Qualitative research in nutrition and dietetics: getting started
Dr J. A. Swift, Division of Nutritional Sciences, School of Biosciences, The University of Nottingham, Sutton Bonington Campus, Loughborough, Leicestershire LE12 5RD, UK.
Tel.: +44 (0)115 951 6178
Fax: +44 (0)115 951 6122
Qualitative research is well placed to answer complex questions about food-related behaviour because it investigates how and why individuals act in certain ways. The field of qualitative health research is undoubtedly gaining momentum and, increasingly, there is a recognition that it should be a vital part of the decision-making processes that direct the development of health policy and practice. Much of the guidance available, however, is difficult to navigate for those new to ‘qualitative research’, and there is little discussion of qualitative research issues specifically in relation to nutrition and dietetics. This review, the first in a series, outlines the field of qualitative enquiry, its potential usefulness in nutrition and dietetics, and how to embark upon this type of research. Furthermore, it describes a process to guide high-quality qualitative research in this area that proceeds from the research question(s) and considers the key philosophical assumptions about ontology, epistemology and methodology that underpin the overall design of a study. Other reviews in this series provide an overview of the principal techniques of data collection and sampling, data analysis, and quality assessment of qualitative work, and provide some practical advice relevant to nutrition and dietetics, along with glossaries of key terms.
Despite claims of a qualitative revolution in some disciplines (Denzin & Lincoln, 2005), health research, including research in nutrition and dietetics, continues to be dominated by quantitative approaches (Fade, 2003; Broom & Willis, 2007). Clearly more needs to be done to engage health researchers such as dietitians and nutritionists with qualitative research. Although there are many excellent texts and dedicated journals available, those new to this type of research can find the literature and qualitative terminology impenetrable. This review is the first in a series that seeks to demystify qualitative research for those who feel more comfortable with conducting and evaluating quantitative research, or are inexperienced in using qualitative approaches. In addition, qualitative research, on the whole, is not produced by passively following predefined ‘recipes’. It can be more aptly considered as an adventure; exciting, challenging and a little unsettling (Willig, 2008). This series will not therefore provide an exhaustive how-to guide to conducting or evaluating qualitative research. Instead, it aims to empower those dietitians and nutritionists who wish to delve deeper.
The present review aims to discuss what qualitative research has to offer the field of nutrition and dietetics. Furthermore, it takes a detailed look at what is meant by the term ‘qualitative research’, and offers advice on how to develop a qualitative research strategy from a research question relating to nutrition and dietetics. Overall, this review presents a model for embarking on qualitative research in this area. Other reviews in this series will provide an overview of the principal techniques of data collection and sampling that may be used for qualitative research in nutrition and dietetics (Draper & Swift, 2010), describe a number of techniques and practical steps that can be taken to provide some structure and focus to the intellectual work of qualitative analysis (Fade & Swift, 2010), and introduce some of the key scientific debates around quality criteria that can be applied to qualitative research (Pilnick & Swift, 2010). In addition, all reviews in the series provide some practical advice for those wishing to engage with qualitative research in nutrition and dietetics and a glossary of key terms.
What does qualitative research have to offer nutrition and dietetics?
‘Food [is] a complex entity, replete with contradictions and oppositions, full of confusions and a potential source of anxiety, particularly in relation to health’
(Chamberlain, 2004: 468)
Perhaps the most basic approach to the study of food is to consider it simply as a source of energy and nutrients: a fuel for living. However, as Chamberlain eloquently points out, this approach is very limiting; food does more than fulfil a biological need (Chamberlain, 2004). Food can be a source of great pleasure, although, for many, it can also cause guilt and anxiety (Coveney, 2006). Food can be used to help cope with stressful situations and negative emotional states (Conner & Armitage, 2002). Food also plays a vital role in how humans communicate within their social world. Food might be used to establish affiliation to particular groups and engender peer acceptance, to express love, sociability and hospitality, to demonstrate civilisation and self-control, and to express power within relationships and social status (Fieldhouse, 1996; Ogden, 2003; Coveney, 2006). To borrow terms from sociology and psychology, it could be said that food is
‘… constructed, negotiated, socialized and contextualised’ (Chamberlain, 2004: 468)
Dietitians work at the heart of this complexity. They are required to bridge the gap between logical, unambiguous scientific data and the complicated, multiple functions that food has in their clients’ lives. For example when working with a client with obesity, a dietitian will not simply advise the client to create a state of negative energy balance by reducing the amount of calories they consume, and by increasing his/her physical activity levels. They might also deal with how the client could achieve weight loss at the same time as continuing to dine out with friends or how to deal with stressful situations without ‘comfort’ eating.
