “resisted the notion that mainstream scientists had the authority and the ability to reveal a single objective reality” (Knapik, 2002). Instead, research should focus on the individual’s lived experience, with self-disclosure as an invaluable source of knowledge. Researchers who operated from a post-modernist standpoint criticized modernist views about the nature of truth, knowledge, and language (Knapik, 2002).
Overall, then, the aim of most forms of qualitative research is to understand how the world is socially constructed by its participants, and what meanings those constructions have for the participants (McCleod, 2001). Therefore, the paradigmatic base underlining (most) qualitative research is phenomenology, naturalism, and/or hermeneutics [6], rather than positivism, which is at the base of (most) quantitative research. The notion of the world being constructed implies that individuals inhabit a social, personal and relational world that is complex, layered, and can be viewed from different perspectives. This social reality can be seen as multiply constructed. The world is constructed through language (stories, conversations), through systems of meaning, through memory, through rituals and institutions that have been created, and through the ways in which the world is physically and materially shaped (McLeod, 2001, p. 2).
In summary, (most) qualitative research has four general emphases in theoretical perspective and research methodology. These emphases differentiate it from quantitative research. First, there is a move away from the construct of scientific objectivity and the consequent focus on the need to acknowledge explicitly (rather than recognize and minimize) the researcher’s own biases in the conduct of research. Second, there is an attempt to conduct research in more naturalistic settings, away from the sometimes rigid and artificial controls imposed by the laboratory [7]. Third, the use of more naturalistic research settings encourages the attempt to understand phenomena as the participants under investigation understand them, being wary of the imposition of categories created by the researcher upon such phenomena. Fourth, there is the acknowledgment that all behaviour occurs within a social, political, and historical context, which in turn shapes behaviour, and the ways in which behaviours are understood. Data collected by researchers operating within a qualitative paradigm are consequently rich, have transferability, and contextual sensitivity.
There are many different forms of qualitative research that might be conducted by researchers (Knapik, 2002; Oberle, 2002). These include ethnography, grounded theory, ethnomethodology, and discourse analysis, among others. Moreover, the practice of qualitative inquiry does not centre solely upon the adoption of particular research methodologies and techniques. Rather, many different types of methodologies and techniques are available to researchers (Knapik, 2002; McCleod, 2001). Each type of qualitative research has its own theoretical perspective, its own approach to the gathering of data, the types of data that constitute viable areas for investigation, and the appropriate types of analyses for these data. For example, grounded theory is used to describe social processes and generate descriptive theory, ethnography describes culture (Oberle, 2002), and discourse analysis aims to uncover the shared, taken-for-granted assumptions that are used in conversation to negotiate power relations in day-to-day life (Wood & Kroger, 2000). There are basically three types of qualitative research, all of which aim to provide new ways of knowing the world. These types are knowledge of the other, knowledge of phenomena, and reflexive knowing (McCleod, 2001). These ways of knowing have been separated for analytical purposes, but often tend to appear together.
Knowledge of the Other
Knowledge of the other is generated by research that takes a category of person (such as student, psychotherapy patient, gang member), and seeks to describe, analyze and interpret the world-view of a sample of people who represent that category. This form of knowledge is highly useful. All individuals are socialized into stereotyped views of many groups, and may have little access to the groups to explore feelings and perceptions of these groups, and occasionally, alternate epistemological universes. Additionally, there may be barriers of class, gender, ethnicity, race, sexuality, and power that inhibit gaining a rich understanding of the world of others. Qualitative research, therefore, gives these groups of individuals a “voice,” and thus potentially empowers them. Gilligan’s work on female morality could be argued as an example of this type of qualitative research.
Knowledge of phenomena
Qualitative research that aims to develop knowledge of phenomena is focused upon categories of events, rather than upon groups of persons. This form of research aims to provide an in-depth understanding of a wide range of phenomena that might be considered psychologically significant, and often offers insight into process. An example of this type of research is an exploration of the transference process in a therapeutic situation.
Reflexive knowing
In research that aims for reflexive knowing, the researcher turns his/her attention to his/her own process of constructing a world, with the aim of saying something new about that world, or increasing awareness of negotiated meaning construction. Reflexive research generally focuses on the use of language, by analyzing assumptions that underpin many shared, common understandings of the way in which the world works. Many researchers within this tradition use terms such as “deconstruction” and “critical” to describe what they are attempting to do. The intention of this research is to subvert everyday ways of seeing the world.
