Ten Fundamentals of Qualitative Research PDF

 2 Ten fundamentals of qualitative research OVERVIEW Qualitative research...

 2 Ten fundamentals of qualitative research OVERVIEW Qualitative research is about meaning, not numbers Qualitative research doesn’t provide a single answer Qualitative research treats context as important Qualitative research can be experiential or critical Qualitative research is underpinned by ontological assumptions Qualitative research is underpinned by epistemological assumptions Qualitative research involves a qualitative methodology Qualitative research uses all sorts of data Qualitative research involves ‘thinking qualitatively’ Qualitative research values subjectivity and reflexivity Knowing what you now know, is qualitative research right for your project? If you’re travelling to a completely foreign country, some basic knowledge – such as what language is spoken and what the key aspects of culture and etiquette are – is vital for a successful trip. This chapter provides such an introduction for qualitative research, so that rather than blundering uninitiated into the wilderness of qualitative research, and potentially getting lost or making some fundamental errors, you can walk confidently, with solid ground beneath your feet. We introduce ten basic things you really need to know about qualitative research before you start to do it, and then discuss how you determine the suitability of qualitative research for a project. Before we begin, it’s important to note that qualitative research is a rich, diverse and complex field (see Madill & Gough, 2008). It can aim to do one or more different things: ‘give voice’ to a group of people or an issue; provide a detailed description of events or 02-Braun & Clarke_Ch-02.indd 19 28/02/2013 7:33:34 PM  20 Successfully getting started in qualitative research experiences; develop theory; interrogate the meaning in texts; identify discourses or dem- onstrate the discursive features of a text; and/or engage in social critique. Qualitative research is not a single thing, although people who don’t understand it often treat it as if it were. QUALITATIVE RESEARCH IS ABOUT MEANING, NOT NUMBERS In a nutshell, if asked what the central thing that distinguishes qualitative research as a field is, our answer would be that it deals with, and is interested in, meaning. At its core, qualitative research is about capturing some aspect of the social or psychological world. It records the messiness of real life, puts an organising framework around it and interprets it in some way. To gain an understanding of what it is, it’s also helpful to understand what it is not. As noted in Chapter 1, qualitative and quantitative research have quite different foci and purposes, and result in quite different knowledge and claims. Box 2.1 provides a useful comparison of two studies on the same broad topic, one using a qualitative approach and one using a quantitative approach; see also Table 1.1 in Chapter 1. Qualitative research is not about testing hypotheses, and not typically about seeking comparisons between groups. This isn’t to say you cannot make comparisons in qualitative analysis, but only quantitative methods provide a framework for testing difference between groups in any concrete or absolute way. And it does not aim for replication, either as a principle, or as the criterion by which the quality of research is established (see Chapter 12). Because of a focus on knowledge as something that comes from, and makes sense within, the con- texts it was generated from (see below), qualitative research does not assume the ‘same’ accounts will always be generated, every time, by any researcher. QUALITATIVE RESEARCH DOESN’T PROVIDE A SINGLE ANSWER If you love certainty, qualitative research is going to present you with some chal- lenges. Among most qualitative researchers, it’s generally agreed upon that there is more than one way of making meaning from the data that we analyse, which means there isn’t a single ‘right’ answer. One of the criticisms of qualitative research from some quantitative researchers is that, if that’s the case, then our analyses are simply ‘made up’ and don’t tell us anything meaningful; that ‘anything goes’ in qualitative research. This is emphatically not the case. An analysis of qualitative data tells one story among many that could be told about the data. The idea that analysis is like a story is a useful concept, but don’t think this means it’s fictional. Imagine you’ve gone on holiday with your family. When you come back, your story of the holiday may be quite different to your parents’ story – they may have had a fantastic time, while you were bored senseless. Each is an equally true story of the holiday. Qualitative researchers recognise that the data analyses we produce are like such stories – they are partial, and 02-Braun & Clarke_Ch-02.indd 20 28/02/2013 7:33:34 PM  Ten fundamentals of qualitative research 21 they are subjective (see below). But any good analysis needs to be plausible, coherent and grounded in the data. You don’t need to be claiming to tell the only or absolute truth to be telling a compelling ‘truth’ about your data. QUALITATIVE RESEARCH TREATS CONTEXT AS IMPORTANT Another key tenet of qualitative research is an appreciation that information and knowledge always come from somewhere. Qualitative data are understood as accounts that are not produced in the ether. Instead, they are seen to be produced in particular contexts, by participants who come from, and are located within, specific contexts. What does that mean? It means that, in contrast to the positivist/quantitative ideal of being able to obtain ‘uncontaminated’ knowledge, with all biases removed, qualita- tive research recognises that these exist, and incorporates them into the analysis. It recognises the subjectivity of the data we analyse, and the analyses we produce. Subjectivity basically refers to the idea that what we see and understand reflects our identities and experiences – the contexts we’ve existed in, a concept sometimes also referred to as ‘perspectival subjectivity’ (Kvale, 1996: 212). Qualitative research does not treat this subjectivity as bias to be eliminated from research, but tends to involve contextualised analysis, which takes this into account. QUALITATIVE RESEARCH CAN BE EXPERIENTIAL OR CRITICAL Qualitative research is exploratory, open-ended and organic, and produces in-depth, rich and detailed data from which to make claims. As a field, it can be divided into two broad camps (Reicher, 2000), which we term experiential and critical. Experiential qualita- tive research validates the meanings, views, perspectives, experiences and/or practices expressed in the data. We call it experiential because participants’ interpretations are prioritised, accepted and focused on, rather than being used as a basis for analysing something else. Critical qualitative research takes an interrogative stance towards the meanings or experiences expressed in the data, and uses them to explore some other phenomenon. Typically, it seeks to understand the factors influencing, and the effects of, the particular meanings or representations expressed. We call it critical because it doesn’t take data at face value. This means that analysts’ interpretations become more important than participants’. We’ll explain these two camps in a bit more detail. EXPERIENTIAL QUALITATIVE RESEARCH: TELLING IT ‘LIKE IT IS’ Experiential research is driven by a desire to know people’s own perspectives and mean- ings, to ‘get inside’ people’s heads as it were, and to prioritise them in reporting the research. Research becomes a process of collecting such information, and then putting an organising, interpretative framework around what is expressed in the data. 02-Braun & Clarke_Ch-02.indd 21 28/02/2013 7:33:34 PM  BOX 2.1 COMPARING A QUALITATIVE AND QUANTITATIVE APPROACH