Qualitative research is particularly relevant for health research that investigates human behaviour and relationships, as described above, because it investigates how and why people behave in certain ways. Much of the work conducted in dietetics and nutrition health promotion is aimed at changing people’s eating behaviour to improve their health. Qualitative research that investigates how and why people eat in certain ways therefore appears ideally placed to support this work. The research questions amenable to qualitative research are numerous and diverse. Examples include ‘Why do some clients with renal disease fail to adhere to a low-potassium diet?’ or ‘How do mothers on low incomes feel about buying economy-brand food?’ It is also important to recognise that qualitative research in nutrition and dietetics is not only limited to eating behaviour. Equally, it can be used to investigate how and why healthcare professionals behave in particular ways in practice; for example, ‘How do dietitians deal with aggressive clients during consultations?’ or ‘Why do nurses fail to feed in-patients who are unable to feed themselves?’ Furthermore, there are applications in education and training such as ‘What are the training needs of community food workers employed in Children’s Centres?’ or ‘How do experienced dietitians mentor and support newly-qualified staff?’
At the heart of questions about the usefulness of research evidence is the issue of generalisability (Morse, 2008). The type of evidence produced by most qualitative research (described in more detail below) is not empirically generalisable (i.e. the findings cannot be used to infer about the characteristics of a wider population) (Mason, 2002). This situation may have lead some to question the role of qualitative research in evidence-based practice. Qualitative research evidence can, however, be theoretically generalisable (i.e. the findings can be used to develop concepts, understand phenomena and theoretical propositions that are relevant to other settings and other groups of individuals) (Draper, 2004; Yardley, 2008).
Although health research continues to be dominated by quantitative approaches (Broom & Willis, 2007), the field of qualitative health research is undoubtedly gaining momentum. The National Institute for Health Research’s Centre for Reviews and Dissemination (CRD) has recently dedicated a whole chapter of their guidance on undertaking systematic reviews in health care, to the role of qualitative evidence (Centre for Reviews and Dissemination, 2009). The CRD’s position is that qualitative research helps to ensure that reviews of quantitative evidence are of maximum value in the decision-making processes that guide the development of health policy and practice by:
- • shaping questions of importance to end users,
- • investigating the mechanisms behind (in)effectiveness,
- • explaining heterogeneous results,
- • identifying factors that impact on implementation of interventions,
- • describing the experiences of people receiving the intervention,
- • detailing participants’ subjective evaluations of outcomes (Centre for Reviews and Dissemination, 2009).
This is quite a departure from the traditional view that the only useful evidence regarding effectiveness comes from randomised controlled trials (Victora et al., 2004) and qualitative research being consigned to the lowest level of evidence hierarchies, alongside consensus and opinion (Grypdonck, 2006).
In the field of nutrition and dietetics, the Food Standards Agency has also demonstrated their commitment to qualitative research by commissioning a number of qualitative projects that investigate, for example, people’s use of food labelling information (Food Standards Agency, 2010a) and public understanding of sodium and salt labelling (Food Standards Agency, 2010b).
What is qualitative research?
As previously stated, qualitative research investigates how and why people behave in certain ways. Although this is a useful position to start from, it obscures the diversity of this area. Qualitative research might be best viewed as an umbrella term; it encompasses a vast range of approaches and techniques, many of which have developed from different disciplines, such as anthropology, philosophy, sociology and psychology (Giles, 2002; Denzin & Lincoln, 2005). In addition, the scope of qualitative research is continually evolving (Lincoln & Denzin, 2005); for example, the recent development of Q methodology (e.g. van Exel & de Graaf, 2005). As such, there is no single, universally accepted, precise definition of qualitative research (Flick, 2007).