Clearly, the adoption of a qualitative framework for the guiding of research practice has implications for the types of methods that are used to collect data and who might serve as participants in research. Each of these issues will be discussed.
Research questions, methods and research design
The epistemologies of the various forms of qualitative research have four implications for research questions, methods, and designs employed by qualitative researchers. First, the research design tends to emerge as the study proceeds, as the researcher responds to the participants and shifts the original research question as information emerges from participants in the course of the research (Oberle, 2002). Therefore, the question in which the researcher is interested may not emerge until the study is half underway (Oberle, 2002). Consequently, it is more appropriate in qualitative research to describe a research problem, rather than pose a research aim or research question.
Second, the social research methods and techniques used by researchers working within this paradigm are necessarily diverse, seeking to capture a wide array of social phenomena. Research methods and techniques may range from the close analysis of documents, such as photographs, letters, or diaries; the observation of teacher-student interaction; the active participation of the researcher in a “gang” or community project; surveys, interviewing, naturalistic observations, to the collection of written or oral responses from a large population.
Third, qualitative research may use different approaches to the conceptual organization of research data, such as biography, autobiography, oral history, life history, longitudinal studies, and the case study. Each of these approaches tends to be associated with specific research techniques and specific philosophical approaches to methods. For example, oral histories are often compiled with the use of in-depth interviews and are often used in feminist research to express the experiences of less powerful members of society, which are not necessarily followed in all cases. There is no a priori relationship between research strategies, techniques, philosophies and methods. Indeed, a rigid expectation of how this relationship should operate would undermine creativity and the possibility of good research design (Collier, Dunlop, Kippax, & Yeatman, 1994).
Fourth, analysis of data generated by qualitative research may involve the review of transcribed interviews, conversational gatherings, or other scripts (e.g., autobiographical accounts), and descriptions of field observations, field notes, and other documents, in search for patterns, meanings, or taken-for-granted assumptions about the ways in which the world works (e.g., discourses) (Oberle, 2002). Quite the contrary, it is the specificity of the circumstances that may be desirable. Consequently, much qualitative research is descriptive, and generative of theory. There is no attempt to draw statistical inferences, or to generalize the results of the research. Therefore, attempts to establish reliability and validity of qualitative research are largely without meaning. Instead, the results of the research should be trustworthy, dependable, and transferable (Oberle, 2002).
Participants
The issue of inclusive sampling, which has become a concern among members of Research Ethics Boards must be understood as having a different relevance and importance when the proposed research is conducted from within a qualitative framework (Collier et al, 1994; Knapik, 2002; Oberle, 2002). There is an accepted tradition within qualitative research that allows for purposeful sampling, that is, the identification of a sample group that reflects the phenomenon being investigated. The decision to use purposeful sampling extends from a desire to gather information rich cases from which to draw data, with the intent to uncover other possible instances where similar circumstances may prevail and to where the experiences of others may be applicable. Moreover, many researchers who operate within a qualitative framework conduct research with members of society whose voices have been under-represented, or are less powerful. Conversely, they may also attempt to indicate means by which elite groups can avoid scrutiny. Consequently, the samples used in qualitative research will often not be inclusive, and this lack of inclusiveness should not be taken as detracting from the value of the research, especially as generalizability of research findings is not a concern within this research tradition (Collier et al, 1994; Knapik, 2002; Oberle, 2002).
In terms of sampling, not only is a purposeful sample (sometimes) required for qualitative research, the exact number of participants required for a study often cannot be specified before the study is conducted (Oberle, 2002). The inability to specify the exact number of required participants for a study stems from the limitation posed by theoretical saturation, that is, the point in the research beyond which no new concepts, themes, or discursive constructions emerge from the analysis (Oberle, 2002). This point cannot be specified a priori [8], in contrast to research designs that use quantitative methodologies, where the number of participants required for a study are specified before the research is conducted to ensure power of the data analysis and external validity of the research, neither of which are concerns in qualitative research.
A list of questions have been framed by Collier et al (1994) to provide a guideline for the assessment of qualitative research proposals [9]. These questions include the following:
- Is the research problem described?