Here we contrast two studies (Christianson, Lalos, Westman, & Johansson, 2007; Herlitz & Ramstedt, 2005) on the same (broad) topic – 02-Braun & Clarke_Ch-02.indd 22 sexual risk – conducted in the same country (Sweden) to give you a sense of the different sorts of understandings that are generated by quantitative and qualitative research. Quantitative Study Qualitative Study Assessment of sexual behaviour, sexual attitudes, and sexual ‘Eyes wide shut’ – sexuality and risk in HIV-positive youth in Sweden risk in Sweden (1989–2003) (Herlitz & Ramstedt, 2005). (Christianson et al., 2007). Research aim: to identify changes in the general Swedish Research aim/question: to explore perceptions of sexual risk taking population’s attitudes, knowledge, beliefs and behaviours among HIV positive youth, and their understandings of why they related to HIV/AIDS, over time. Specific hypotheses: a) contracted HIV. sexually risky behaviour would have decreased; b) that attitudes to sex would be more conservative, due to risk of HIV/AIDS. Sample: a random sample, stratified for age, was generated Sample: a purposive sample of 10 HIV positive Swedish residents from the general population in 1989, 1994, 1997 and 2003 (five female; five male; seven born in Sweden; three born abroad) (n = 4000 each year), and a sample in 2000 randomly selected aged between 17 and 24. Participants were recruited through three but weighted toward urban dwellers (n = 6000). Overall HIV clinics/organisations. response rate was 63%, and total n = 13,762. Method of data collection: quantitative questionnaire (closed Method of data collection: in-depth semi-structured interviews (see response options), consisting of 85–90 items, delivered by mail Chapter 3); tape recorded; transcribed ‘verbatim’ (all utterances to sample. Up to three reminders were sent. transcribed as spoken). Method of data analysis: statistical. Multiple logistical Method of data analysis: grounded theory (see Chapter 8). Coding regression, a statistical method that allows determination and analysis began from the first interview. Analysis involved of the (relative) influence of multiple variables (e.g. age, multiple stages of (open) coding and re-coding and organising the sex, education level) on a particular outcome (e.g. practising data into core categories and subcategories; for credibility, four unsafe sex), in order to predict the likelihood of that participants also read and commented on the preliminary analysis outcome. (Chapter 12 discusses ‘member checking’). 28/02/2013 7:33:34 PM  Key results: neither hypothesis supported: significant Key results: identified two main clusters of factors that limited increases in casual sexual contacts without condoms, and the individuals’ possibilities for agency in sexual interactions: a) multiple partners, between 1989 and 2003; attitudes to ‘sex’ ‘sociocultural blinds’ referred to factors which make safer sex a 02-Braun & Clarke_Ch-02.indd 23 outside relationships more permissive in 2003 than 1989; hard topic to broach, like the idea that ‘being in love’ protects you attitudes were more liberal, with regard to acceptability of from sexual risk; b) ‘from consensual to forced sex’ referred to casual sex, than behaviour; youngest participants more likely factors within consensual encounters, like pleasure and trust, and to use condoms for casual sex than older participants. coercion, which resulted in risky sex. Conclusions: need for continuous, extensive sexuality education Conclusions: the data and analysis emphasise the context-bound to help reduce sexual risk which can be controlled by behaviour nature of sexual experiences and practices, and the way power and (e.g., condom-use). gender inflect most experiences. Contrary to the idea that informs health promotion, of a rational agent who makes (informed) choices about their behaviour, these accounts show that agency can be compromised by various factors beyond the individual’s control. This type of research can show: changes in sexual attitudes This type of research can show: the richness of (reported) real lived and practices (at a population level); factors that might experiences; nuances and diversity within accounts; patterns across predict particular outcomes a researcher might be accounts. Can offer insights into the lived complexity of negotiating interested in – can be useful for targeting interventions. safer sex practices. Can help understand how and why young people are at risk for HIV. This type of research cannot show: the meanings of different This type of research cannot show: general patterns across the experiences; why these changes may have occurred. population; cause and effect relationships. Evaluation: this sort of research is useful for mapping large Evaluation: this sort of research is useful for providing a compelling population-level patterns in behaviour – it provides a ‘breadth’ sense of what sexual risk/safety (or any other topic) really is like, of knowledge. The focus on association between factors can for individual people in their lives. It provides both rich and deep be very useful for targeting interventions. However, it does understandings of the ways people make sense of, and put into practice, not provide ‘deep’ or ‘rich’ understanding around sexual scientific ‘facts’ about sexual risk and safety. While it can inform attitudes, perceptions of risk, or behaviour, and so cannot interventions, it cannot be generalised to all people, and so cannot be offer understandings of why people do what they do. used to make population-level claims about sexual safety and risk. 28/02/2013 7:33:34 PM  24 Successfully getting started in qualitative research There are many reasons why a qualitative approach is more suitable than a quantita- tive one when trying to understand people’s meanings: it allows us to retain a focus on people’s own framing around issues, and their own terms of reference, rather than having it pre-framed by the researcher (e.g. items in a questionnaire); it allows a far richer (fuller, multi-faceted) or deeper understanding of a phenom- enon than using numbers, not least because the complexity of people’s meanings or experiences is revealed and retained in qualitative data; reality, meaning and experience for people often tend to be messy and contradic- tory; qualitative research can ‘embrace this messiness’ (Shaw et al., 2008: 188). Participants’ language can reveal both ‘mess’ and contradiction in a way quantita- tive methods cannot; as it can be open-ended, exploratory, organic and flexible, it can evolve to suit the needs of the project (such as accommodating unanticipated ideas expressed by participants); by collecting and analysing such data, we can find out things that we might never have imagined; things that would be lost using quantitative methods. This means the scope of knowledge and understanding is opened up considerably. What we can understand with qualitative research is not limited by the researcher’s imagination and existing knowledge in the field. Instead, participants’ experiences and meanings (personal and wider societal meanings) drive experiential qualitative research. For example, researchers wanting to understand more about young women and eating might conduct interviews or focus groups (FGs) to understand the meanings that ‘food’ has for a small group of young women, the place it has in their lives, and their experiences of eating (or not eating) food, instead of getting young women to complete a quantitative survey about food and eating which would involve them responding to categories and options pre-determined by the researcher. So experiential qualitative research seeks to make sense of how the world is seen, under- stood and experienced from the person’s perspective. Language is treated as if it provides a window to the person’s interior; it is understood as the way people report their experi- ences, practices and meanings in a straightforward fashion; it is the vehicle researchers use to access and make sense of that inner world. Research often involves what is talked about as ‘mapping’ or ‘giving voice’ to ‘the rich tapestry of people’s lives’ – analysis provides ‘rich’ or ‘thick’ descriptions of meaning and experience (see Box 2.2). BOX 2.2 ‘THICK’ OR ‘RICH’ DESCRIPTION The idea that qualitative research should provide thick description came from US anthro­pologist Clifford Geertz’s writing about ethnography in the 1970s (Geertz, 1973). Description was ‘thick’ when contexts of behaviour were described; it was ‘thin’ when context was excluded. Subsequently taken up in different ways throughout qualitative research, the idea now is often used synonymously with the term ‘rich’ to refer to detailed descriptions of the object of study, in which the complexity and contradictions of participants’ stories of their lives are included. 02-Braun & Clarke_Ch-02.indd 24 28/02/2013 7:33:35 PM  Ten fundamentals of qualitative research 25 CRITICAL QUALITATIVE RESEARCH: INTERROGATING THE STORIES WE COLLECT In critical qualitative research, in contrast, the focus is not on language as a means to get inside the person’s head, but on language as it is used ‘out there’ in the world. Its interest is in how language gives shape to certain social realities – and the impact of these. While critical qualitative research is essentially about language as a mode of communication, interest shifts away from only looking at the semantic content (the objects the words refer to). Rather, language is understood as the main mode by which the reality of our world is created, and so researchers within this tradition use language to explore the ways different versions of reality are created. They take what is called a constitutive or productive view of language; its central premise is that language creates rather than reflects reality (Weedon, 1997). For example, a project concerned with young women and eating might examine the ways young women talk about food in ways which construct distinct categories around different food types and eating patterns and habits (obvious ones might be healthy/unhealthy; good/bad; fattening/non-fattening; controlled/uncontrolled). Unlike experiential qualitative research, critical qualitative research doesn’t see such talk as offering a window into how these young women really feel about food/eating. Rather, talk is seen as depicting a reality about food that they are creating or constructing through the way they talk about it, and which reflects broader ways of understanding available in their sociocultural contexts (see Chapter 7 for more on this approach). Within this approach, research can broadly be divided between that interested in representation and construction and that interested in language practice. An inter- est in representation and/or construction is an interest in factors which shape or create meaning and the effects and implications of particular patterns of meaning. Language is one of the main means by which representation and construction occur; qualitative research is therefore ideal for researchers interested these. Qualitative research is used to understand the ways language (or imagery) tells particular stories about research objects. (Research objects – the things we study – can be concrete things, like clothing, or abstract things, like love.) To continue the previous example, a project concerned with young women and eating might analyse the ways weight and dieting are repre- sented in teen magazines, and the explicit and implicit ideas about food, weight and eating that exist there. One common representation might be of food as something that is a threat to health, self and well-being – with the implication that eating should be approached with caution, with the right types of food consumed in a controlled fashion. A key assumption is that there are numerous ways objects could be represented, and that different representations have different implications for individuals and society (Hall, 1997). For instance, the representations of body size as the result of genetics and hormones or of individual eating and exercise behaviour carry quite different inferences about how a person can or should feel, and indeed behave, regarding their body size. For a fat person, a genetic/hormonal account means that they are seen as not responsible for their fatness; an eating/exercise account makes them potentially responsible and blameworthy. Some research in this tradition involves the practice of deconstruction (Norris, C., 2002a; Parker, 1988), whereby texts are ‘taken apart’ and interrogated for the dominant and hidden assumptions (or oppositions) they rely on. 02-Braun & Clarke_Ch-02.indd 25 28/02/2013 7:33:35 PM  26 Successfully getting started in qualitative research In terms of language practice, qualitative research seeks to examine the ways lan- guage is used to create particular versions of reality. The analytic focus ranges widely: some is quite micro, with a focus on the detail of language use, such as the function of particular features of talk and texts (Hutchby, 2002). Hong Kong linguist Amy Tsui (1991), for instance, examined the ways the expression ‘I don’t know’ functions in con- versation not (just) to express a cognitive state, but to avoid or ameliorate certain deli- cate activities, such as a disagreement or making an assessment about something (see also Potter, 1997). Some is more macro, and considers the ways language produces a certain version of reality. For example, researchers have examined the ways people deploy different constructions of identity at different points in a conversation, identi- fying that these serve different purposes for the speaker (as Edley & Wetherell, 1997, show around heterosexual masculinities). For example, consider a fat man talking to a doctor about his desire to lose some weight. At one point in the conversation he may use language which suggests he is an ‘in control’ independent person who can choose to act in ways which will determine his weight. At another point, he may use language which suggests his fatness is not his ‘fault’, and that it results from forces beyond his control, such as biology or culture. Each construction has different implications, allow- ing the doctor to blame him, or not, for his weight, and to feel confident offering one of a range of weight-loss suggestions. A BRIEF SUMMARY, AND AN INTRODUCTION TO THEORY So qualitative research is concerned with words, and sometimes images, and is typically either experiential or critical. Each camp contains diverse interests (e.g. the critical camp is concerned with representation, construction or language practices). In whatever form it takes, qualitative research accesses the richness of the worlds we all exist in – whether they are the worlds that exist ‘in our heads’, or the social and physical worlds exter- nal to us. Regardless of which camp it’s in, qualitative research overall tends to come from a different theoretical position than quantitative and experimental research. And then different qualitative methodologies also have their own particular theoretical frame- works and do different sorts of things with data. In setting out a framework for research practice, methodology relies on ontology and epistemology (Ramazanoglu & Holland, 2002). These complicated-sounding words refer, respectively, to theories about the nature of reality or being and about the nature of knowledge. Each demarcates what can and Relativism Critical realism Realism ‘Reality’ is A pre-social A pre-social dependent on reality exists but reality exists that the ways we we can only ever we can access come to know it partially know it through research Figure 2.1 The ontology continuum 02-Braun & Clarke_Ch-02.indd 26 28/02/2013 7:33:35 PM  Ten fundamentals of qualitative research 27 cannot count as meaningful knowledge and informs our methodology and the process of producing that knowledge. Qualitative research often departs radically from quantitative research in both of these. QUALITATIVE RESEARCH IS UNDERPINNED BY ONTOLOGICAL ASSUMPTIONS Ontological positions specify the relationship between the world and our human inter- pretations and practices. Ontology determines whether or not we think reality exists entirely separate from human practices and understandings – including the research we conduct to find such things out – or whether we think it cannot be separated from human practices, and so knowledge is always going to reflect our perspective. There are many variations, which range along a continuum from a view where ‘reality’ is entirely independent of human ways of knowing about it – what has been described as a ‘mind- independent truth’ (Tebes, 2005) (known as realism) – to a view where reality is entirely depends on human interpretation and knowledge (known as relativism) (see Figure 2.1). Realism assumes a knowable world, which is comprehensible through research – that the truth (and there is only one) is ‘out there’ and can be accessed by the appropri- ate application of research techniques. In its most extreme or ‘naive’ version, it has been referred to as ‘a correspondence theory of truth’ (Madill, Jordan, & Shirley, 2000: 3), where what we observe is assumed to mirror truthfully what is there. Realism is the ontol- ogy underpinning most quantitative research, but it rarely informs qualitative research. Relativism, in contrast, argues that there are multiple constructed realities, rather than a single, pre-social reality or mind-independent truth, and that we can never get beyond these constructions (Cromby & Nightingale, 1999). Rather than being universal, what is ‘real’ and ‘true’ differs across time and context, so that what we can know reflects where and how knowledge is generated. A relativist ontology underpins some qualitative approaches, including some versions of discourse analysis, but it rarely informs quan- titative research. Somewhere in between sit critical realist positions, now quite com- monly adopted in qualitative research, which also invoke a real and knowable world which sits ‘behind’ the subjective and socially-located knowledge a researcher can access (Madill et al., 2000). Because knowledge is viewed as socially influenced, it is thought to reflect a separate reality that we can only partially access. The critical realist position holds that we need to claim that some ‘authentic’ reality exists to produce knowledge that might ‘make a difference’ (Stainton Rogers & Stainton Rogers, 1997). For example, we would need to be able to claim that the shame and embarrassment that (some) fat people experience in airline seats (Farrell, 2011) is real to produce knowledge that might mean airlines change standard seating sizes. In this position, an external reality (people’s feel- ing of shame) provides a foundation for knowledge. A critical realist position underpins a number of different qualitative approaches, including some versions of thematic analy- sis, grounded theory, discourse analysis and interpretative phenomenological analysis (IPA). These different positions are illustrated by analogy in Box 2.3. 02-Braun & Clarke_Ch-02.indd 27 28/02/2013 7:33:35 PM  28 Successfully getting started in qualitative research BOX 2.3 ONTOLOGY THROUGH THE LOOKING GLASS The analogy of looking at a view can be useful to illustrate realism, relativism and critical realism. Realism would be akin to looking at a view through a perfect glass window in your house. The information you access from this perfect glass window corresponds exactly to what really is outside – if you go outside, the path and garden you have seen would be there; you can walk along the path, smell the roses, and thus verify the truth of what you have viewed. Your window has given you a way to determine the reality that exists beyond it, a way to measure what is there. Relativism is better captured by the idea of prisoners looking at a view from their prison cells. Prisoners housed in different cells will see different views of the world outside the prison, but there is no way of prioritising one prisoner’s view as more real than another’s. Moreover, although the views appear real, they could be a projection or a hologram. A prisoner has no way to ascertain the truth of the information they have about what is outside the prison. It is impossible to step outside to determine if their view corresponds to the real landscape outside; there is no way of measuring the relationship between the physical landscape and the perception of it, and thus no way of knowing how true it is. Critical realism would be like looking at a view where the only way to see it is through a prism, so what is seen is nuanced by the shape of the prism (the prism is culture, history, etc.). If you could just get rid of that prism, you’d be able to see what lies behind it (the truth), but you never can get beyond it. Where you sit on the ontology continuum determines whether you think that you cannot remove the process of knowledge production from what you’re studying (relativism – we can only ever view the scene from some perspective, and never know if the knowl- edge we have of it is the only or ‘right’ one) or whether you think you can (realism – we can ascertain the true nature of the scene if only we use the right tools, i.e. look through a perfect glass window). QUALITATIVE RESEARCH IS UNDERPINNED BY EPISTEMOLOGICAL ASSUMPTIONS The central concern of epistemology is what counts as legitimate ‘knowledge’: in a world where all sorts of knowledge exist, how do we know which to trust, which are meaningful? As an illustration, consider two ideas about what we are like as persons: personality theory and astrology. Astrology tells us our ‘essence’ comes from the align- ment of astronomical bodies at the time of our birth – most basically, through year (in the Chinese horoscope) or month (in the western horoscope). Both Virginia and Victoria are (western astrological) Librans, which means we have high aesthetic senses, are hope- lessly indecisive, hate conflict and crave harmony. Astrology makes predictions on our 02-Braun & Clarke_Ch-02.indd 28 28/02/2013 7:33:35 PM  Ten fundamentals of qualitative research 29 daily (and indeed lifetime) experiences, based on planetary and other alignment. Such astrological predictions are popular, and feature in most western mainstream media. In contrast, personality theory, developed out of scientific psychology, argues that psycho- metric testing can determine our location along certain dimensions of personality (the most ‘robust’ being the ‘big five’ traits: neuroticism, extraversion, agreeableness, con- scientiousness and openness to experience; Digman, 1990). These traits form a stable core essence of who we are as a person, and can be used to predict how we behave in the world. While many in the contemporary west might ‘scoff’ at astrology (less than a quarter of Britons believe in horoscopes, Gill, Hadaway, & Marler, 1998), and dismiss it as an invalid way to theorise and understand human personality, far fewer would ques- tion the validity of psychological science. Barring some ‘critical psychologists’ (see Burr, 2002), not many westerners question the idea that we each have a personality. This is because a scientific epistemology is the dominant epistemological position in the west at present: true knowledge is determined by science; other forms of knowledge are framed as biased, untrustworthy and ‘unscientific’. So epistemology determines what counts as valid, trustworthy, ‘true’ knowledge within a community and, conversely, what is seen as not valid knowledge. If we lived in a culture where true knowledge was determined not by science, but by astrologers, the theory of personality might be treated with the same derision that astrology is by many people. So epistemology is about the nature of knowledge, and addresses the question of what it is possible to know. What counts as knowledge determines how meaningful knowledge can be generated (and what it is seen to represent). Epistemology can also be realist or relativist – a realist epistemological position assumes that it is possible to obtain ‘the truth’ through valid knowledge production; a relativist epistemological position states that, theoretically, knowledge is always perspectival and therefore a singular, absolute truth is impossible. A basic distinction between epistemological positions is whether we think real- ity (be that external or personal) is discovered through the process of research, or whether we think reality is created through the process of research. In one camp, we could characterise the researcher as archaeologist who digs to uncover a reality (knowledge) that exists quite independent of their practice (the [post-] positivist camp). In the other camp, we could characterise the researcher as a sculptor who creates a reality with their sculpture, so is involved in the production of that reality (the constructionist camp). We will briefly outline these two broad epistemological positions found within psychology and the social sciences, and a third, contextualism (Henwood & Pidgeon, 1994). POSITIVISM Positivism assumes a straightforward relationship between the world and our percep- tion of it. Closely aligned with empiricism, it separates out the practice of observation, the observer and that which is observed, and requires demonstration of reality through objective (unbiased) collection of data. Valid knowledge is obtained through the applica- tion of established scientific methods which control variables and remove various forms of contamination and bias. Appropriate application will discover ‘the truth that is out there’. In the social sciences, it would now be hard to find people who adhere to a pure 02-Braun & Clarke_Ch-02.indd 29 28/02/2013 7:33:35 PM  30 Successfully getting started in qualitative research form of positivism; instead a less pure version termed postpositivism tends to dominate. Postpositivism, which first emerged from the critiques of science by scholars like Karl Popper (Popper, 1959), still holds onto the search for the truth and sees this as, by and large, achievable, but acknowledges that researchers are influenced by their contexts, and influence research – that facts are not neutral reflections of the truth, but are theoretically influenced (Clark, 1998; Guba & Lincoln, 2005). Where this differs from construction- ist epistemologies is that postpositivist researchers still believe in, and aim to know, that singular truth, and thus seek to control for or remove subjective influences on knowledge production as much as possible. Qualitative research is sometimes undertaken within this paradigm, and some argue that postpositivism should take its place as an appropriate epistemology for qualitative research (Michell, 2004); others disagree. CONSTRUCTIONISM Other approaches question idea that knowledge is an objective (or as objective as possible) reflection of reality – but instead see our ways of knowing the world as tied to the (social) world in which we live. Constructionist epistemologies argue that the world and what we know of it do not reflect an ‘out there’ true nature of the world (or an ‘in there’ true psychology (waiting to be discovered, but that what we know of the world, and ourselves and other objects in the world is constructed (produced) through various discourses and systems of meaning we all reside within (Berger & Luckmann, 1967; Burr, 2003; Gergen, 1985, 1999). As these change, truth changes, meaning there is no one truth which a certain method allows access to – there are knowledges, rather than knowledge. The terms in which the world is understood are seen to be related to specific social, cultural contexts. Knowledges are viewed as social artefacts, and are therefore seen as social, cultural, moral, ideological and political. A critical stance tends to be taken regarding perceived truths and taken-for-granted knowledge. Given this position, there are numerous possible ways to create truths. However, constructionism as an epis- temological position is not saying that knowledge is just ‘made up’ and ‘anything goes’ (nor that there is nothing ‘outside the text’, meaning no material or experiential reality exists; Edley, 2001b). What it is saying is that knowledge of how things are is a product of how we come to understand it – for example, the ‘knowledge’ that people really do have personalities is the product of a long history of theorising ‘the person’ and per- sonality research, rather than independent fact (Burr, 2002). The process of knowledge production is still (often) empirical in that it is grounded in data, and understanding of some kind is sought. Where this differs from realist and positivist positions is that there is no singular underlying reality that is theorised as providing the foundation for true knowledge – constructionism is a non-foundational view of knowledge. CONTEXTUALISM Sitting between these poles, and somewhat akin to critical realism, contextualism is another epistemology which has a foot in both camps (Henwood & Pidgeon, 1994). Tebes cites Pepper (1942) to describe the central metaphor of contextualism as ‘the human act in con- text’ (2005: 216). It can be seen as a version of constructionism (constructionism-[very]-lite), 02-Braun & Clarke_Ch-02.indd 30 28/02/2013 7:33:35 PM  Ten fundamentals of qualitative research 31 in that it doesn’t assume a single reality, and sees knowledge as emerging from contexts (hence the name) and reflecting the researcher’s positions, so that it is local, situated, and therefore always provisional (Madill et al., 2000; Tebes, 2005). But it does retain an interest in understanding truth, and hence has a realist dimension. It argues that while no single method can get to the truth (Tebes, 2005), knowledge will be true (valid) in certain contexts. So it retains a notion of ‘the truth’ which constructionism rejects. ONTOLOGY, EPISTEMOLOGY, METHODOLOGY AND METHOD Ontology and epistemology are far from independent of each other, and they lead into particular methodologies and together constrain the methods that are appropriate for your research. For instance, if you take a relativist/constructionist position where you don’t think that ‘personality’ or ‘attitudes’ are stable constructs that reside in people’s heads and exist independently of context, or the process of finding out about them, you simply cannot ‘measure’ them using personality tests or attitude scales. Instead of seeing ‘personality’ or ‘attitudes’ as individual variables (the realist/positivist take), they can be understood as things which become meaningful as we present ourselves in the world and interact with others, which can be understood, captured and analysed by looking at language and discourse (see Burr, 2002; Potter & Wetherell, 1987). The point here is not that one of these approaches is inherently right and one is wrong (although many people align with one or the other and indeed think that), but that different sorts of knowledge are generated within different theoretical and methodological frameworks – as US psychologist Jeanne Marecek (2003: 54) has observed, it’s about asking (and answering) the question: ‘What kind of truth am I interested in hearing?’ What’s important is coherence within your over- all research design (see Chapter 3). QUALITATIVE RESEARCH INVOLVES A QUALITATIVE METHODOLOGY Two terms often used interchangeably when talking about research are method and meth- odology. But they are different. Method refers to a tool or technique for collecting or analysing data: the interview and the survey are methods of collecting data; thematic analysis and analysis of variance are methods of analysing data. Methods are quite spe- cific and applied in specific ways. Methodology is broader, and refers to the framework within which our research is conducted. It consists of theories and practices for how we go about conducting research. It provides a package of assumptions about what counts as research and how it is conducted, and the sorts of claims you can make about your data. It tells us which methods are appropriate for our research and which are not. Methodology can be understood as a theory of how research needs to proceed, to produce valid knowledge about the psychological and social world. It is what makes our research make sense, both in terms of design, and in terms of process. Methodology provides a framework for making a series of decisions about your research, including: 02-Braun & Clarke_Ch-02.indd 31 28/02/2013 7:33:35 PM  32 Successfully getting started in qualitative research How can participants be selected? What methods of data collection and analysis are appropriate? Who can or should conduct research? What is the role of the researcher? The obvious question, then, is what are the features of a qualitative methodology? Unfortunately, there’s no single answer. One thing you’ll quickly learn about qualita- tive research is that simple straightforward answers aren’t that common – sorry! This is because ‘qualitative inquiry is rife with ambiguity’ (Patton, 2002: 242), because there are lots of methodologies, and because many qualitative data analysis ‘methods’ are more accurately understood as methodologies. For example, IPA is much more than a method of analysing data – it sets out a whole framework for conducting a study; grounded theory and discourse analysis likewise. Although each qualitative methodology is unique, they share many similar features, like siblings in a big family. And like siblings, some meth- odologies are more similar than others, and ‘get along’ better with other methodologies (they share similar core assumptions). This stems from the fact that they all come under the umbrella of a qualitative paradigm (see Box 2.4). BOX 2.4 METHODOLOGIES AND PARADIGMS A qualitative paradigm provides the essential – and generic – criteria for what counts as qualitative research. But then each specific methodology varies from this to provide a framework for how an actual study is conducted and the data interpreted. Different methodologies share many features from the generic qualitative paradigm, but not necessarily all, and they may be expressed differently. Think of the relationship as akin to the difference between the generic criteria for making a cake, and the particulars of a specific recipe. A cake generically consists of sugar, butter, eggs, flour and a raising agent. The sugar and butter are creamed, the eggs beaten in and the dry ingredients gently added. The mixture is poured into a single tin, and baked immediately in a pre-heated oven. Almost any cake that is made will resemble this process, but the specifics will vary: butter may be replaced with oil; the eggs may be separated and the whites beaten first; different ingredients (e.g. chocolate, carrots, vanilla) may be added for flavour and texture. The resulting outcome is recognisably a cake, but is not identical to other cakes, either in making or in outcome. Similarly, research conducted using a qualitative methodology like discursive psychology is recognisably qualitative, but quite different from research conducted using a different qualitative methodology, such as IPA (see Chapter 8). BRIEF SUMMARY Phew! You’ve survived a whirlwind introduction to some of the fundamental things you need to know about what qualitative research is and what it is not, and made it through 02-Braun & Clarke_Ch-02.indd 32 28/02/2013 7:33:36 PM  Ten fundamentals of qualitative research 33 the most complex discussion of theory you’ll find in the book. You should by now understand that qualitative research is more than just collecting and analysing qualita- tive data; it refers to a cluster of different methodologies which offer frameworks for conducting research and producing valid knowledge. What a qualitative paradigm tells us is that useful knowledge can be generated by looking at meaning, with small samples, and that the researcher should not theorise themselves as absent or removed from this process. It also locates knowledge as contextual, and always partial, and as linked to particular theoretical and methodological commitments. QUALITATIVE RESEARCH USES ALL SORTS OF DATA Data are the bedrock of the social sciences – they are what we use to answer the ques- tions we have, and generate new and useful understandings of phenomena in the world. Without data, we’d all be theorists. Not that there’s anything wrong with being a theo- rist, but the social sciences are located within an empirical tradition, where knowledge is grounded in data. Qualitative research is just as empirical as quantitative, as its knowl- edge generation is also based on data. Within qualitative research, the data you collect are relatively ‘naturalistic’ in the sense that they aren’t put into pre-existing categories – they are not pre-coded and cat- egorised at the point of collection (like responses to a scale on a questionnaire would be) (Lincoln & Guba, 1985). The raw (uncoded) data retain the messiness of real life. However, they are not necessarily ‘natural’ in that we ourselves often generate the data we use (what counts as naturalistic data is a debate in some areas of qualitative psychol- ogy, e.g. Potter & Hepburn, 2007; Speer, 2002a). The possibilities for data are seemingly endless: the most common – and very useful – methods of data collection are individual interviews (for examples, see Ahmed, Reavey, & Majumdar, 2009; Whitehead & Kurz, 2009) and group discussions (e.g. see Augoustinos, Tuffin, & Every, 2005; Schulze & Angermeyer, 2003) – done face-to-face, over the phone or via email. Data could also come from sources like diaries (e.g. Milligan, Bingley, & Gatrell, 2005); qualitative surveys (e.g. Toerien & Wilkinson, 2004); parliamentary debates (e.g. Ellis & Kitzinger, 2002); women’s magazines (e.g. Roy, 2008); internet discussion sites (e.g. Wilson, Weatherall, & Butler, 2004); television documentaries (e.g. Hodgetts & Chamberlain, 1999); dinner-time conversations (e.g. Wiggins, 2004); or even the text on a tube of children’s toothpaste (e.g. Parker, 1996). Others are increasingly using visual imagery as well – from photo-elicitation techniques, where photographs are used in conjunction with interviews (e.g. Oliffe & Bottorff, 2007), to the use of the visual in its own right (e.g. see contributions to a special issue of Qualitative Research in Psychology on this topic, Frith et al., 2005) – but, as noted in Chapter 1, we focus primarily on text in this book. SO WHAT ‘COUNTS’ AS DATA? So with seemingly endless possibilities, how can we know what ‘counts’ as data? Data come into being either through production, where we generate them ourselves by what 02-Braun & Clarke_Ch-02.indd 33 28/02/2013 7:33:36 PM  34 Successfully getting started in qualitative research we get participants to do, or through selection from existing materials, such as media reports or blogs. With production, we develop criteria for how we generate the data, and who we have as participants; with selection, we develop criteria through which we select a particular collection of instances, which become our data. Sampling criteria are cov- ered in the next chapter, alongside the question of how much data one needs – a tricky question in qualitative research (Morse, 2000) (see the design tables, Tables 3.1–3.3, in Chapter 3). NOT ALL DATA ARE CREATED EQUAL How do you know if data are any ‘good’? If they are well selected (see Chapter 3), that’s a start. If you can easily see things to say about them, that’s a good indication, too. Another aspect that often gets talked about is ‘rich’ or ‘shallow’ data – with rich data generally better. Shallow or ‘thin’ data are those which only really access the surface of a topic, the everyday or obvious stories about it; rich data are those which offer a more thorough, thoughtful or unexpected commentary on the topic. Imagine two friends had eaten a meal they’d enjoyed. When you asked what it was like, one said, ‘My steak was really good’, and the other said, ‘I had an asparagus and lemon risotto and it was incred- ible. It combined different colours, flavours and textures so that it was a whole eating experience, heaven to my eyes, nose and mouth. For me, texture’s the most important thing in a meal, even more than flavour. The asparagus was barely cooked while the rice melted, and so it was crunchy and creamy all at once. I couldn’t stop eating it, but had to slow down to savour every mouthful so it wouldn’t end. The only problem was, they didn’t give me enough. It’s one of the best things I’ve ever eaten.’ The first description is ‘thin’; the second, ‘rich’. Rich data are preferred in qualitative research. They are data which reach below the surface, and allow the researcher to gain a deep understanding of the topic of interest. They can be generated, for instance, by a participant who reflects on their experience in a thoughtful manner (Morse, 2000), by an interviewer who asks critical probing ques- tions, or by text which articulates in depth the dominant meanings about a topic (or goes beyond them). Importantly, richness of data is also determined by context, purpose and mode of data collection: what counts as rich in a qualitative survey would be quite different from what counts as rich in an in-depth interview. Unfortunately, it’s hard to provide an absolute definition of what makes data rich, but with experience you will know it when you see it. The weight and obesity FG on the companion website contains rich data – see Sally’s first extract (starting line 79), and how we coded it for our IPA example (Table 9.3 in Chapter 9) for an example of rich data. To be good, data also have to serve the purpose of the research. What is primarily important is that the data allow you to address your research question (see Chapter 3). To continue the idea of healthy eating, if you are interested in what might stop men from ‘healthy eating’, for instance (e.g. Gough & Conner, 2006), the best place to start looking is with men themselves, and what they say. Regardless of what data we use, we always need to be able to justify our choice of data: why, and how, did we select or produce them, and why are they useful to enable us to answer our research question? 02-Braun & Clarke_Ch-02.indd 34 28/02/2013 7:33:36 PM  Ten fundamentals of qualitative research 35 QUALITATIVE RESEARCH INVOLVES ‘THINKING QUALITATIVELY’ Research is a culture, and different cultures exist for different research paradigms and methodologies. After a while you become enculturated, knowing the language and implicitly understanding the norms and practices of a particular research culture. If you’ve studied much psychology, for instance, you’re probably already somewhat enculturated into another research culture: a ‘“quantitative” culture’ (Gough, Lawton, Maddill, & Stratton, 2003: 5), with the scientific experiment as pinnacle of research excellence. This culture includes norms and expectations about hypotheses, objectivity, bias, (statistical) significance, replication, and so on. Reading that form of psychology requires understanding these concepts; conducting that form of research (well) requires putting those into practice; writing it up means using the appropriate language and style. Qualitative research culture is quite different (e.g. see Marecek, 2003) and part of becoming a qualitative researcher is learning to think qualitatively (Anderson, 2008; Morse & Richards, 2002). In this section we discuss some of the language and concepts of qualitative research, to start you on that process. As we’ve discussed already, the purpose of qualitative research is to understand or explore meaning and the ways people make meaning, rather than to prove a theory or deter- mine a relationship between factors. This means that while qualitative research has research questions (see Chapter 3), and aims and objectives (although sometimes expressed in a different language), and even theories it examines, it does not typically have hypotheses which are empirically tested and (dis)proved. So in developing a qualitative orientation and project, you want ‘understanding’ as your key driver. For instance, your project might have ‘understanding fat women’s emotional experiences around clothes shopping’ (Colls, 2006) as its aim. Once you have analysed your data, how should you talk about your analytic insights? Quantitative research uses the language of results and findings, terms which can be problematic for qualitative research, as are words like revealed, discovered and uncovered. This is because they all stem from the (positivist empiricist) model of research as a process of excavation – research as archaeology – where, if you dig deeply, using the appropriate skills and tools, in an organised fashion, you will dis- cover the truth hidden within. Qualitative methods stem from a different model of research, which is more akin to the process of sculpting or patchwork quilting (some suggest the qualitative researcher should be seen as a quilt-maker; Denzin & Lincoln, 2005a). Let’s consider the patchwork analogy. You start from a position where you have lots of shapes of material (your data), some of which are similar to others, and some of which are different, and then you work at arranging them in to a particular pattern to tell a particular story (the patterned quilt – i.e. analysis – you produce). This metaphor is useful in dealing with the suggestion that ‘anything goes’ in quali- tative research. While we could make different patterns on our quilt, through using the fabric (data) in different ways, the end product would still represent what we had 02-Braun & Clarke_Ch-02.indd 35 28/02/2013 7:33:36 PM  36 Successfully getting started in qualitative research started with (our raw data). Furthermore, some quilted patterns will ‘work’ (e.g. look great, be coherent and well organised), while others won’t (e.g. look bad, be random or chaotic); all will be constrained by the raw materials – just like with qualitative data analysis. While this is partly a matter of taste, it is not just individual – the community agrees in general on what counts as good and bad quilting (competitions get judged, prizes get awarded) – there are standards; the same goes for qualitative analysis (see Chapter 12). One of the best ways to learn the language of qualitative research is through reading qualitative research, but keep in mind that what we’ve introduced here is best practice for qualitative research reporting. People don’t always do this; even very experienced qualitative researchers can find it difficult to ‘let go’ of their quantitative training, and sometimes academic journals, particularly ones less welcoming to Big Q qualitative research, may require the use of more quantitative language. QUALITATIVE RESEARCH VALUES SUBJECTIVITY AND REFLEXIVITY There are two other important issues to grasp before becoming immersed in qualitative research culture. The first of these is subjectivity in research; the second, closely related, is reflexivity. Within a positivist-empiricist model of research, where objectivity is val- ued, avoiding bias is a prime concern. Bias refers to the idea that the researcher might (inadvertently) have influenced the results, so that they cannot be trusted. This could be through poor sample selection, poor instrument design, or their research practice (for example, the idea of interviewer effects, where characteristics of the interviewer can be shown to influence data collection; e.g. Singer, Frankel, & Glassman, 1983). Within a qualitative paradigm, the question ‘How might the research be biased?’ fails to make sense, as all research activity is seen as influenced, and the influence of the researcher is just one of many influences, albeit often a significant one. In the same way objectivity is valued in a quantitative paradigm, subjectivity is positively valued in the qualitative paradigm (e.g. consider books like Fine, 1992; Hollway, 1989). Research is understood as a subjective process; we, as researchers, bring our own histories, values, assumptions, perspectives, politics and mannerisms into the research – and we cannot leave those at the door. The topics we find interesting to research, and ways we ask questions about them, the aspects of our data that excite us – these (and many other factors) reflect who we are; our subjectivity. Therefore, any knowledge produced is going to reflect that, even if only in some very minor way. The same has to be said for participants in our research; they bring their own experiences, perspectives and values to the research. They’re not robots; we’re not robots – we’re all living, breathing, subjective human beings, partial in our knowledge, and flawed. Is this a problem? The short answer is absolutely not. The model of knowledge in qualitative research is one where a robot would make a terrible researcher; in qualita- tive research, our humanness, our subjectivity, can be used as a research tool. But to 02-Braun & Clarke_Ch-02.indd 36 28/02/2013 7:33:36 PM  Ten fundamentals of qualitative research 37 do qualitative research well, and to use subjectivity in this way, it needs to be thought about and considered. The way to do this is by being reflexive. Reflexivity is an essential requirement for good qualitative research (and usefully there are whole books dedicated to helping us become reflexive; e.g. Etherington, 2004; Finlay & Gough, 2003). Reflexivity in a research context refers to the process of critically reflecting on the knowledge we produce, and our role in producing that knowledge. It can be useful to distinguish between two forms of reflexivity – functional and personal (Wilkinson, 1988). Functional reflexivity involves giving critical attention to the way our research tools and process may have influenced the research. For instance, this might include consideration of the ways the stories that people tell about dieting might be influenced by the method we choose – for instance, data collected via FGs or researcher-directed diaries are likely to be different: an FG involves face-to-face interaction and discussion with others, whose comments and bodies might impact on what any one participant shares; a diary invokes a mode of directed yet somehow private, intimate disclosure, where the researcher is less directly present (see Chapters 5 and 6). Personal reflexivity in research is about bringing the researcher into the research, making us visible as part of the research process – unlike in quantitative research, where the researcher is typically invisible (that robot). At the very least, it might involve acknowledging who we are as researchers (much like Victoria and Virginia did in Chapter 1). More than that, it involves considering how factors like our embodiment (our physical bodies and what we do with them) can influence the pro- duction of knowledge within research (e.g. Burns, 2003, 2006; Rice, 2009), or how our assumptions can shape the knowledge produced. For example, Virginia (Braun, 2000) examined heterosexism (the assumption that everyone is heterosexual) within her doctoral research – both her own and that of her participants – and discussed both how this had ‘shut down’ possible avenues of inquiry, and the implications for research practice more broadly. Reflexivity is essential in all qualitative research, and can be seen as part of ‘quality control’ in qualitative research (see Chapter 12). That said, it’s not easy to do (particularly if you’re embedded in a quantitative research culture), so we encourage you to start this process right away. One way to assist this is to keep a research journal which records your thoughts, feelings, and reflections about your process. We encourage all qualitative researchers – however experienced – to keep a reflexive research journal throughout the process of their research (see Chapter 3). KNOWING WHAT YOU NOW KNOW, IS QUALITATIVE RESEARCH RIGHT FOR YOUR PROJECT? Researchers choose their research methodology for a combination of epistemologi- cal, political and practical reasons. Epistemology delimits the types of approaches that are possible; politics may guide the researcher towards methods that can give answers 02-Braun & Clarke_Ch-02.indd 37 28/02/2013 7:33:36 PM  38 Successfully getting started in qualitative research that can be used for particular ends; and practical factors, such as our skills and the resources we have, constrain research design in various ways. Even if the questions ‘What do I want to know?’ and ‘Why do I want to know that?’ drive our research, the questions themselves often already reflect epistemological, ontological and political positionings. For example, if you want to try to understand the July 2011 far-right terrorist attack in Norway, in which Anders Behring Breivik killed 77 mostly teenage Norwegians, by measuring levels of racism in Norway, your research is already situ- ated with a (post)positivist framework; if you wanted to understand this act by analys- ing the construction of Norwegian national identity among online postings by members of ultra-right-wing organisations, you’re situated within a qualitative paradigm. So depending on your theoretical commitments, what you want to know, and why, qualita- tive approaches might not be suitable. They might not be suitable because you want to establish some sort of relationship between various factors (say between prevalence of depression and race/ethnicity in the US; Riolo, Nguyen, Greden, & King, 2005). Or they might not be suitable because you want to ascertain how common something is within a particular population (e.g. the prevalence of eating disorders in Norwegian adolescents [Kjelsås, Bjørnstrøm, & Götestam, 2004] or in lesbian, gay and bisexual populations [Feldman & Meyer, 2007]). Chapter 3 discusses research questions in more depth. Once you’ve identified the benefits of a qualitative approach, then you have to deter- mine which particular approaches within the paradigm are most useful. Some qualita- tive writers have described an (ideal) process for qualitative research as one of bricolage where the situated, subjective, knowledgeable, inventive researcher selects and uses the best of a wide variety of tools, techniques and theories at their disposal to collect data and tell a story about their research object which answers their research questions (Denzin & Lincoln, 2005a). This book will give you the confidence to understand and use a range of qualitative methods and approaches. CHAPTER SUMMARY This chapter: introduced ten fundamental things you need to grasp about qualitative research before you start do it, including: {{ knowing that qualitative research is about meaning; {{ recognising that qualitative research doesn’t seek a single answer or single truth; {{ recognising that context is really important in making sense of qualitative data; {{ understanding that qualitative research is not a single approach or method, and different forms have different purposes; {{ recognising that ontology and epistemology are important for research, and understanding some of the main variations within psychology; {{ grasping the concept of methodology; {{ knowing what different sorts of data suit qualitative research, and what con- stitute ‘good’ qualitative data; 02-Braun & Clarke_Ch-02.indd 38 28/02/2013 7:33:36 PM