To begin to understand what qualitative research is, it is important to first consider the philosophical assumptions that underpin it. This may appear to be an unnecessary diversion (after all, researchers are practical animals, they do research) but an awareness of these assumptions will ensure that choices about methods of data collection and analysis can be fully justified and answer the study’s research question(s) (Carter & Little, 2007). Ultimately, failure to understand these assumptions may result in poor quality research (Yardley, 2006). Creswell (1998) suggests that there are three main assumptions that underpin qualitative research relating to: (i) ontology; (ii) epistemology; and (iii) methodology. He proposes that, broadly speaking, qualitative research tends to: (i) have a relativist ontology; (ii) be located in a epistemology that embraces subjectivity; and (iii) favour research designs based on inductive reasoning (Creswell, 1998). Because these philosophical assumptions may be unfamiliar to readers, a detailed description is offered below.
Ontology: what is there to know?
Ontology relates to a person’s understanding of the nature of the world. Typical responses to the question ‘why engage in research?’ might be; to search for knowledge, to establish facts, to systematically investigate whether particular ideas about the world are true or false. All of these responses assume that there is a single, stable reality ‘out there’ waiting to be understood and that it can be understood if it is subject to rigorous investigations. Researchers who subscribe to a single, stable reality can be described as having a realist ontological position (Willig, 2008).
By contrast, a relativist ontology does not rely on there being a single, stable reality ‘out there’ waiting to be discovered but considers reality to be socially constructed. This means that how the world is perceived and a person’s thoughts about it are always influenced by social factors such as culture, history and language (Willig, 2008). No matter how controlled research techniques are, there will always be bias because they are conducted and interpreted by human beings, and human beings are incapable of being completely objective. Researchers inevitably have a relationship with the subject they are studying; they choose the research questions, decide what to measure, how to measure, when to measure, how to interpret and how to communicate findings. Imagine a 100-mL flask containing 50 mL of water – the flask could equally be described as half full or half empty. Qualitative researchers tend to be reflexive (i.e. explicit about their role and its impact upon the research that they undertake). Researchers who consider reality to be socially constructed can be described as having a relativist ontological position.
Epistemology: how can people know?
Just as it is possible to have different ideas about what there is to know (i.e. different ontological positions), there are also different ideas about how knowledge can be produced (i.e. different epistemic positions). Epistemology refers to a branch of philosophy concerned with the theory of knowledge. As described above, an important characteristic of a socially constructed reality is the recognition that researchers inevitably have a relationship with the subject they are studying that makes objective measurements impossible. A researcher with a relativist ontological position might therefore hold an epistemic position in which he/she does not attempt to produce objective findings but instead embraces subjectivity (Ezzy, 2001). So rather than attempting to filter out and control sources of bias, such as the gender and professional background of the researcher, these would be reflected upon and welcomed as adding vitally important context. The qualitative researcher therefore needs to be reflexive; to recognise his/her role in the research process via critical self-scrutiny (Mason, 2002; Draper & Swift, 2010). Studying social phenomena in natural, real-life settings rather than experimental, ideal situations helps researchers claim that they are being true to the nature of the phenomenon under investigation (Bryman, 2004). Although it adds to the richness of the data, acknowledging and working with subjectivity can lead to researchers being accused of being unscientific and anecdotal (Abusabha & Woelfel, 2003). It is, however, strongly argued that the quality of qualitative research should be critiqued as rigorously as any other kind of research, albeit against appropriate criteria (Seale, 1999; Pilnick & Swift, 2010).
Methodology: how should research proceed?
Knowledge production in health research has been dominated by deductive reasoning (Broom & Willis, 2007) and the traditional process is to consult an established theory regarding the phenomena of interest, in order to develop a clear hypothesis. This hypothesis is then tested by collecting data from which conclusions regarding the veracity of the theory are inferred; a process called hypothetical-deductivism or deductive reasoning. The conclusions are considered to be undeniable if the data are derived using objective, unbiased methods. To do this, the researcher is required to position his or herself ‘outside of’ the subject matter and employs a range of techniques to ensure that he or she does not inadvertently influence the results obtained. Techniques include the use of reliable and valid measurement devices, standardised protocols, and double-blind randomised controlled trial study designs. However, as noted above, not all researchers aim to produce objective findings about a single, stable reality. Those who consider reality to be socially constructed, and aim to produce subjective findings, favour the process of inductive reasoning. This refers to findings that are derived from the data itself and contrasts with deduction in which research hypotheses are usually applied to the data and the data are used to confirm or reject the hypotheses. Instead of collecting data to test specific hypotheses, data collection and analysis is guided by one or more, open-ended research questions (Willig, 2008). The important distinction is that research questions direct the study rather than predict what may be found. For example, some researchers will ask how something happens and aim to produce detailed descriptions of participants’ feelings and experiences, whereas others will also ask why something happens and aim to produce explanations or arguments (Fade & Swift, 2010). Although attitudes and behaviour can also legitimately be investigated using quantitative research methods, such as surveys, psychometric scales and statistical analysis, numbers are unlikely to produce data that provide detailed descriptions or generate explanations. Qualitative research therefore normally deals with words; indeed qualitative research is often defined as ‘research beyond numbers’ (Greenhalgh & Taylor, 1997).