- What is an appropriate theoretical framework to adopt in order to address the research problem?
- What methods will be used and how will they address the research problem and overall approach?
- What form of analysis is envisaged?
- How will evidence for conclusions be demonstrated?
- How will the results be integrated into a body of scholarship?
- Have the ethical problems that the study represents to the participants as well as the community at large been identified? What safeguards have been built into the research process to minimize these? (Ethical problems common to qualitative research, and suggestions for solutions to these problems are provided in the subsequent section of this document).
- Is the research design appropriate to the research problem?
- What problems are envisaged with accessing the data and how might these be overcome?
- Is there a logically coherent relationship between the research problem, the research methodology, and the likely outcomes of the research?
It has often been an assumption that because qualitative research does not involve physical manipulation or intrusive procedures, it poses less than minimal risks to participants. However, there are three areas of potential harms to participants that must be addressed in any qualitative research proposal: assessing harms and benefits, informed consent, and privacy and confidentiality.
Assessing harms and benefits
In research involving in-depth interviews or focus group discussions on sensitive issues, the researcher often develops a relationship of trust with his/her participants. This relationship may be misinterpreted by the participant as a therapist-client relationship. Moreover, interpretation of the data collected in the course of qualitative inquiry involves having one’s life categorized by an “authority” and in some forms of qualitative research, reframed into larger theoretical constructs (Oberle, 2002). This process may mean that interpretations are unflattering or damaging to participants.
Solutions
Oberle (2004) suggests that in order to address these issues, the Participant Information Letter and Consent Form should contain the following statements, where necessary:
- Interviews/focus group discussions are not meant to be therapeutic;
- Data obtained in the course of the study will be reframed into a larger context, and the data must accurately reflect what interviewees reveal, even if reports are unflattering to participants [10];
- Checks may be built into the research process, so that the researcher can check any interpretations that have been made with the participant;
- A plan outlining how the research will proceed if, in (3), the participant disagrees with the interpretation made by the researcher;
- If any problems arise for the participant during the course of the interview or focus group discussion, what provisions have been made for counselling and support.
Informed Consent
Risks to individuals from participating in qualitative research may not be able to be anticipated when the method and research questions are evolving throughout the research. It is important, therefore, that the possibility of these unanticipated risks is made clear to potential research participants.
Solutions
Oberle (2004) suggests that in order to address this issue, the Participant Information Letter and Consent Form should contain the following statements, where necessary:
- That participation in the research may have unanticipated consequences;
- That consent to participate in the research will be sought at various points throughout the research process; and how often that consent will be sought;
- What will be done with the data if the participant decides to withdraw from the study.
Privacy and Confidentiality
Qualitative research often generates questions about the ability to protect confidentiality (Oberle, 2002). These questions arise particularly when publication of the results of qualitative research include lengthy quotations from participants. Even when names and other identifiers are excluded in the data collected in qualitative research, subtle cues might enable those familiar with the research context to identify the participant (very much a problem in research conducted in small communities) (Knapik, 2002; Oberle, 2002) [11].
Solutions
Oberle (2004) suggests that in order to address these issues, the Participant Information Letter and Consent Form should contain the following statements, where necessary:
- Assurance of confidentiality of the information;
- Researcher’s intentions for data processing and storing;
- That pseudonyms for participants will be used, and how these will be selected;
- What demographic data will be reported, so that participants may judge for themselves whether or not they may be identifiable by this data, allowing them to make an informed decision to continue as participants;
- Any plans to air video or audio-tapes in public, or place materials on the Internet;
- An offer to allow participants to view any material intended for public broadcast before it is broadcast.
[1] The ethical concerns that are currently argued as needing to be addressed by qualitative research proposals are themselves representative of an ethical standard, which presently is established by the Tri-Council Policy Statement. The Tri-Council Policy Statement is based upon a biomedical (and therefore, experimental) model of research, which has its own unique ethical limitations (Krautwurst, personal communication, 2005). A discussion of these limitations is beyond the scope of the present document.
[2] Many researchers in the physical, chemical, and biological sciences who use quantitative methods do not subscribe to positivism as an underlying philosophy. According to Yu (2001) this misconception that quantitative research is based upon positivism derives from two misunderstandings: first, the oversimplification of quantitative research as hypothesis testing, and second, the failure to realize that quantitative research has been shaped by more than one philosophy.