As noted above, Creswell describes qualitative research as tending to recognise multiple realities (relativist ontology), embrace subjectivity and favour inductive reasoning (Creswell, 1998). Qualitative research is therefore often presented as opposite to quantitative research (e.g. Abusabha & Woelfel, 2003). Although the distinction between quantitative and qualitative research is a convenient device for introductory texts, it has been criticised as overly simplistic and misleading (e.g. Green & Thorogood, 2004; Rolfe, 2006; Holliday, 2007). Qualitative researchers can vary in their ontological, epistemological and methodological positions (Willig, 2008). For example, a researcher conducting Grounded Theory can take either a realist approach in which they attempt to capture social reality, or a relativist approach, which views the analysis as a social construction of reality (Willig, 2008).
Qualitative strategies of inquiry
Many qualitative researchers work within a particular strategy of inquiry (also termed tradition of inquiry or qualitative methodology). Strategies of inquiry vary in the extent to which they promote the use of particular research methods (i.e. the techniques for collecting and analysing data) but all provide the researcher with an overall strategy for formulating, articulating, analysing and evaluating their methods (Carter & Little, 2007). Working within different strategies of inquiry will produce different findings because they act as a ‘lens’ by which a researcher discovers a particular aspect of the phenomena (Wu & Volker, 2009).
The most commonly used strategies of inquiry in health research include Phenomenology, Grounded Theory, Discourse Analysis, Ethnography, Ethnomethodology and Action Research (see Glossary) (Dew, 2007). There are, however, many different strategies, of which there are many different variations (Denzin & Lincoln, 2005) and, unfortunately, no single categorisation system is definitive (Madill & Gough, 2008). This diversity offers range and flexibility when designing a research project, although may equally be overwhelming (Easterby-Smith et al., 2008), particularly for novice researchers.
So how should a researcher go about choosing a particular strategy of inquiry? There are a number of approaches to take; for example, the researcher might be driven by his/her disciplinary background or the aim of the study might be to develop a theory from the data (Giles, 2008). A common approach is to proceed from the research question (Madill & Gough, 2008) and this flexible method appears to be particularly relevant for health research. It is beyond the scope of the present review to describe all possible strategies of inquiry; indeed, whole texts are devoted to this enterprise (e.g. Denzin & Lincoln, 2005; Stainton-Rogers & Willig, 2008). In addition, articles have detailed the most common strategies in qualitative health research (e.g. Dew, 2007), whereas others have discussed them in relation to human nutrition and dietetics (e.g. Harris et al., 2009). Instead, the present review illustrates the process by which a researcher might choose a particular strategy of inquiry and ensures that the choices about methods of data collection and analysis can be fully justified and do not clash with the assumptions underpinning the overall design of a study. This process should be considered explicitly because it impacts on the quality of the research produced (Carter & Little, 2007; Dew, 2007) and qualitative research is often criticised for lack of rigour (Maggs-Rapport, 2001). Table 1 describes the process using a hypothetical research question related to nutrition and dietetics.
Table 1. Developing a qualitative research strategy from a research question: an example from nutrition and dietetics
|Research question||The starting point is the research question, which in this hypothetical example is ‘What is the experience of obese adolescents who access adult weight management services?’|
|Ontological position||The researcher, in this case, a National Health Service dietitian, might assume a relativist ontological position. This means that he/she accepts that the experience of obesity in this population is socially constructed (i.e. is influenced by factors such as the individual’s culture and family history). Other considerations would be the concerns and background of the researcher. In this case the dietitian’s professional position and work context will impact on how the data are approached and analysed (e.g. the data may be analysed with a view to developing new services for the population of interest)|
|Epistemological position||In keeping with his/her relativist ontological position, the researcher might hold an epistemic position in which he/she does not attempt to produce objective findings but, instead, embraces subjectivity. He/she may view subjectivity as being crucial to this particular piece of research because it acknowledges that different individuals will experience the ‘same’ environment differently. For example, one participant may find adult services intimidating and the advice given irrelevant, whereas another participant might find such services self affirming because it validates his/her burgeoning status as an adult. Correspondingly, the researcher would reflect on factors that might influence the participants’ experience and the impact of the researcher’s own views and values, as outlined above|
|Methodological position||Because the researcher considers reality to be socially constructed and aims to produce subjective findings, he/she favours the process of inductive reasoning. Rather than setting up a series of hypotheses, the research is guided by an open-ended research question. In this way, the researcher does not make any predictions about the experiences of participants, instead he/she uses the data to generate explanations|
|Qualitative strategy of inquiry||The research question focuses on the examination of experience itself, which suggests that a phenomenological strategy of inquiry is appropriate. The philosophical branch of phenomenology (see Husserl) suggests that experience is separate from reality and is concerned with individual meanings and interpretations, the so-called lifeworld of an individual (Ashworth, 2008). An appropriate qualitative technique would therefore be Interpretative Phenomenological Analysis (IPA) (Fade, 2004; Smith et al., 2009). IPA is rooted in phenomenology, that is, it is concerned with individuals’ lived experience and how they make sense of that experience. It adopts a double hermeneutic position, which means that it foregrounds the reflexive position of the researcher making sense of the participant’s experiences|
The approach of proceeding from the research question, however, requires the researcher to critically appraise the numerous strategies available and choose the ‘best fit’ (Wu & Volker, 2009). Qualitative health researchers are not, however, always experts in qualitative strategies of inquiry (Willis et al., 2007; Harris et al., 2009). This can lead to method slurring where a researcher claims to adhere to a particular strategy of but, in practice, fails to do so (Holloway, 1997). Some studies have suggested that a seasoned qualitative researcher should be included as part of the research team (e.g. Willis et al., 2007; Harris et al., 2009). Although apprenticeship can play an important role in learning about qualitative research (Robson, 1993), experienced qualitative researchers can passionately advocate particular approaches. Concerns about quality may lead novice researchers to conform to recommended and/or highly specified strategies of inquiry (Lambert, 2006) without fully appreciating the philosophical assumptions that underpin it. It is therefore crucial that anyone embarking on qualitative research – novice and experienced researchers alike – thinks independently about his or her own views and develops an epistemological position from which to ground the most appropriate strategy of enquiry.
It is also worth considering that some researchers argue that qualitative research does not have to be conducted within a particular, highly specified strategy of inquiry (e.g. Avis, 2003; Braun & Clarke, 2006). This is relevant in the field of health research where researchers often cross disciplinary boundaries and where research is applied, making it useful to have the flexibility to select from a range of appropriate research methods (Braun & Clarke, 2006). This is compatible with a pragmatic approach; for example, the promotion of qualitative researchers as bricoleurs (i.e. multiskilled individuals who assemble the tools and materials necessary to get the job done) (Denzin & Lincoln, 2005). Flexibility does not, however, excuse an ad hoc approach to the research process (Mason, 2002). As always, choices about research methods must be appropriate for the research question(s) and must not clash with the assumptions underpinning the overall design of a study. Flexibility also does not excuse poor reporting of theoretical issues.
Mixing qualitative and quantitative research
Although some researchers will champion exclusively quantitative or qualitative research, a less divisive and more pragmatic stance is that neither is superior and that, depending on the research question, both quantitative and qualitative research methods could appropriately be employed; an approach termed mixed methods (Broom & Willis, 2007). Increasingly, health researchers are taking a ‘soft’ constructivist approach by acknowledging both the subjective and purely physical aspects of disease and illness, and are combining quantitative and qualitative methods (Broom & Willis, 2007). Although this may appear to be a neat way to navigate between two different views of research, mixed methods research is not without its critics. Those engaging in such work must be able to appreciate and bring together the assumptions underpinning both qualitative and quantitative research (Yardley & Bishop, 2007).
|Action research||Both research and intervention. The goals of action researchers are to understand something about the social world and change it (Dew, 2007). Takes into account the characteristics of the community, culture, researcher and research methods when evaluating the intervention’s impact (Harris et al., 2009)|
|Discourse analysis||Focuses on features of language, styles of argumentation and the way language is used to represent a particular phenomenon or issue (Dew, 2007)|
|Empirical generalisability||The extent to which research findings can be used to infer about the characteristics of a wider population (Mason, 2002)|
|Epistemology||The branch of philosophy concerned with the theory of knowledge (Carter & Little, 2007). Epistemic position refers to adherence to a particular idea about how researchers can produce knowledge|
|Ethnography||Derived from anthropology and focuses on cultural beliefs and practices. It typically involves close observation and often participation in the social life of the group being researched, often referred to as fieldwork (Dew, 2007)|
|Ethnomethodology||Derived from sociology and focuses on social practices. The primary question for ethnomethodology is ‘how is the organisation of society achieved?’ (Dew, 2007)|
|Grounded theory||Derived from sociology and focuses on generating theory that explains the data (Dew, 2007)|
|Hypothetico-deductivism||The process by which a hypothesis is tested by collecting data from which conclusions regarding the veracity of the theory are inferred|
|Inductive reasoning||The process by which research findings are derived from the data itself with no consideration to pre-existing theory|
|Methodology||Theories about how research should proceed (Carter & Little, 2007)|
|Naturalistic||Being true to the nature of the phenomenon under investigation (Bryman, 2004)|
|Ontology||Relates to people's understanding of the nature of the world (Willig, 2008)|
|Phenomenology||Derived from philosophy and seeks to capture phenomena as they are immediately experienced (Dew, 2007)|
|Q methodology||A method applied to the study of subjectivity, including a person’s viewpoint, opinion, beliefs and attitude. Involves ranking of statements, the results of which are subject to factor analysis (van Exel & de Graaf, 2005)|
|Realism||Belief in a single, stable reality ‘out there’ waiting to be discovered (Willig, 2008)|
|Relativism||Belief in a socially constructed reality. Relativism accepts that how people perceive their world and their thoughts about it are always influenced by social factors such as culture, history and language (Willig, 2008)|
|Research methods||Research action; the techniques for collecting and analysing data (Carter & Little, 2007)|
|Research question||An open-ended question that guides data collection and analysis (Willig, 2008). An important distinction is that research questions direct the study rather than predict what may be found|
|Reflexivity||Consideration of the researcher’s own role in the research process via critical self-scrutiny (Mason, 2002)|
|Strategies of inquiry||Provide the researcher with an overall strategy for formulating, articulating, analysing, and evaluating their methods (Carter & Little, 2007). Common strategies of inquiry in health research are Phenomenology, Grounded Theory, Discourse Analysis, Ethnography, Ethnomethodology and Action Research (Dew, 2007)|
|Subjective||A subjective fact is one that is only true under certain conditions, at certain times, in certain places, or for certain people. This is opposed to an objective fact that is true in every circumstance|
|Theoretical generalisability||The extent to which research findings can be used to develop concepts, understand phenomena and theoretical propositions that are relevant to other settings and other groups of individuals (Draper, 2004; Yardley, 2008)|
Qualitative research undoubtedly has much to offer nutrition and dietetics. Food has multiple functions in a person’s life and qualitative research is particularly well placed to deal with this complexity because it investigates how and why people behave in certain ways. Historically, qualitative research has been positioned low down hierarchies of evidence, although there is a growing recognition that it should be a vital part of the decision-making processes that guide the development of health policy and practice.
Despite this, qualitative research has yet to be fully embraced by nutrition and dietetics. This is perhaps unsurprising because those new to this type of research can struggle with an impenetrable literature and a lack of consensus regarding what constitutes appropriate research methods. The advice presented here is pragmatic: proceed from the research question(s) and ensure that choices about research methods are appropriate and do not clash with the philosophical assumptions about ontology, epistemology and methodology that underpin the overall design of a study.
Conflict of interests, source of funding and authorship
Both authors were involved in the preparation of this manuscript. The authors report no conflict of interest.