[3] Scientific research does not always proceed using the scientific method, that is, following an hypothetico-deductive pathway. Indeed, many traditional physical sciences (such as biology, zoology, astronomy, physics, medicine, and pathology) will use a descriptive methodology in research before attempting to frame theory and hypotheses, and testing these. In descriptive methodology, observations of some phenomenon are made, the phenomenon is described, categorization (description of groups based on similarities and differences) is made, leading, in turn, to the identification of specific relationships or cause-and-effect chains. Once this methodology has been followed, hypotheses can be generated, which can then be tested, leading to verification (Horney, personal communication, 2005).
[4] The attempt to establish causal relationships among variables using the scientific method is an approach used traditionally in many sub-disciplines of psychology, and not in all natural and social disciplines (Richards, personal communication, 2005).
[5] It has been suggested that this criticism of quantitative research is problematic, as much of quantitative research (and positivism) never claimed to be concerned with phenomenal states (Murray, 2004, personal communication).
[6] The term phenomenology is not used in all disciplines. Some disciplines (e.g., nursing) prefer the use of the term naturalism (Bryanton, 2004, personal communication). Hermeneutics is also an epistemological base for some forms of qualitative research (Krautwurst, personal communication, 2005).
[7] Not all quantitative research is conducted within a laboratory setting. This point, then, relates only to those disciplines in which research has attempted to impose “rigid and artificial controls” of a laboratory setting (Richards, personal communication, 2005).
[8] It has been suggested by researchers working from with the grounded theory approach in qualitative research that theoretical saturation is reached usually with a sample of ten to twelve participants (Glaser & Strauss, 1967).
[9] It has been suggested that the major focus of a Research Ethics Board should be on point 7, instead of on points 1 through 6, and points 8 through 10. Point 7 is related directly to ethical concerns raised by the research, whereas points 1 through 6 and 8 through 10 should be dealt with by researchers and their advisors and committee members within their disciplines (Hopkirk, personal communication, 2005).
[10] Points 1 and 2 should also take into account the nature of the research. Although guidelines are important, each research ethics proposal must be handled situationally (Krautwurst, personal communication, 2005).
[11] In some research situations, participants may wish to be named in the work that results from the research procedure. This desire raises the issue of the researcher’s ethical obligation to name those who wish to be named. This issue is one that may need to be addressed by Research Ethics Boards in the future (Krautwurst, personal communication, 2005).
Collier, F., Dunlop, R., Kippax, S., & Yeatman, A. (1994). Notes on qualitative social research for the Macquarie University Research Committee. North Ryde, Australia: Macquarie University.
Gergen, M (2001). Feminist reconstructions in psychology: Narrative, gender, and performance. Thousand Oaks: Sage Publications.
Glaser, B.G & Strauss, A.L. (1967). The discovery of grounded theory: Strategies for qualitative research. Chicago, Aldine Publishing Company
Knapik, M. (2002). Ethics in qualitative research: Searching for practice guidelines. Paper presented at the University of Calgary, centre for Leadership in learning Conference, Calgary, Alberta, July 6, 2002.
McLeod, J. (2001). Qualitative research in counselling and psychotherapy. London: Sage.
Oberle, K. (2002). Ethics in qualitative health research. Annals RCPSC, 35(8), 563-566.
Phoenix, A. (1990). Social research in the context of feminist psychology. In E. Burman (Ed.), Feminists and psychological practice. (pp. 89 - 103). London: Sage Publications.
Rabinowitz, V.C., & Sechzer, J. A. (1993). Feminist perspectives on research methods. In F.L. Denmark, & M. A. Paludi (Eds.), Psychology of women: A handbook of issues and theories. (pp. 23-66). Westport, CT: Greenwood Press.
Wood, L.A. & Kroger, R.O. (2000). Doing discourse analysis: Methods for studying action in talk and text. Thousand Oaks: Sage Publications.
Yu, C.H. (2002). Misconceived relationships between logical positivism and quantitative research. Paper presented at the Annual Meeting of the 2001 American Educational research Association, Seattle, WA.
Preparation and Acknowledgements
This document was prepared by Dr Fiona Papps (2005), of the Department of Psychology at the University of Prince Edward Island. Suggestions and critical and editorial comments were contributed by:
