7.1 Sociological and psychological investigations
Oxford Textbook of Public Health
Sociological and psychological investigations
Myfanwy Morgan, John Spicer, and Margaret Reid
Areas of sociological investigation
Areas of psychological investigation
Sociology and psychology share a number of common interests and areas of study in the public health field. These include issues relating to the psychosocial environment as causes of disease, help-seeking behaviours and adherence with treatment, lifestyle changes and effective health promotion interventions, ageing, disability and mental illness, and the measurement of health status and quality of life. However, each discipline is characterized by its own perspectives and concepts, and by differences in their focus and explanations. Whereas psychology is primarily concerned with the personality, motivations, and behaviours of individuals, sociology focuses on the characteristics of social groupings and the wider socio-economic and political structures within which they are embedded.
This chapter describes the theoretical perspectives and methods of investigation that characterize sociological and psychological studies in the health field. Each section also has examples of the applications of these disciplines to particular areas of public health; topics relating to the sociology part are chronic illness, ethnicity, and the professions, and those relating to psychology are coronary heart disease and condom usage.
Medical sociology as a specialized field of sociological enquiry has partly developed alongside public health medicine. Both disciplines have brought their particular expertise to examine such issues as the causes and distribution of ill health and the organization and provision of care, often working in multidisciplinary teams. Sociology and public health medicine frequently trace the origins of their common interests in the role of economic, social, political, and cultural factors in health and disease patterns to the nineteenth-century struggles of physicians and public health reformers like Virchov in Germany, Chadwick in the United Kingdom, Shattuck in the United States, and Coronel in The Netherlands. The development by social scientists and public health specialists of systematic studies to examine the social distribution and causes of non-infectious conditions (such as lung cancer, bronchitis and coronary heart disease), as well as questions relating to the organization and provision of health services, dates from the late 1940s in both Britain and the United States (Hoffer and Schuler 1948; Pearce and Crocker 1949). This was encouraged by changing patterns of disease, increasing concerns about the social conditions and health of populations, and developments in social science survey techniques.
The period of rapid growth of medical sociology started in the 1950s in the United States and about a decade later in the United Kingdom. The founding of an international journal, Social Science and Medicine, which forms a marker in the development of the discipline, occurred in 1966. However, from the early 1960s the paths of sociology and public health medicine began to diverge, reflecting their separate concerns and disciplinary developments. During this period medical sociology was increasingly critical of the assumptions and limitations of the biomedical model. For example, sociologists drew attention to the wider societal determinants of individual behaviours such as smoking, diet, and non-uptake of services, and were critical of what was viewed as a ‘victim blaming’ stance and overly individualistic approaches to health promotion (McKinlay 1975). They also promoted the rights of consumers/patients to make choices and be listened to in relation to the provision of medical care, highlighted areas of potential disagreement and conflicts between professionals and patients (Bloor and Horobin 1975; Oakley 1980), and drew attention to the adverse effects of long-term care in institutions run on custodial lines on patients’ self-concept and behaviours (Goffman 1961; Wing and Brown 1970).
From the late 1970s, sociology and public health medicine experienced a new phase of shared interests. This largely reflected the adoption by public health of a broader social model of the causes of ill health and appropriate interventions, as exemplified by the development of what is often referred to as the New Public Health (Ashton and Seymour 1988; Martin and McQueen 1989). This is an approach that sees many contemporary health problems as social rather than individual problems, and therefore requires healthy public policies (in terms of housing, the environment, food policies, and so on) to support the promotion of health. The move towards a more social and participative model of health care was also encouraged by broader changes in society. This includes the greater importance attached to local communities having a voice in determining local needs and priorities, and to patients being afforded a more active role in treatment choices and in assessing the outcomes of health care.
The rate of growth of medical sociology (or what is now appropriately termed the sociology of health and illness) has varied between countries. However, currently there is considerable activity across different parts of the world. For example, the first three issues of Social Science and Medicine in 2000 (Volume 50, numbers 1 to 3) contained 33 papers that were distributed geographically (based on the location of the first author) as follows: Western Europe (13 of which 11 from United Kingdom), North America (11), Southeast Asia and Pacific Rim (five), Africa (three), and South America (one). Many of the topics covered are similar across these areas, although research takes account of particular local circumstances, cultural beliefs, and characteristics of health systems.
Research undertaken by sociologists in the health field ranges from empiricist quantitative surveys, to more sociologically informed analyses of health and care issues, to research which is primarily concerned with developing new sociological theories and explanations. Similarly, sociologists may often employ quantitative methods of investigation and undertake large-scale surveys or participate in randomized controlled trials, as well as conducting more in-depth studies using qualitative methods of data collection and analysis. While acknowledging the considerable overlap and collaboration between the disciplines of public health medicine and sociology, this section nevertheless focuses on what is distinctive about sociological investigations in terms of their perspectives, the methods of investigation employed, and the questions addressed.
The broad conceptual framework for sociological investigations comprises theories of the nature and workings of society. However, sociology is characterized by a range of theories rather than a single unified approach. These theories influence ways of ‘seeing’ the social world and inform the questions asked about particular aspects of health, medicine, and health care and the types of explanations given, as well as influencing the methods of research.
Sociological theories can be broadly divided into structural and social action theories. Structural theories emphasize the characteristics and operation of the broad structures of society, which are viewed as comprising an external social reality that shapes prevailing beliefs, values, and behaviours of social groups. In contrast, social action theories acknowledge the existence of multiple realities, and are concerned with the individual’s subjective perception of their social world and actions on the basis of meanings. However, each of these broad theoretical perspectives comprises distinct sociological traditions that differ in their focus and explanations of the nature and workings of society. This diversity forms an important strength of sociological investigations in public health as in other fields of enquiry. Moreover, all sociological research involves a theoretical position, albeit this is often implicit. Thus although sociological research in the field of health and medicine has a strong empirical tradition, it is always based on some assumptions regarding the nature of social reality which in turn influences the questions posed and how evidence is collected, analysed, understood, and used. As Alderson (1998) observes, ‘hypotheses are explicit, but when theories are implicit their power to clarify or to confuse, and to reveal or obscure new insights, can work unnoticed’.
Structural theories share a common focus on the structure of society and its institutions, and regard society as constraining and shaping the beliefs, values, and patterns of behaviour of social groups. They also emphasize the objective nature of social phenomena and the possibility of applying positivist methods in their study (see section on sociological methods). However, structural theories differ in their view of the nature of society and the relationships between groups. They can be broadly divided into consensual theories that assume a consensus of values and goals which forms the basis of social solidarity and co-operation (and is therefore compatible with a fairly medicocentric approach), and conflict theories that emphasize the struggle for advantage between different groups in society.
Consensual theories, most notably that of functionalism, played an important role in the early development of sociology. Particularly influential was Talcott Parsons (1902–1979), who undertook a theoretical analysis of society from a functionalist perspective and applied this to the question of how society contains the potentially disruptive effects of illness (Parsons 1951). Parsons regarded illness as dysfunctional for the smooth running of society because sick people are not able to fulfil their normal social roles (work roles, familial roles, community activities, and so on). Parsons thus depicted illness as a social as well as a biological phenomenon. He identified the mechanisms evolved by society to manage and control the amount of illness in society as being achieved through the special status and role assigned to sick people (the ‘sick role’) and the complementary role for doctors. Specific expectations of the sick person are identified as the need to seek professional help (in relation to the nature and severity of the illness) and to co-operate with the doctor. Privileges are the granting of exemption from performance of normal duties and responsibilities relative to the nature and severity of the illness (such as time off work), and being viewed as in need of care. Reciprocal expectations of doctors include the need to apply a high degree of knowledge and skill to problems of illness, to act for the welfare of the patient rather than their own self-interest, to be objective and emotionally detached, and to be guided by the rules of professional conduct. These expectations of both doctor and patient both ensure that the patient is restored to health and hence normal role performance as quickly as possible, and also that the privileges of the sick role are not abused through inappropriate entry or malingering.
Parsons’ theory has been criticized for not being applicable to chronic illness and for being ethnocentric. However, his ideas extended the view of illness from a purely biological state to the social sphere and have since been widely employed. For example, the ambivalent responses often displayed by the public and medical staff to conditions such as alcoholism, overdoses, and AIDS can be explained by their unwillingness to view such conditions purely as sickness requiring treatment and therefore to grant such people the privileges of the sick role. Instead, there is a tendency to ascribe notions of responsibility and thus of punishment to such patients and possibly to their close relatives (Pollack et al. 1992). Similarly, issues regarding the appropriate expectations of family members following major illness, and the adverse effects on recovery of placing them in a dependent sick role, are of relevance for successful rehabilitation.
Whereas functionalist theory with its emphasis on explaining social stability was influential in the early development of sociology, by the late 1960s it had been largely replaced by conflict theories which were more in keeping with the tenor and mood of the times. Marxist theory (historical materialism) focused on the question of how social change happens. Karl Marx (1818–1883) depicted society as progressing through a series of stages and revolutionary transformations based on the development of systems of production. In brief, the economic infrastructure was seen as determining the dominant social relations and major institutions of society, including its political structures, laws, and ideology, with the basic social division in modern capitalist society being on class lines. The relationship between the owners and the propertyless workers, the proletariat, was depicted as an antagonistic and exploitative one, sustained by the inequality in power and resources.
Materialist theories were influential in the health field during the 1970s and early 1980s, especially in relation to analyses of the social causes and distribution of disease. This political economy framework shifted the focus for understanding inequalities in health from the activities of the individual and the institution of medicine, to the effects of inequalities in income and the organization of work in capitalist economies. Rates of ill health among workers and their families was partly attributed to physical health risks associated with work, including exposure to toxic substances and risks of accidents and injury, and reflected the contradiction that exists between the pursuit of health and the pursuit of profit (Doyal 1981). Similarly, the stresses associated with workers’ lack of autonomy and situation of alienation in the sphere of production was viewed as a cause of psychological health risks (Schwalbe and Staples 1986). Workers’ lack of satisfaction in the sphere of production was also seen as encouraging the pursuit of satisfaction in the sphere of consumption, thus leading to high rates of cigarette smoking and other behaviours harmful to health (Navarro 1982).
Historical materialism is now in decline as a general theory, reflecting both the decline of communism and the restructuring of capitalism. However, it has drawn attention to sectional interests and social conflict as a routine feature of modern societies, and to power and domination as a means through which social integration and the consent of subordinate groups may be secured. This emphasis on conflict and competing interests was associated in the health field with the demystifying of medicine and an emphasis on its iatrogenic effects (Illich 1977), and assessments of the limited contribution of medicine to the decline in mortality rates (Powles 1973). In terms of the doctor–patient relationship there was a new emphasis on the conflicts that may occur between doctor and patient, given their differing interests and knowledge (Freidson 1970). Studies of health policies also demonstrated the effects of competing interests, with different coalitions occurring in relation to different issues (for example, restrictions on tobacco sales or controls on environmental pollution, and so on) (Calnan 1984). A further change was the emergence of feminist analyses which regard gender relations rather than class relations as generating fundamental inequalities and oppression (Jagger and Rothenberg 1984). This focused attention on gender issues in relation to health and health care, and particularly the effects on health of women’s domestic position and labour force participation. In addition, greater attention has been paid in the United Kingdom to the significance of ethnicity for health and health care following the recording of self-assessed ethnicity in the 1991 census, while issues of race and ethnicity had traditionally received greater attention in the United States.
Whereas consensual and conflict theories were developed to explain the nature and workings of industrial society, new challenges for sociological theory have been presented by the recent period of rapid and extensive social change leading to a ‘postmodern’ era. This process has been likened to the industrial revolution, in that it involves the transformation of the major institutions of society and permeates all aspects of everyday life and self-identity. Movement into an era of postmodernity is driven by a process of ‘globalization’, which involves changes in the economic, technological, and social spheres. This includes the deregulation of markets and economic liberalization, the diffusion of technology and increased speed of communications, increased regionalization, greater involvement of the state in intergovernmental business, and the development of a global culture which crosses divides of time and space. Such changes weaken traditional institutions (including traditional structures of the family and the character of work) and the powerful collective identities that went with them. As a result society is becoming more fragmented, heterogeneous, and ‘individualized’, with society forming a collection of factions competing for power and resources (Giddens 1991, 1994; Beck 1992). For example, in contrast to the traditional emphasis on social class as a major social division, there is much debate as to whether social class continues to be a meaningful category or can be viewed as just one of several patterns of power and inequality (Savage and Miles 1994). Increased emphasis is therefore now given in the health field, as in other branches of sociology, to the variety of social groupings, defined by age, gender, and ethnicity, as well as socio-economic circumstances. However, it is argued that such groupings, although of relevance for people, are no longer the units that determine individual experience. Instead the heterogeneity of social life is emphasized, with individuals no longer primarily defined in terms of occupation and class but seen as constructing ‘self-identities’ through consumption and lifestyle which form the major cleavage in contemporary society (Baumann 1987). However, Higgs and Scambler (1998) caution that postmodern perspectives give insufficient attention to the continuing significance of class, including the salience of an excluded underclass and an inconspicuous ruling/capitalist class.
Social action theories
Social action theories are based on a fundamentally different view of the social world from structural theories. The latter take a realist perspective and regard the world as having an existence independent of our perception of it. The object of social science (as with the natural sciences) is therefore to establish the truth about how the social world operates. In contrast, social action theories regard the social world as consisting of a series of representations. Their aim is therefore to understand subjective meanings, and they regard social action as a product of how individuals interpret the world and interact with others on the basis of these meanings. They also emphasize the fluidity that exists, with individuals having the ability to select, interpret, and bestow meaning upon their interactions with others. However, self-direction is seen as always limited by the actions and expectations of others, with the result that individuals engage in processes of ‘negotiation, impression management and meaning creation’ (Fine 1993).
The social action perspective had its origins in the writings of Max Weber (1864–1920), George Herbert Mead (1863–1931), Husserl (1890–1938), and Schutz (1899–1959). However, its rise to prominence occurred in the 1960s and 1970s and it is now extremely influential within the health field. Like other perspectives, it encompasses different theoretical positions, most notably those of symbolic interactionism, phenomenology, ethnomethodology, and social constructionism. These traditions although all emphasizing the subjective nature of the social world, nevertheless differ in their origins and underlying philosophical assumptions, and give differing accounts of the ways in which social actors make social life. Symbolic interactionism, for example, developed from the work of Mead (1934). This theoretical approach argues that human social conduct has a symbolic character, and what permits human beings to interact and form relationships and society is their ability to understand one another’s gestures and responses. Symbolic interactionism also emphasizes the importance of learning about the culture or subculture of the people under study and therefore frequently uses observation and ethnographic analysis. Phenomenology, which developed from the writings of Schutz (1962, 1964), is primarily concerned with interaction between individuals (for example, between doctor and patient) or within small groups, which is regarded as being governed by a principle of reciprocity of perspectives. This involves assumptions about what each person makes about the other and the means by which social actors construct intersubjective understandings. According to Schutz this is achieved through possession of a common stock of knowledge which is constantly modified and refined on the basis of our experiences and interactions.
Interpretive studies in the health field are sometimes theoretically grounded in a particular sociological tradition. However, much qualitative research is more generally based on the assumptions of a social action approach, in terms of an emphasis on uncovering subjective meanings and the acceptance of multiple realities. A major area of study in the health field has been the meanings held by lay people and the reasons for their actions, with such research often serving to render intelligible what appeared irrational based on a medical perspective. An example is research undertaken to explain the high levels of patient non-compliance. Quantitative studies indicate that on average 40 to 50 per cent of patients treated for chronic conditions are non-compliant with drug regimes. This phenomenon was traditionally viewed in terms of a framework of patient ‘default’, and attributed to patient’s lack of medical knowledge, problems of forgetting, and the complexity of drug regimes. Responses mainly involved giving patients more medical information, providing drug counters and other aids to remembering, simplifying drug regimes, and improving doctors’ communication skills. In contrast, qualitative studies informed by a social action approach acknowledged the existence of multiple realities and explored patients’ own meanings and the reasons for their actions. For example, Donovan and Blake (1992) conducted open interviews with 54 patients attending a rheumatology clinic and also observed consultations. Altogether one-half of these respondents were considered to be non-compliant, with common reasons being worries about side-effects, worries about becoming dependent on the drugs or that their bodies would become tolerant, and worries that the drugs would lose their effectiveness. They noted that rather than ‘defaulting’,
[p]atients carried out their cost–benefit analyses in theory and many put them into practice. Some would decide that the risks outweighed the benefits of the drugs…. In most cases, however, drugs would be taken ‘on trial’ for whatever period seemed most appropriate to patients. (Donovan and Blake 1992)
Patients’ medication use was thus more adequately described as ‘reasoned decision-making’ rather than as ‘non-compliance’. Similar findings across a wide range of conditions and treatments have led to an emphasis on the more active involvement of patients in the consultation. This requires giving greater priority to patients’ beliefs, preferences, and co-operation through patient-centred consultations and achieving a ‘concordance’ between doctor and patient in relation to treatment decisions (Chewning and Sleath 1996). The question has thus been raised of whether the term ‘patient’, which portrays passivity, compliance, and an unequal relationship between the user of health-care services and the provider, is still an appropriate term (Neuberger and Tallis 1999).
Other examples of research that stems from this notion of the more active participating ‘patient’, and makes intelligible what are viewed in medical terms as ‘inappropriate’ patient behaviours, include qualitative research examining ‘problems’ of the uptake or non-uptake of medical services and preventive measures (Kai 1996; Cornford and Morgan 1999), studies of lay definitions of ‘health’ (Williams 1983), lay images of disease and of persons ‘at risk’ (Davison et al. 1991), and the meanings of stigmatizing illness (Scambler and Hopkins 1986). There has been much less emphasis on the study of professionals’ beliefs and practices, although an interpretive perspective and qualitative methods are increasingly employed in organizational evaluations to elicit the perspectives of different stakeholders (see section on professionals perspectives).
So far this section has focused on approaches that may be characterized as subtle realist, and which emphasize subjectivity and the existence of multiple realities (Pope and Mays 2000). However, another sociological tradition adopts an extreme relativist position (radical relativism or ‘antirealist’) and holds that social and human reality are mind-created (or socially constructed realities) and not merely mind-shaped and created through interaction. This tradition is thus based on the assumption that it is not possible to have any knowledge of a phenomenon apart from our own experience of it, and all accounts are regarded as of equal status and validity. This approach is exemplified by the writings of Michel Foucault (1926–1984). A major strand of Foucault’s work is his constructionist analysis of the history of medical ideas in France (Foucault 1976). This involved charting the change from speculatively based medicine of the eighteenth century to the scientifically based medicine of the nineteenth. Foucault explained this change in terms of the emergence of a new ‘clinical gaze’ or way of seeing the patient’s body, rather than forming the result of reasoned development and progress. For example, developments in pathological anatomy and the invention of the stethoscope with the opportunities they offered for viewing the internal workings of the body are regarded as giving rise to a change in the reality of disease. This involved a shift from the galenic humoral theory in which disease is seen as a disturbance in the body’s balance towards the belief in the solid invariate reality of the body, and the status of a disease as a ‘thing’ separated from the self and how the individual feels. Disease entities therefore form products of social reasoning and social practices rather than something ‘real’. To call a set of symptoms ‘bronchitis’ or ‘coronary heart disease’ does not therefore mean that a disease entity exists independently of social context, but is how medical science, in a given time and place with the aid of laboratory tests and theories, has come to define it. All knowledge is therefore contingent, with medical belief systems, like other belief systems, being created as a result of reasonings that are socially imbued. Moreover, Foucault suggests that the way the body is seen, described, and constructed might be called ‘political anatomy’. It is political because changes in the way the body is viewed are based on certain mechanisms of power and wider changes in society. This approach therefore involves the study of language and texts as forms of discourse, which help to create and reproduce systems of social meaning (Tonkiss 1998).
A major critique of radical constructionist (and postmodernist) approaches is that if all knowledge is regarded as socially relative, we have no standards by which to judge the merits of different types of treatment or technological intervention. Indeed, Bury (1986) questions whether in this context the constructionist view itself can have any validity. For example, data showing a relationship between social class and health would be seen from a radical constructionist perspective as merely the effects of particular epidemiological techniques, rather than as something ‘real’ requiring social action. An extreme relativist position therefore means that research cannot derive any unequivocal insights relevant to action, and as a result commands little support among applied health service researchers. Nevertheless, this perspective is important in drawing attention to ways in which interpretations and responses to the sick body can be viewed as fabrications of powerful discourses rather than discoveries of ‘truths’, and as such has been particularly influential in relation to feminist analyses of the body as a key site of oppression. For example, the construction of disease categories (e.g. hysteria) and assumptions about women’s bodies can be viewed as maintaining gender relations and reinforcing traditional roles (Shuttleworth 1990). Similarly, definitions of childbirth as a ‘medical problem’ and associated discourses in terms of risk and clinical safety are viewed as supporting the control of childbirth by the medical profession (Kitzinger 1992).
Structure and agency
A recent development in sociological theory is an attempt to overcome traditional theoretical divisions, and to bring together notions of structure with a social action perspective. Key theoretical approaches are Giddens’ (1984) theory of structuration and Bhaskar’s (1989) critical realism.
The attempt to link structure and agency in the health field is illustrated by the proposals of Popay et al. (1998b) for future research on inequalities in health. They argue that the existing quantitative work in this field consists mainly of unidimensional explanations within ‘risk factor’ epidemiology and do not encompass the full complexity of social processes. Conversely, qualitative studies of individual behaviours (although emphasizing beliefs and actions) tend to be divorced from context, in terms of the structures of power and control in which individuals are embedded. Popay et al. argue that future research should take account of material conditions in terms of the places people inhabit (for example, run-down area of large city, suburban housing area, local health services and recreational facilities, and so on), to explore the ways in which structures affect the dynamics of everyday life. Examples of questions to be addressed, include ‘How do individuals living in the most materially disadvantaged areas of society make sense of and act upon their environments, with what consequences for their health and those they care for?’ and ‘What is the relationship between material risk, individual experience and action at the individual or group level?’ Popay et al. explain that:
[a]ttention to the meanings people attach to their experience of places and how this shapes social action could provide a missing link in our understanding of the causes of inequalities in health. In particular, the articulation of these meanings—which we refer to as lay knowledge—in narrative form could provide invaluable insights into the dynamic relationships between human agency and wider social structures that underpin inequalities in health. (Popay et al. 1998b)
Sociological theory and medical sociology
This section has drawn attention to the diversity of perspectives within sociology and identified some of the current debates. The latter includes the changes in society associated with movement into an era of postmodernity, differing assumptions regarding relativism as exemplified by interpretive and radical constructionist approaches, and attempts to link agency and structure. These developments, together with the increasing prominence of interpretive approaches and qualitative methods, are reflected in work within the health field and provide new insights in relation to both clinical and public health issues as illustrated in subsequent sections (see section on areas of sociological investigation).
For sociologists working in the health field, the distinction between sociology of medicine and sociology in medicine still holds. The former comprises sociologists whose main interest is to advance theoretical knowledge through research in the field of health and illness. The latter comprises sociologists primarily concerned with addressing medical or health service issues, who in the United Kingdom are mainly located in departments of public health or other medical departments. A recent change has been the increasing emphasis on work that may be classified as sociology of medicine. This has stemmed particularly from research that is now referred to as the sociology of the body, which brings together developments in Foucauldian, postmodernist, and feminist analyses, and addresses such issues as sociology’s replication of the mind/body dualism, the influence of the Foucauldian reconceptualization of power, and the conditions of life in postmodernity. One aspect of the latter is the possibility of controlling the body (through drugs, transplants, and so on) in ways that were never possible in the past, but which also raise new issues about what the body actually is. Other issues are concerns with the shaping of the body through lifestyle choices, physical interventions and cultural expectations, the lived body in terms of chronic illness, and political dimensions of the regulation and control of bodies in relation to for example reproduction, death, and the provision of physical capital (Nettleton 1995).
Sociology in medicine also continues as a strong tradition and applies sociological theories and insights to public health issues. Sociologists engaged in this activity sometimes employ quantitative methods, specializing particularly in survey research. Such studies conform closely to epidemiological work, especially in relation to research examining the distribution and explanations of social inequalities in health. However, increasingly sociologists in medicine apply an interpretive perspective and employ qualitative methods to examine clinical and public health issues, particularly in relation to the perspectives and evaluations of patients and consumers. This framework and methodology, although differing from the assumptions and methods of epidemiological research, is now accorded official recognition and acceptance. This is reflected in the recent acknowledgement by medical journals of the contribution of this approach to the understanding of health-care and health services issues (Black 1994).
Sociological investigations employ a range of methods which can be classified as belonging to two main research traditions, those of quantitative and qualitative research. Each is derived from a set of philosophical assumptions, with issues around qualitative research being more debated and discussed than for the former. The logic of each tradition does not apply to the other, although it is common for novice researchers to ask inappropriate questions of qualitative research, which derive from the quantitative tradition. This section therefore defines the key features of these traditions and how they contribute to the understanding of public health. It pays particular attention to qualitative methods, however, as its logic and techniques are less familiar to epidemiologists and other clinical scientists, and involve a radical departure from the assumptions and methods of the scientific model that form the basis of quantitative research.
Quantitative methods derive from a positivist philosophy, of which a key element is the view that social phenomena are objective and external to the individual and thus take the form of social ‘facts’. It will be suggested later in the chapter that there is a move in the sociology of health and illness towards greater use of qualitative methods, but it remains the case that there is a solid contribution to public health by social scientists using quantitative techniques.
The model for quantitative research derives from laboratory-based science. Key features include the assumption that there is no relationship between the researcher and the researched, that the ‘social facts’ remain constant over the study period, and that the well-conducted study will have high reliability (repeatability) and validity. Taken out of the laboratory context, this set of assumptions provides difficulties for researchers in the public health and social science fields who wish to carry out good quality research in a social environment. The problems hold true for many different methods, including designing randomized controlled trials, which although ‘classically’ designed to test drug treatments are sometimes employed to evaluate the introduction of a new form of service. Thus, among other things, researchers pay special attention to problems associated with reliability and validity, and issues of bias in the data collection.
Data employed in a quantitative study derives from a number of sources. Some may be drawn from ‘secondary’ sources; for example using routinely collected hospital statistics, government surveys, and census data. Social scientists, however, are commonly involved in the collection of primary data using a variety of methods, such as randomized controlled trials, health needs assessments, and large-scale surveys. Sociologists bring to such research a knowledge of the meanings and significance of concepts such as social class, gender, ethnicity, and age, which are of value in examining and interpreting health inequalities, variations in uptake of services, or variations in the incidence of particular disease conditions between males and females. Sociology has also contributed other concepts and interpretations in studying the social determinants of health and health outcomes, including relative (and absolute) poverty and social exclusion, social networks and support, lay beliefs about health and illness behaviours, and the possible effects of occupational/social mobility and diagnostic decision-making on observed patterns of disease.
The quantitative tradition is based upon the scientific method, which is distinct along a number of dimensions to qualitative methods, as set out on Table 1. Studies using quantitative methods require the research question to be predefined, so that the study tests a precise hypothesis or set of research questions. It is also essential at this stage to define the variables, such as social class or ‘working women’, since these variables have to be translated into categories that can be incorporated into the study design (for example, the questionnaires) and which have meaning for the study. Thus to operationalize the category of ‘working women’ in a study of the effects of pregnancy on women in paid employment, the researchers would have to decide where to set cut-offs for hours worked per week (for example, up to 10 h/week, 11 to 20 h/week, and so on), or to simplify matters, perhaps to make the distinction between women working full or part time. The appropriate definition of key variables is crucial to the success of the study. For example, researchers may wish to draw up their categories in the knowledge that 17.5 h is defined as half-time employment, and that a category of 11 to 20 h would cut across this important distinction. By limiting the study to working women the researchers should be aware that they will miss the inclusion of various categories of women who carry out unpaid work (housewives, women working unpaid in the family business, and so on) but for whom the health impact may be similar to those with paid work.
Table 1 Quantitative and qualitative research
Setting out the research questions and defining the independent variables are critical stages in the research process, and these allow no second chances if poorly defined. Thus in order to carry out a quantitative study, the researcher(s) must be familiar with the field of study. For example, it is important to be aware of all the options that are available to the subject responding to the questionnaire, since lack of fit between the subject’s preference and the options set out in the questionnaire decreases the validity of the study. It is normal to pilot any questionnaire, sometimes twice, to ascertain the validity of the questions.
Most quantitative research in this field requires collecting ‘primary’ data using a self-completed or interview-administered questionnaire. The questionnaire is a central part of the researcher’s armoury in the maintenance of reliability since all respondents will receive the same questionnaire and the same letter of introduction. As importantly, the measurement (for example, alcohol consumption, attitudes towards pain) should work in a consistent way with each individual. If the questionnaire is interviewer administered, the researcher should be trained to check that all procedures are systematically carried out with all respondents to reduce possible biases within the data collection. Questions are generally precoded and allow only a limited range of responses, while some studies also incorporate prevalidated measures including psychometric tests (for example, measuring anxiety or depression) and other measures relating to social support, quality of life, health status, and so on (Bowling 1991).
Questionnaires vary along several continuums. They may be sent through the post (self-administered), or carried out with a trained interviewer administering the schedule (for example, on the doorstep or in the shopping mall, or, increasingly, by telephone). The method selected will depend in part upon the research issue, intended sample size, and funding available. All approaches have their benefits and disadvantages. Response rates may vary depending on the approach, the group being questioned, and their interest in the topic. With the increasing number of studies being carried out, some health service staff and patients are in a danger of being over-researched. It is therefore important to make the questionnaire clearly laid out, as brief as possible, and of interest to the respondents to increase the possibility of a high response rate (issues of questionnaire wording, sequencing, appearance, and enhancing the response rate are dealt with clearly by McColl et al. (1998)). Sampling is a complex issue; researchers often wish to gain a representative sample of a population (since total population sampling is usually out of the question) but there are a number of routes to achievement of the most appropriate sample, which are discussed by Arber (1993) among others.
The key to the success of quantitative research is the ascertainment of statistical relationships between variables. Analysis may be complex, and requires statistical knowledge, although computer packages such as Minitab and the Statistical Package for the Social Sciences (SPSS) can aid the process. However, the choice of statistical tests is affected by the distribution of the data and it is important that social scientists either become statistically literate and/or work with statisticians.
Quantitative research in sociology thus conforms to the broader principles of scientific research shared by public health and other clinical sciences. It assumes that there is an objective reality which can be measured, and that studies are reliable (repeatable) and valid. Many public health issues have been tackled with the survey approach uniting sociologist and epidemiologist. Sociologists have less often become involved with other quantitative approaches such as case–control or cohort studies. It is increasingly common, however, for those designing randomized controlled trials to acknowledge the importance of measuring ‘consumer satisfaction’ or ‘professionals views’ of the intervention and to co-opt a sociologist onto the project (for example, Tucker et al. 1996; Turnbull et al. 1996). While this is not ideal, it does ensure that randomized controlled trials include these concepts in their design, as well as incorporating prevalidated measures of health outcome based on health status or quality of life measures.
Qualitative research is the methodological tradition used more often by interpretist or social action perspectives described above. These methods have a long tradition in sociology and characterized many of the early sociological studies undertaken in the United States in the 1940s, as well as forming the general methodology of anthropological research. However, acknowledgement of the legitimacy of the approach in the medical field is more recent. A series of articles published in the British Medical Journal in the summer of 1994 (subsequently published as an edited collection (Mays and Pope 1996)) could be seen as a marker of acceptance. More generally, with a broadening definition of public health to include topics such as social exclusion, social networks, and other social determinants of health, has come recognition of the contribution of subjective experience to health planning and policy. Today qualitative methods are frequently employed in clinical and health services research, and in a wide range of community health studies and health needs and health impact assessments. These approaches, where the respondent’s beliefs and experiences are prioritized, are appropriate where there is a gap in the existing literature and where a new line of thinking needs to be established. This may relate to an understudied group or population, or where there is a need to develop a greater in-depth understanding of an issue.
There are several main sociological traditions that have an affinity with qualitative research, and are separated by subtle differences in analysis and emphasis. Whilst it is not appropriate in this chapter to provide an in-depth discussion of the methodological approaches associated with each tradition, the key features of qualitative research will be discussed.
Qualitative approaches are most commonly field based and include interviews, focus groups, and observational work—’words and actions’ (Lofland and Lofland 1995). Secondary source analysis may also be included, such as library-based research drawing upon discourse analysis, and historical and documentary research. Finally, action-based methods in terms of process evaluation and action research are now considered valuable contributions to public health research.
Qualitative research (contrary to the assumptions of quantitative methods) acknowledges that there is always an interaction between the researcher and the researched. Researchers in this tradition are trained to be more reflexive about their role, reflecting critically upon the ways in which interaction with the respondents/field situation may influence the data collected. Researchers continue to debate the extent to which commonality between researcher and researched may enhance data collection. Some have argued that shared gender (for example, woman researcher/woman respondent) is highly influential in establishing trust between the two (Oakley 1981). Others have suggested that the supremacy of gender should not be assumed and that other factors, for example, ‘ethnicity’, may be more important than gender to some respondents (Douglas 1992). The essence is that these potential influences are acknowledged and discussed as they may ultimately affect the nature of the interview and the data collected as a result.
Interviews are a central tool of the qualitative researcher, although it is important to be wary of the increasing use of the term ‘interview’ to describe a questionnaire-designed study in which the interviewer has talked the respondent through the schedule. It should also be noted that the key to identifying a qualitative study is the data analysis, which for a qualitative study would usually be in the form of thematic analysis from interview transcripts.
Interviews in qualitative research may take a number of forms, although most often used is the semi-structured (or semi-standardized) interview. In this form of interview the researcher starts with a list of topics and questions to be asked at every interview (often identified through earlier pilot interviews), although the order may alter and the researcher may probe for more information and follow-up a line of thinking with a respondent. At the other end of the ‘structure’ continuum is the non-standardized or unstructured interview, representing what Lofland and Lofland (1995) call ‘the guided conversation’. This form of interview gives considerable flexibility in terms of the range and order in which issues are discussed, and is particularly useful as a strategy for investigating new subjects where respondents’ views are unknown.
Interviews in qualitative research are usually tape-recorded and subsequently transcribed. The process is lengthy; interviews may last from 30 min to 2 or 3 h with a respondent but it may take as long as 5 to 10 h to transcribe and check the transcription. This process contributes to the perception of qualitative research as slow and time-consuming. Analysis (see below) is also a lengthy process.
Personal interviews are the most common method of data collection in qualitative research and have been employed with increasing frequency to study a wide range of issues of importance to public health. These include charting the meaning of illness (Kelly 1992), analysing the ways in which low-income families construct their daily diet (Dowler and Calvert 1995), and examining what individuals think about risk (Hart and Flowers 1996).
Brief mention has already been made of the sample in qualitative research. Numbers involved in interview studies are usually relatively small compared with sample sizes associated with quantitative studies. Indeed, the underlying assumptions of sampling is quite different, with the aim being to sample in such a way as to achieve a good representation of those characteristics thought to be important for the study. There are different sampling techniques, including purposive or systematic sampling, which is the deliberate choice of respondents, subjects, or setting to represent a wide range of opinion or experience (for example, one might take a consecutive sample of attenders at a clinic or a general practice). Theoretical sampling is based on a previously developed hypothesis or theory (for example religion is an important factor, so sampling would take place with attention to religious affiliation). A third type of sampling, ‘snowball’ sampling, is used when the individuals under study are difficult to access through normal routes (for example, studying homeless individuals) and respondents may therefore introduce other potential respondents.
Focus groups form an alternative to the personal interview. They have been employed by sociologists since the 1950s, but became known primarily as the tool of market researchers where they were used to carry out rapid checks on public attitudes (for example, in relation to consumer goods or political attitudes). However, focus groups have recently achieved a new status as a valid research tool (Barbour and Kitzinger 1999). The method brings together a group of individuals (usually six to eight), often selected with reference to age, gender, and other criteria, to discuss a series of topics, using the group or collective interaction to produce insights. Focus groups are regarded as having particular application for studying, among other things, ethnic groups, sensitive topics (Kitzinger 1994; Kitzinger and Farquer 1999), and organizational change (Barbour 1999). They have also found a niche as a research tool for investigating consumer views, as demonstrated in studies by Schwarz et al. (2000), Lanza and Ericsson (2000), French and Chopra (1999), Lansbury (2000), and Williams and Clark (2000). In all these examples, and especially the first two, focus groups are seen to provide valuable data for use in the development of health policy. Several of these studies also served as a first stage in the development of a larger study to generate hypotheses based on informants’ insights. For example, Lanza and Ericsson’s consumer-led focus groups on clinical guidelines preceded a survey on the topic.
With a skilled facilitator (who may use a number of techniques to encourage participation from group members), focus groups can provide rich data and elicit not only the views and experiences of group members but also areas where conflicts may arise and disagreements are expressed. The importance of trust has already been raised in relation to interviews and it is equally true here. Respondents are invited to participate in a focus group, sometimes with a small payment as an inducement. To facilitate free discussion and communication during the short time of the group, the researcher must establish trust between himself or herself and members of the group, emphasizing the confidential nature of the study and ensuring that individuals understand the purpose of the group (Cunningham-Burley et al. 1999).
Focus groups suffer from a number of drawbacks, including the evident ‘self-selection’ of individuals who may be willing to turn up to discuss their views and beliefs with a series of strangers. Many who feel marginal or who lack confidence may simply not respond to the invitation, or be reluctant to express their views, leaving the group dominated by the ‘socially included’ in society. One way to avoid this difficulty is to use an existing group as the basis for the focus group, such as a group of pensioners, mother and toddlers, or a support group who meet regularly and who are familiar with each other (Kitzinger 1994). However, it is important that ‘gatekeepers’ do not select individuals seen to be appropriate for the exercise.
As Hammersley and Atkinson (1997) note, ‘all social researchers are participant observers’. All qualitative researchers may use observation as an adjunct to other methods of data collection, for example, when visiting an interviewee in their home or their work place, the ‘observant’ interviewer would note the context of the interviewee’s life. However, participant observation (or observation, since not all researchers have the opportunity to become actively participant in the social context under enquiry) is a specific method with its own logic and rules. Participant observation is based on the structured observation of social life as it occurs; or, to put it more formally, ‘participant observation refers to the process in which an investigator establishes and sustains a many-sided and relatively long-term relationship with a human association in its natural setting for the purpose of developing a scientific understanding of that association’ (Lofland and Lofland 1995).
Researchers using this method are based within the context under study, spending as much time as possible becoming familiar with the lives of the respondents. Research using this method is often described as ethnographic. Ethnographic research is traditionally concerned with the study of culture and cultural meanings, although Hammersley and Atkinson (1997) emphasize the looseness of definition of the term ‘ethnography’. In keeping with its anthropological origins, there is a strong emphasis on first-hand field study, that is, research involving considerable periods based in social contexts and settings, and involving informal interviews, discussions, and observation. Ethnographic methods relating to public health issues have been usefully employed to understand students training to be health professionals (Atkinson 1981), working-class family life in London (Cornwell 1984), and the culture of drug takers (McKeganey and Barnard 1992). Part of the process of establishing credibility with the researched group is learning through rigorous observation as well as informal interviews what the important values are of the group. As with other methods, the researcher sets out to gain the trust of the researched. Thus McKeganey and Barnard (1992) spent a year ‘in the field’ in their ethnographic study of street-workers in Glasgow. Pavis and Cunningham-Burley (1999), exploring street culture among young people in East Lothian, reported 70 h of participant observation.
Observational data collected in the form of field-notes (as in anthropology) is often accompanied by additional data collected through interviews and possibly questionnaires. Thus Pope (1991), in her study which examines the process by which hospital waiting lists are created and maintained, reports a range of methods, although the data collected through observation are key to her analysis and explanation of the waiting list phenomenon. In such situations the researcher often does not know in advance the area and issues to focus upon, and therefore progressive focusing takes place. Thus after reviewing the initial field-notes, the researcher may decide to ‘focus down’ on specific areas of investigation. Although observational studies provide valuable data on actions tied to particular situations and circumstance, they have the drawback of being labour intensive and expensive.
The term ‘action research’ is normally attributed to Lewin (1946) who defined action research as ‘a way of generating knowledge about a social system while, at the same time, attempting to change it’. Meyer (2000) brings out its particular qualities, and describes action research as ‘a style of research rather than a specific method’, and one which contributes simultaneously to social science and social change. Three key features of action research are identified as its participatory character (requires participants to perceive a need for change and be willing to play an active part in the research and change process), its democracy (requires participants to be seen as equals with the researcher and to be consulted on the action process and methods of evaluation which is negotiated with participants), and its contribution to both social science knowledge and local change (the former requires that the report is presented with considerable contextual detail to aid generalization).
An action research approach is most common in community work. However, it has also been employed in identifying problems in clinical practice, helping to develop potential solutions, and facilitating change within the health services, especially by members of the nursing profession (Titchen and Binnie 1993; East and Robinson 1994). East and Robinson (1994) have described a 2-year action research project that aimed to facilitate the management of change in a district general hospital. The study brought out clearly the differing agendas of the participants both within the hospital (demonstrated by a case study of the nurses’ experience of change at ward level) and those of the clinical nursing staff and the general managers. Similarly, a study by the Royal College of Physicians in England is exploring the roles of clinicians, clinical audit staff, and managers in implementing clinical audit and ways of overcoming organizational barriers to audit (Berger 1998).
Action research is not tied to a particular methodology and may use a number of different methods to capture the process of change and to evaluate the nature of any changes that have occurred. However, it generally employs qualitative methods with data collected through interviews and observation as well as documentary sources. There is a move towards stressing the value of documenting the experience of change, and many of the more exciting action research projects in nursing, education, and social work have incorporated this element.
P>A ‘mixed method’ study design is increasingly common and viewed as desirable, with each method drawing out distinct features of the issue under study. Thus a questionnaire study may identify the spread of views on an issue, while an additional set of interviews with a sample of subjects might provide a greater understanding of why different views are held. The concept of ‘triangulation’ is often used in this respect, with the researchers using multiple and different methods which may involve a mix of interviews, observation, and routine data to approach the same issue. Triangulation is a form of validation in research, with the assumption that the emergent findings using one method will support those achieved from another method in corroborating an overall interpretation. However, the way in which methodologies from different traditions actually combine together as a form of validation is still debated.
Data analysis and validity
The question is not whether the data are biased but to what extent [the researcher has] rendered transparent the processes by which data have been collected, analysed and presented. (Popay et al. 1998a)
Data analyses are an on-going feature of qualitative research, bound up with the collection and processing of the data from the study. The iterative process of data collection and hypothesis testing is explicit in ‘grounded theory’ but is essentially part of all qualitative methods. The raw material for qualitative researchers is words; as ‘data’ these can be in the form of transcriptions from (tape-recorded) interviews or focus groups, notes written by the researcher(s) from interviews and field-notes in which they have recorded their fieldwork experiences, and occasionally diaries kept by the respondents themselves.
From early on in the study, the researcher will begin to think about concepts that seem important, useful, or in some way to provide insight into the respondent’s experiences. At the beginning of analysis, the researcher will identify a set of codes or categories, which derive from the data. Some of the codes may be practical (for example, family relations, descriptions of heart attack, contact with general practitioner), but others may be more abstract such as ‘independence’, ‘coping strategies’, ‘bad days’, and so on. Within an extended set of codes it may be appropriate to construct subdivisions, such as dividing up the types of ‘coping strategies’ or different types of communication with the general practitioner, if the data provide examples of such divisions. There is no hard-and-fast rule about coding the data, some preferring quite complex systems and others starting with quite a broad basic framework. The important—and distinctive—feature of most qualitative research is that the data suggest the codes, rather than the researchers working to a prior hypothesis of what they wish to examine. The researcher will then apply this coding system to all the transcripts, either by hand, or by using one of a number of computer packages available to aid this process. These packages can be time-consuming to master but generally help the next stage of the analysis, which is to develop themes from the codes. These themes may be based around one or more codes; thus one theme in a study of patients with heart disease could be the different ways in which patients manage their medication taking. Much research rests at this point in the analysis and remains quite ‘literal’, reporting at a fairly concrete level of what the data have shown. However, many qualitative researchers will try to create some kind of ‘higher-order’ theory from their research and to draw out at a more abstract level of meaning from the data. Thus in our example, the researchers may look at notions of patients’ independence and patient strategies to remain self-monitoring of their condition. Similarly in a study of the experiences of relatives of critically ill patients, Walters (1995) summarized the findings into four themes or sets of concepts, ‘making sense’, ‘plain talk’, ‘being with’, and ‘seeing’. At an even higher level of abstraction one could see whether these concepts might apply to other conditions and situations. By so doing, it creates opportunities for the research to achieve greater generalizability, that is, generalizability at a theoretical level.
Grounded theory as a form of analysis was developed and elaborated in the 1960s by Glaser and Strauss (1965, 1967). At that time it was proposed as a form of analysis in which data analysed inductively in a rigorous manner resulted in the production of substantive and formal theoretical statements. It is fair to say that since then the term has been widely misused and that even its original proponents themselves do not agree about the process (see Melia’s thoughtful discussion of the varying approaches to grounded theory (Melia 1997)). Grounded theory essentially refers to theory which has been ‘worked up’ from an iterative process, reviewing the data to derive hypotheses which are then tested in the field, repeating the process with revised hypotheses until the theory is seen to be supported fully within the data. Although many researchers report that they use this method, true grounded theory demands a commitment to data analysis and time in the field that is beyond the limit of many researchers.
The issue of validity in qualitative research most sharply distinguishes these methods from quantitative research. The assessment of the quality of qualitative research remains the focus of debate (Hammersley 1990; Murphy et al. 1998) and views of those within the field diverge. Some dislike the term ‘validity’ with its connotations of ‘being able to establish the truth of any research beyond all possible doubt’, and prefer to suggest that the way in which the soundness of study should be assessed is by using concepts of plausibility, credibility, trustworthiness, and the nature of generalizations made from the data. Ways of establishing ‘validity’ are difficult to summarize succinctly, and are subject to debate over form as well as terminology (Murphy et al. 1998; Seale 1999). Clarity of exposition of method is the key. However, the increasing publication of qualitative research in journal article form with their assigned word limits makes it difficult to provide a full description and reflection on the methods employed.
Areas of sociological investigation
The sociology of health and illness covers a wide and diverse range of topics. For example, a recent textbook covers the following topics: social construction of medical knowledge, lay perceptions of health and illness, the experience of health and illness, the sociology of the body, interactions between patients and health-care providers, and the patterning of health and illness in relation to social class, gender, race and ethnicity, and other social groupings, and the social organization of both formal and informal health care (Nettleton 1995). A number of authors have attempted to classify the concerns of sociology based on some organizing principle. A useful example is Turner’s (1987) level of analysis approach. He argues that a comprehensive sociology of health and illness must involve the study of health and illness at three levels: the ‘individual’ level (which examines perceptions of health and illness), the ‘social’ level (which examines the social creation of disease categories and health-care organizations), and the ‘societal’ level (which examines health-care systems within the political context).
This section provides examples of the application of sociological perspectives and methods of investigation in relation to three areas—chronic illness and disability, ethnicity and health, and professionals’ perspectives. The aim is not to provide a comprehensive review of these topics, but rather to illustrate the ways in which different perspectives and approaches focus attention on different issues and types of explanations, and to demonstrate the range of sociological concerns and methods of research.
Chronic illness and disability
The social dimensions of chronic illness and disability have formed a major area of study by public health specialists and sociologists from the late 1960s. This reflects the increasing prevalence of chronic disease and the opportunities such conditions offer for examining the social aspects of sickness. Research by sociologists has involved both collaborative studies with public health colleagues concerned with the conceptualization and measurement of ‘disability’, and qualitative research to investigate individuals’ personal experiences and strategies in relation to chronic illness.
Conceptualization and assessment of ‘disability’
An early stream of sociological research examined the consequences of chronic illness for both patients and families (Blaxter 1976). In addition, social scientists collaborated with clinicians and public health specialists in assessing the prevalence of disability in the population and needs of disabled people (Jefferys et al. 1969). These bodies of research informed the development of the World Health Organization’s (WHO) model of the consequences of chronic illness—Classification of Impairment, Disabilities, and Handicaps (WHO 1980). This model provided a consistent terminology in the field and formally acknowledged that the consequences of chronic illness may extend from the medical to the social sphere. The WHO model distinguished three categories: (a) individual impairment, in terms of abnormality in the structure or function of the body whether through disease or trauma; (b) disability in everyday life (that is, restriction in ability to perform tasks, especially those associated with everyday life and self-care activities); (c) the experience of handicap, which refers to the disadvantage for a given individual, arising out of impairment and disability, that limits or prevents the fulfilment of a role that is normal (such as handicap in relation to work, education, or social activities). Impairments often lead directly to disability and may in turn place an individual at a disadvantage (handicap) in one or more spheres of life. However, the WHO model recognized that an impairment may not cause any disability (for example, severe facial scar, or controlled epilepsy) but nevertheless lead to considerable disadvantage or ‘handicap’, reflecting the societal meanings of the condition and consequent feelings of ‘stigma’ and shame.
The WHO model provided a framework for community-based studies undertaken by sociologists and others to examine the impacts of disability, including research examining the ways in which material and social circumstances may mediate the relationship between impairment, disability, and handicap (Locker 1983; Patrick and Peach 1989). Policies relating to the planning and provision of social care, community services, and financial benefits for people with disabling conditions also led to considerable emphasis on the measurement of the prevalence and severity of ‘disability’ and the conduct of national and local community-based studies. However, over time official definitions and measures of disability have increasingly broadened and now define greater numbers of people as ‘disabled’, thus underlining the social construction of this category (Martin et al. 1988).
The dominance of the WHO model has recently been challenged by the emergence of a ‘disability movement’ and the writings of academic disability theorists. These groups adopt a radical sociopolitical perspective derived from a conflict model, and criticize the WHO framework for being an overly ‘medical’ (or individualistic) model, as it locates the ‘problem’ of disability in terms of individual impairment (Oliver 1996). In contrast they depict disability as a social phenomenon, arising from material and cultural forces that effectively ‘dis-able’ people. As Barnes (1991) explains:
In our view, it is society which disables physically impaired people. Disability is something imposed on top of our impairments, by the way we are unnecessarily isolated and excluded from full participation in society. Disabled people are therefore an oppressed group in society.
The social model thus regards disability as socially created and the product of the ways in which society is organized and its exclusionary practices, rather than the product of individual disabilities. For example, the built environment generally assumes that people can climb stairs, see instructions, and take extended steps on buses. The social model of disability therefore provides a radically different perspective to the traditional framework in emphasizing the need for social action rather than individual interventions. The former includes the achievement of improved rights, greater social acceptance, and changes in the material and social conditions of life that increase independence and opportunities for participating in the mainstream of society. The social model also requires that the measurement of disability should focus on the extent to which environments are disabling, rather than the assessment of individual’s functional abilities (Abberley 1992).
Bury (1996, 1997), a medical sociologist who has written extensively in the field of disability and chronic illness, has drawn attention to what he regards as an overemphasis of the social model on the societal determinants of disability, with insufficient attention given to the varying experiences of people with different types of impairments and the significance of their underlying medical condition. For example, pain and associated activity restrictions may form an important aspect of some people’s experience of chronic illness. French (1993) (herself part of the group of disabled disability theorists) also observes that there can be limits to the extent to which restrictions may be removed through social measures (such as the problems she experiences in interaction as a consequence of being blind). The social model has also been criticized for treating disabled people as a socially homogeneous group, with little attention given to the social meanings and effects of gender, ethnicity, and other social characteristics (Lonsdale 1990).
Despite such criticisms, the social barrier model has been important in directing attention to the importance of environmental barriers in limiting activity and participation. This model also raises issues of the control of professionals over what is regarded as a dependent and devalued client group, through defining the problems of this group in their own (medical) terms. Social barriers theorists thus challenge the goal of medicine in terms of achieving ‘normality’, and instead emphasize the need for greater acceptance of diversity and for the involvement of disabled people themselves in determining policies and priorities (Oliver 1996). The social model also stresses the importance of individual autonomy and choice, rather than seeing disabled people as passive recipients of services, and emphasizes the need for greater acceptance of the ordinary needs of disabled people in terms of medical care, social activities and family life, and so on.
The social model of disability is increasingly reflected in work regulations, building regulations, and other public, health, and social policies (Campbell and Oliver 1996). This conceptualization is also informing a revision of the WHO International Classification of Impairments, Disabilities, and Handicaps (ICIDH-2) which is described as providing a synthesis of the medical and social approaches to disablement in terms of a ‘biopsychosocial’ model (Bickenbach et al. 1999). This new model conceptualizes each dimension of disablement (impairments, activity limitations, and participation restrictions) as an interaction between intrinsic features of the individual and that person’s social and physical environment. It also accords each dimension equal significance as representing different and independent facets of disablement, and recognizes that each requires different social responses or interventions.
This analysis of formal models of disability thus illustrates the ways in which accounts are often partial, and emphasizes different aspects of social reality. It also demonstrates the way in which the dominant framework serves to influence the nature of health/social policies, the formulation of research questions, and associated measurement issues.
Individual meanings and experiences
An interpretive perspective has informed a second strand of sociological research. This has involved in-depth interview-based studies of the meanings of chronic illness and disability for individuals, and the ways in which such meanings are constructed, negotiated, and transmitted, and often reconstructed through the course of illness. Early work focused on the crisis created by major disability. Examples include Davis’ classic study of childhood polio, entitled Passage Through Crisis (Davis 1963), Scott’s study of how people learn to be ‘blind people’ (Scott 1969), and Voysey’s study of parental experiences of children with disabilities and the problems of coping with uncertainty (Voysey 1975). More recently, a number of studies have provided a temporal framework to take account of the emergent character of chronic illness, and followed the pathway from the initial disruption of illness, to the process of explanation and legitimation, treatment, and adaptation.
A key concept describing the impact of chronic illness is Bury’s (1982) notion of ‘biographical disruption’ that derived from his in-depth study of people with rheumatoid arthritis. Biographical disruption recognizes that the onset of chronic illness represents an assault both on the person’s physical self and on his or her self-identity. This occurs at two levels of meaning. One is the consequence for the individual of the onset of disruptive symptoms on everyday life at home or work. Studies of specific disorders such as multiple sclerosis (Robinson 1988), diabetes (Kelleher 1988), and ulcerative colitis (Kelly 1992) demonstrate how symptoms disrupt the flow of normal everyday life and introduce considerable uncertainty into it. Patients therefore expend considerable effort in managing their symptoms and limiting their practical consequences. For example, Kelly’s (1992) study of ulcerative colitis sufferers drew attention to their problems of the unpredictability of symptoms, especially diarrhoea, who may therefore manage this by avoiding eating and drinking in public. Similarly, people with epilepsy try to avoid stress and other situations that increase risks of a fit. More generally, people seek to establish a way of life with symptoms or disability, often involving changes in daily activities and making tasks more manageable, while still leading valued lives. A second level of biographical disruption arises from the symbolic significance of chronic illness in terms of its implications for self-identity. For example, disfigurement and deformity associated with skin conditions and arthritis can have a major impact on self-identity, especially for women given the cultural value placed on attractiveness (Londsdale 1990). In addition, many conditions including epilepsy, mental illness, and cancer of the bowel are surrounded by negative social meanings or ‘stigma’, and result in feelings of shame and fears of avoidance and exclusion. In-depth studies of people with particular chronic conditions (Macdonald 1988; Scambler 1989; Pinder 1990) have identified a number of strategies that people may adopt to avoid or minimize these negative social responses in interaction with others. Sufferers also need to come to terms and ‘make sense’ of their condition. An in-depth study by Williams (1984) of 30 arthritis sufferers indicated that this was achieved through a process of what he termed ‘narrative reconstruction’. This involves people making sense of the onset of their illness through reorganizing their own biography. Williams suggests that this process enables people to reconstruct a sense of order from the fragmentation produced by chronic illness, and also notes that clinical explanations are often inadequate for this purpose.
The success of the individual’s adaptation to chronic conditions is demonstrated by a recent survey that indicated that 54 per cent of people with serious disabilities reported that they experience a good or excellent quality of life (Albrecht and Devlieger 1999). Factors associated with such positive assessments were identified as people’s ability to understand their condition, take control, and introduce order and predictability. Successful adaptation may be enhanced by an individual’s ability to mobilize supportive social ties and material resources (Locker 1983; Morgan et al. 1984), although the effects of these social circumstances have been given relatively little attention.
For the future, Kelly and Field (1998) suggest that chronic illness identities may become less coercive and constraining than at present, with the development of a more fluid and open society in the postmodern world, a greater tolerance of difference, and a kind of relativism in the aesthetics of appearance. Moreover, the continually changing and uncertain nature of postmodern society means that chronic illness may become just one more burden an individual has to contend with rather than a central aspect of identity. Another important change is the increasing challenge to the hegemony of medicine and the greater range of discourses available to draw on in making sense of chronic illness. These include those associated with various types of alternative and complementary medicine, local and national self-help organizations, and entertainment and the mass media. Complementing these changing patterns of social life are advances in effective drug regimes and surgery, which may increasingly help to limit and contain the disruption and discontinuity associated with chronic illness.
Ethnicity and health
In the 1970s sociologists in the United States carried out small-scale studies of different ethnic groups (for example, Jewish and Italian immigrants) and demonstrated the way in which concepts such as pain, illness, and health experiences were culturally created and expressed (Zola 1975). The inclusion of ‘race’ as a variable in sociological and epidemiological research remains important in the United States, and represents a key form of stratification within the society. In the United Kingdom by contrast, research has concentrated on the majority culture. Until the 1980s very little social and biomedical research acknowledged the ethnic diversity of the United Kingdom population or explored issues such as health status in relation to ethnicity. By the 1980s, sociologists began to investigate the health experiences of different ethnic groups, with this research generally being qualitative, small scale, and in-depth. Within the most recent decades, large-scale surveys have also contributed considerably to the knowledge of the health patterning of specific ethnic groups (Marmot et al. 1984; Balarajan and Balusu 1990; Nazroo 1997), with the last of these notably bringing a sociological awareness to the investigation at this level of ethnicity and health by introducing concepts such as poverty and socio-economic status, and discussion about the relevance of formal measure of class to ethnic minority groups.
As with disability research (see above), the emerging literature on ethnicity and health contains a number of highly debated topics, with researchers disagreeing on terminology, focus, and approach. This section can only briefly identify the key issues, note the areas of contention, and identify ways in which sociology can contribute to the further development of studies of ethnicity and health. What is evident is that sociological research has played a major role in the understanding of factors affecting the health of ethnic minorities within the United Kingdom, and also that a large research agenda remains to be developed.
Terminology in this sensitive area is important, ethnocentric, and ever changing. As a concept ethnicity—in theory—can be applied to members of any group since it may be understood as ‘cultural practices and outlooks that distinguish a given community of people’ (Giddens 1989; Smaje 1995). In reality, however, the study of ethnicity in the United Kingdom generally refers to specific groups, not usually (as in the earlier American research) to Jewish, Italian, or Irish groups (although there are exceptions), but most frequently to the more recent migrated ‘black’ ethnic minorities. The largest groups among the latter are South Asian (Indian, Pakistani, Bangladeshi), African-Caribbean (previously termed ‘West Indian’, then ‘Afro-Caribbean’), and Chinese, together constituting about 5 per cent of the United Kingdom population.
Problems exist in the allocation of individuals to ethnic groups. Previously, this has been crudely undertaken in the United Kingdom using country of birth of the individual (and their parents). Only in 1991 did the United Kingdom census include a question on self-assessed ethnicity. However, this classification, which identified eight main ‘ethnic’ groups, was problematic in a number of respects, including the mixture of ethnic and racial categories and an assumed homogeneity within the ‘white’ group (represented by a single category to include people from Cyprus, Turkey, the Republic of Ireland, and other areas). Considerable numbers of people also allocated themselves to ‘other’ categories. More generally, the approach to ethnic classification adopted and the categories identified serves to construct official meanings of ethnicity. As Bhopal (1997) observes, the variations in definitions employed make cross-national and intercountry comparisons difficult.
Social action theory underlies much of the early qualitative research carried out with ethnic minority groups. Research studies have described the cultural beliefs, experiences, and lifestyles of individuals and groups from ethnic minorities now established in the United Kingdom. To take a few examples, Homans (1983) studied food choice among young pregnant Asian women living in Britain, exploring amongst other issues notions of ‘hot’ and ‘cold’ foods. Work by other female researchers identified aspects of South Asian life that had previously remained hidden. This includes research by Currer (1986) and by Fenton and Sadik-Sangster (1996) on mental health concerns of women of South Asian origin. These studies highlighted the culturally specific nature of diagnoses and challenged psychological research, which suggested (using psychological tests) that the rates of psychological distress and psychiatric admissions were lower for those of South Asian origin compared with the white population. More recently, Bradby (1997) carried out research with young South Asian women in Glasgow to investigate their understanding about diet and foods, prompted by concerns about the role of diet in the higher incidence of coronary heart disease among South Asians. Bradby spent time during the research period visiting and informally talking to her respondents as well as carrying out formal interviews. Her research demonstrates the value of a qualitative approach in exploring strongly held cultural beliefs. Bradby proposed that there was a conceptual division in her sample’s perception of food groups, between what was seen as ‘our foods’ (a systemic or Ayurvedic model of food which is based on assessing the merit of different foods in terms of an equilibrium), and ‘their foods’ (the Western-based model which breaks foods down into their components such as proteins, vitamins, and so on) and which judges the value of foods on this basis. Bradby argues that health promotion activities need to take into account cultural understanding of such beliefs to be effective in influencing behaviours and practices. Overall, the importance of these studies lies in their attempt to understand and respect different cultural groups and to explain behaviours previously hidden or misinterpreted. Researchers have tended, however, to underline the differences in culture rather than to explore the similarities.
The issue of racism within the health services is a very difficult area to research, for as Pearson (1989) notes, less favourable treatment ‘is not demonstrated through hard outcome measures or performance indicators’. Observational data and interviews with midwives contributed to Bowler’s (1993) analysis of the stereotyping of Asian women attending for maternity care in a United Kingdom hospital. Midwives comments in interviews, but as importantly as revealed through observational work, reflected their views that Asian women lacked communication skills and were not ‘compliant’ in attending for care. Their behaviour in hospital (for example in labour) was also interpreted as being antisocial. The negative stereotypes that became attached to Asian women indicated a lack of understanding of cultural variation and resulted in women receiving less helpful care.
A second strand of research has employed quantitative techniques and has been underpinned by a structuralist approach, although it is difficult to trace the precise theoretical origins of some work. In the 1980s and early 1990s there were a number of localized small-scale studies focusing upon the uptake of certain services, such as the maternity services, mental health services, and the relationship between maternal and child health (Smaje 1995). Findings from these studies were difficult to combine to gain a wider overview of ethnic health, since projects varied in their definitions of ethnicity and in the groups researched (or ignored), and offered little in the way of explanations for perceived variations.
Several larger-scale epidemiological surveys provide a broad overview of ethnic minority health. Marmot et al. (1984) reviewed the mortality of ethnic minorities compared with that of the white population in the Immigrant Mortality Study of the Office of Population Census and Study (OPCS), and Balarajan and Bulusu (1990) provided an analysis of mortality as part of the OPCS Mortality and Geography study among foreign-born residents of England and Wales, 1979 to 1983. Both studies were methodologically flawed (Smaje 1995) but have been seen as useful contributions to the dialogue about the health status of ethnic groups. A more complex picture of morbidity emerged from Nazroo’s (1997) national survey of the ethnic minorities (5196 respondents) with a comparison sample of the white population (2867 respondents). This study demonstrated considerable ill health among ethnic minority men and women. One-third reported ‘fair or poor’ health, a fifth reported a long-standing illness, and one-sixth reported that their performance of moderately exerting activities was curtailed by poor health. Overall, Pakistanis and Bangladeshis had the worst reported health, and these groups also reported the highest rates of heart disease. In this study health status was investigated both for differences and similarities between ethnic groups, with the author noting that the Chinese, Indian, and African-Asians experienced better health than other minority groups. Furthermore, Nazroo highlighted the important contribution of socio-economic status as a predictor of health. This and other work (Amin and Oppenheim 1992; Jones 1992) challenges the accepted link between ethnicity and poor health status by emphasizing the significance of the intervening variable of low socio-economic status. Many ethnic minorities live in considerable poverty for a number of reasons, including difficulty of gaining higher-grade employment, poor working conditions, and poorer-grade housing in both the private and local authority markets. However, some studies have failed to include any measures that might pick up deprivation status, instead accounting for poor health solely by the variable ‘ethnicity’.
Certain key messages emerge for any social or biomedical researcher. The complexities of research on ethnic minority groups in the United Kingdom are beginning to be understood and it now seems that much of the earlier work, particularly in the quantitative tradition, was simplistic and naive. More recent research and debate, often by sociologists, have suggested that there is a large research agenda still to be tackled. Douglas (1998), for example, has argued that interest in the United Kingdom has focused mainly on the largest of the groups, those of South Asian origin, often to the neglect of other minorities, for example, the Chinese and Vietnamese, African-Caribbean, and other smaller groups (including refugees). Research has also focused upon selected issues, for example, heart disease, ‘Asian rickets’, and the haemoglobinopathies, and ignored other conditions such as cancer.
It has also been argued, with powerful examples supporting the thesis, that the study of ethnic health is incomplete unless comparisons are drawn between ‘ethnic’ health status and that of the white population (Bhopal 1997; Nazroo 1997). Previously, many studies simply reported on the health of a specific ethnic group with no attempt to compare the findings with those for white groups whose health status may be similar, worse, or better. Furthermore, it is important to draw out similarities between the health of white and ethnic minority groups since these are as revealing as the differences. It has also been noted that when comparisons are made with the white population the assumption is always that the white population formed the ‘norm’ against which other ethnic groups should be compared. Moreover, the assumption of homogeneity within a single ethnic group has been shown to be misleading with differences associated with varying socio-economic circumstances, length of migration, and intergenerational changes, whilst the heterogeneity of the white population has also seldom been researched.
Finally, a research agenda for those adopting a qualitative approach to work in this field relates to questions concerning the quality of health care experienced by ethnic minority men and women, as well as how major common conditions are perceived, acted upon, and managed within ethnic minority families. More broadly, studies of social networks and of family life of ethnic minority groups would contribute considerably to the understanding of ethnic minority determinants of health.
Sociological work on health professionals dates back many years, when a rather academic debate existed about the structure and role of ‘professions’. Early analyses were shaped by functionalist theories, later becoming replaced by a number of ‘structuralist’ critiques. These were generally directly towards all professions in a discussion of ‘professionalism’, a device by which certain groups achieve control over occupational standing and expertise in society. More specifically, sociologists directed their attention to analysing the medical profession which held a high and, some would argue, overvalued position in society (Freidson 1970; Johnson 1972; Illich 1977). Central issues in this literature included an examination of how medicine maintained its status and autonomy in society and its dominance over other professions, such as nursing (Freidson 1970). As Annandale (1998) observed, Freidson’s writing ‘captured the collective imagination of a generation of sociologists’. However, not all agreed with Freidson’s analysis, and McKinlay and Stoeckle (1988) took the controversial view that today doctors sell their labour power on the market in the same way as any other worker. However, the general tenor of this debate was not to see members of the medical profession as members of the proletariat, but rather as a group that had until recently managed to maintain considerable power in society, manifest in their degree of self-governance. Studies published in the form of monographs described aspects of the self-monitoring process by the members of the medical profession, including Atkinson’s (1981) study of medical education and Bosk’s (1981) study of the management of medical errors.
The extent to which political developments have more recently curtailed medical power is an interesting question, and one that brings out the relevance of what may originally have been seen to be an intellectual exercise. Freidson (1994) himself acknowledges the challenges to the profession but feels that medicine is still attempting to police itself. In the United Kingdom this occurs through the General Medical Council, although recent ‘scandals’ relating to professional practice have resulted in questions about the effective fulfilment of this role, and in future there is likely to be considerably more public scrutiny of the medical profession.
Most importantly, in the United Kingdom over the last decades there has developed within the National Health Service (NHS) a new culture in which managerialism, rather than professionalism, is the dominant discourse. The rise of general management within the NHS and the construction of a series of institutions of clinical governance (for example, the National Institute for Clinical Excellence) have curtailed the autonomy of the medical profession. Social scientists continue to chart the changes over the past decade and have noted that the claim of the medical profession to rely upon ‘clinical judgement’ as the basis for self-governance is threatened by evidence-based medicine which has ‘achieved the status of official policy in the NHS’ (Harrison and Pollitt 1995).
Health professionals—a social action perspective
Continued, and some would argue continual, change within the health services in the United Kingdom means that traditional hierarchies are no longer fixed, and health professionals are at times asked to play dual roles, such as clinician and manager (Owens and Petch 1995). These circumstances have led a number of sociologists to investigate—using a social action perspective—the manner in which health professionals negotiate their work, create and maintain professional boundaries, and respond to change. This approach, whilst overtly often fairly atheoretical, ultimately derives from seminal work carried out in the 1960s and 1970s by researchers interested in understanding how professional groups interacted within the organization of the hospital. For example, Strauss et al. (1963) studied, among other things, the ‘negotiated order’ of hospital life, looking behind what was taken for granted about professional work and the division of labour, while Mechanic (1962) wrote perceptively about the informal power of lower participants of organizations. Research on the health service now takes a similar approach, examining not the formal roles of health professionals but rather looking at what lies ‘behind’ the new policies and changing roles of health service staff. Examples relate to a wide range of topics, including the study of medical professionals involved in management roles (Harrison and Pollitt 1995; Wilcocks 1999), and issues of interprofessional collaboration and conflict (Owens et al. 1995).
Work on this topic is important, since constant reorganization within the health service is accompanied by shifting roles and occupational boundaries, and a social action approach allows the research to ‘unpack’ the social context in some detail. For example, in a study using a mix of quantitative and qualitative methods, Fenwick et al. (1998) were able to uncover concerns held by the study general practitioners over new midwife-led schemes introduced into the community as a result of the policy document Changing Childbirth (Department of Health 1993). Many of the general practitioners questioned reported that they saw their pregnant patients fewer times during the pregnancy, and the majority were concerned that the midwives were not offering an informed choice to the women or were even dissuading them from visiting their general practitioner. Interviews with a sample of the practitioners also revealed the importance that they attached to their maternity work. The study suggests, as others have done (Cheyne et al. 1995), that changes in the organization of health services may often be accompanied by a redrawing of professional boundaries and tasks. The ensuing difficulties and conflicts may well inhibit the overall success of these high-level policy changes and deserve to be studied in considerable detail as a guide to successful policy changes in the future.
Professional views of official policy also provide another area where the social action approach has been applied. Challenging the view that the application of policies is a non-problematic exercise, the sociologist attempts to understand what these policies mean to professionals by asking such questions as, ‘What do they understand by the policy?’, ‘Do they agree with it?’, and ‘How may it affect work routines?’ In a study of maternity hospital policy, Beeken and Waterston (1992) investigated the attitudes of hospital and community midwives and health visitors towards breast feeding in a Newcastle hospital by sending the staff a postal questionnaire (92 per cent responded). Although the majority of staff reported that hospital policies were intended to facilitate breast feeding (for example through rooming-in) the researchers found that some nurses reported opinions about breast feeding which were at variance with hospital policy. For example, 47 per cent disagreed or strongly disagreed with the statement that ‘Milk company advertising should be banned in the antenatal clinic’ and over a quarter disagreed with the statement that ‘Healthy babies are breast-fed babies’.
Other examples that strikingly demonstrate the gulf between official policy and the views of the rank and file workers include Williams and Calnan’s (1994) study of general practitioners’ views about health promotion surrounding coronary heart disease. In face-to-face interviews the general practitioners reported considerable ambivalence towards the efficacy of behavioural choice, noting also time constraints in carrying out health promotion activities and the difficulties of identifying individuals at risk. They also reported that whilst they recognized that prevention was important they found health promotion ‘tedious, dull, and boring’, and had little time or interest to attend to this aspect of their work.
Relatively little research has examined the agency of professionals and the ways in which they may attempt to maintain control within their working situation, and this is an area that would benefit from greater study. More generally, considerable attention has been paid to the consumer as a recipient of health care, as a respondent in the research process, and as a patient whose views are increasingly worthy of attention, but the voice of health professionals is less heard. Health professionals are key players in health service research (for example, in randomized controlled trials, health promotion campaigns); they are routinely expected to help recruit patients, implement interventions, and facilitate and interpret such research. However, ways in which the research may have an impact on their workload or the intervention may affect their working practices is little explored, and very little research has examined the complexities of the work situation of professionals and the uncertainties associated with their redefined roles.
Medical sociology has traditionally worked closely with public health and clinical specialists, and this is likely to continue, with such issues as the social causes and distribution of disease, the organization and provision of medical care, and the development and implementation of preventive programmes, all having important social aspects and dimensions. The contributions of sociology to these areas arises both from its techniques of enquiry (particularly the application of qualitative methods), and its body of knowledge concerning the nature and influences of social factors and processes which operate both at a micro-level and in terms of broader social structures. However, it is increasingly difficult to determine precise disciplinary boundaries, as sociology, psychology, and public health medicine each draw on concepts and explanations developed in the other disciplines in addressing public health issues, and over time these become part of the accepted wisdom in the field. For example, the sociological distinction between disease and patients’ experience of illness is now commonplace and subjective health status measures are employed routinely in assessing the outcome of medical care. Similarly, effective health education is also accepted to require not merely the provision of health advice but also to address people’s personal beliefs and take account of the situation and context of their lives. More generally, whereas qualitative methods were traditionally largely confined to sociologists and anthropologists, they are increasingly employed by public health doctors and other medical specialists to investigate the perceptions and experiences of patients and health professionals.
Areas of specialization are influenced by prevailing concerns in the health field and priorities for research funding, and therefore vary over time and between countries. However, current trends suggest that in the United Kingdom issues of social inequalities in health, in terms not only of socio-economic position but also the effects of ethnicity and gender, will form important foci of research and collaboration between sociologists and public health doctors, as ill issues of community priorities and rationing, and assessments of the quality and acceptability of health care. In addition, assessment of the quality of qualitative research and requirements for synthesizing qualitative studies and producing systematic reviews are likely to receive increased attention.
Medical sociology has traditionally had an applied focus (especially in the United Kingdom) and this is likely to continue. However, increasingly this subfield is also characterized by greater emphasis on theoretical concerns and developments. Indeed, Turner (1992) suggests that the sociology of health and illness has the potential to ‘become the leading edge of contemporary sociological theory’. This is because the body, which is central to health and illness, is forming an important dimension of sociological debates and raises in acute form issues of the relationship between body and mind, culture and nature, and self and society. Such developments in the theoretical basis of the discipline will provide new insights and concepts that may later be applied to questions posed by clinicians, public health specialists, and policy makers. Thus future work in sociology in medicine is likely to be strengthened by developments both in theory and method.
Throughout its history of over a century the discipline of psychology has included health issues within its ambit of research and practice. Not surprisingly, most attention has been paid to mental health. The present chapter, however, focuses on psychology’s contributions to the realm of physical health, illness, and disease. Prior to the 1960s, the most notable contribution of psychologists to public health investigations was the development of psychosomatic theories of disease. In the 1960s psychological variables and measures began to appear in large-scale epidemiological studies. For example, at this time the prospective Framingham Heart Study included assessment of type A behaviour (excessive aggression, ambition, competitiveness, and time urgency), and subsequently showed that the attribute significantly increased the risk of coronary heart disease in men and women (Haynes et al. 1980). This type of research became known as behavioural or psychosocial epidemiology, explicitly recognizing psychological contributions to the broad epidemiological picture and to public health issues (Kaplan et al. 1990).
The following decade saw the emergence of the fields known as behavioural medicine, behavioural health, and health psychology (Matarazzo 1980). Definitions of the first two vary, but broadly refer respectively to the integrated contributions of the behavioural and biomedical sciences to the diagnosis, treatment, and rehabilitation of the sick, and to the maintenance of health and prevention of sickness in the healthy. Within this multidisciplinary framework, health psychology has been defined conventionally as the
educational, scientific and professional contributions of the discipline of psychology to the promotion and maintenance of health, the prevention and treatment of illness, the identification of etiologic and diagnostic correlates of health, illness, and related dysfunction, and the improvement of the health care system and health policy formation. (Matarazzo 1980)
In the last 20 years behavioural medicine, behavioural health, and especially health psychology have developed at an extraordinary pace. Professional societies, research institutes, specialist journals, textbooks, training courses, conferences, and positions in universities and health agencies have proliferated in many countries (Jansen and Weinman 1991). In Europe alone the development of health psychology has recently prompted a Health Psychology 2000 report by a special Task Force of the European Federation of Professional Psychologists’ Associations (Marks et al. 1998). Not surprisingly health psychologists have so far focused their attentions mainly on the health problems of developed countries, conducting extensive research on chronic diseases such as coronary heart disease and cancer, and on risk factors such as a high-fat diet and a sedentary lifestyle. However, some interest is now emerging among health psychologists into the health issues faced by developing countries such as malnutrition and family planning (Aboud 1998).
The extreme diversity and proliferation of psychologists’ contributions to investigations of health in recent times means that this chapter can only provide a broad outline of their methodological approaches, and a few selected examples of substantive findings. The next section outlines the characteristic ways in which psychologists formulate theory about health phenomena. Following this, some of their distinctive approaches to research design, measurement, and statistical analysis are discussed.
Authors of texts in health psychology and behavioural medicine conventionally characterize their fields as theoretically located in the ‘biopsychosocial model’ (Schneiderman and Tapp 1985; Taylor 1995). The model, which is attributed to Engel (1977), envisages health and illness as the product of the complex interactions of biological, psychological, and social processes. It is usually contrasted with the biomedical model, which is seen as reductionistic and focused on illness and disease rather than health. The term ‘biopsychosocial’ is regarded not just as a shorthand way of signalling the need to take account of three different types of variables, but more importantly as a reminder that such variables are to be seen as comprising systems. From this perspective, explanations and interventions need to be conceptualized at the system level of the individual and of the social systems in which he or she is embedded rather than at the level of the variable. For example, a biopsychosocial account of healthy dietary practices cannot be reduced to nutritional components, but must incorporate relevant cognitive, emotional and behavioural attributes of the individual, and the social and cultural forces that shape them. This would be a commonplace observation to a medical anthropologist or sociologist, but is a much-needed reminder to psychologists who, by virtue of their disciplinary focus, tend towards individualistic if not reductionistic explanations.
The biopsychosocial model has been promoted by health psychology and behavioural medicine, with until recently surprisingly little critical reflection. McLaren (1998) has suggested that the biopsychosocial model is not a model at all, and provides no real theoretical force or constraint. Even those who believe that the biopsychosocial model has something to offer as a general guide question whether psychologists are actually heeding its messages. Thus concerns have been raised that in practice psychologists in the health field continue to study variables rather than systems, decontextualizing and fragmenting individuals, and then trying to reassemble them with statistical analysis (Spicer and Chamberlain 1996; Ogden 1997; Spicer 1997). A more radical response has been to reject the biopsychosocial model, on the grounds that any natural systems approach based on quantitative analysis of variables cannot capture the meanings and contexts of health-related experience. Instead, qualitative paradigms and methods are needed to interpret the discourses and narratives of people’s health worlds (Murray and Chamberlain 1998, 1999).
Perhaps not surprisingly in a new and rapidly growing area, the theoretical foundations of health psychology and behavioural medicine have yet to settle. The biopsychosocial model has at least served a useful rhetorical function in helping psychologists establish an identity in the health arena. But the serious questions as to its theoretical and methodological implications, and its ideological relationship with the biomedical model, are only beginning to be addressed. Given the extreme diversity of the issues embraced by health psychology and behavioural medicine, it seems unlikely that a single foundation will be found. More likely a variety of theories and methods will be adopted depending on the question at hand, though the vexed question of amalgamating findings from disparate approaches remains (Buchanan 1992).
Despite creaking foundations, there is no shortage of theories in health psychology which address public health concerns. Two examples of influential theories will give the flavour of psychological theorizing in this area: the health belief model and the transactional theory of stress and coping. Health behaviours are seen as any actions taken to prevent or detect disease or to improve well being (Conner and Norman 1996). As such, they encompass such activities as dietary control, physical exercise, visiting health professionals, and adhering to prescribed medications. The predominant strategy for attempting to uncover the determinants of health behaviours in psychology has been to devise so-called social cognition models. These theoretical models assume that the most important determinants of social behaviours such as those concerning health are to be found in people’s cognitions: their rational cost–benefit analyses of whether or not to engage in the behaviour. One of the earliest and most influential social cognitive models was the health belief model (Becker 1974; Sheeran and Abraham 1996). This envisages the decision to engage in smoking, for example, as a function of the individual’s perceived susceptibility to and the severity of the health consequences of smoking, as well as the perceived benefits and barriers relating to the behaviour. The model also includes general motivations concerning health, triggers to action, and background demographic, personality, and social factors, but the cognitive variables are seen as the key determinants. As Conner and Norman (1998) note in a helpful review of the health belief model and other social cognitive models, the model has been tested on a diverse range of health behaviours including smoking, alcohol use, dietary practices, exercise, health screening activities, contraceptive use, compliance with medical regimens, and visits to health professionals. Meta-analyses of these studies suggest that many health behaviours can be consistently predicted by each of the health belief model cognitions, most successfully with perceived barriers, followed by susceptibility, benefits, and severity. No single component of the model, however, predicts more than 4 per cent of variance in health behaviour so, even when considered overall, the health belief model provides a limited capacity for predicting health behaviours.
The partial success of the health belief model is not surprising given its limitations. The broad limitations, discussed by Conner and Norman (1998), are worth noting here because they are often voiced as criticisms of theories in health psychology generally. A first limitation is the small set of constructs included in the model. Conner and Norman (1998) highlight in particular the absence of personal control and perceived social pressure as important determinants of health behaviours. More generally, the concern is often expressed that psychological theories of health are too circumscribed. To be comprehensible and testable, theories have to be selective, but not to the extent of omitting important constructs. A second limitation is the lack of a theoretical account that explains how the cognitions are connected with each other and with behaviour. Without such an explanation, the model is little more than a list of constructs, which provides a weak basis for interventions. The third limitation is the variety of ways in which investigators have measured the health belief model cognitions—a latitude which can be partially explained by the lack of theory clearly specifying the constructs and their relationships. This loose linkage between constructs and measures is a concern which frequently surfaces in health psychology. Finally, the health belief model and many other theories in health psychology have been criticized for their failure to capture process. Giving up smoking, for example, is clearly a multistage process involving reflection, planning, execution, and maintenance. Cognitions are undoubtedly relevant at each stage and across each transition, but determining which and when particular cognitions are important requires a theory which goes beyond the static simplicity of the health belief model. Some other theories of health behaviour that avoid some of these limitations are discussed in the applications section below.
A very different approach to theorizing in health psychology can be found in Lazarus’s theory of stress and coping (Lazarus and Folkman 1984), though it did not originate in this field. The concept of stress has loomed large in the health arena for many years, and Lazarus’s psychological theory has provided one of the most sophisticated and influential explanatory accounts. The theory is transactional in the sense that it attempts to characterize the continual transactions between person and environment, focusing on those defined as stressful. Interpretations of events and situations are given a central role in the model via the concept of appraisal. Primary appraisal refers to the person’s evaluation of whether they are confronting a demand on their resources and, if so, whether the demand constitutes a threat of harm, loss, or challenge. Secondary appraisal refers to a person’s evaluation of their ability to cope with the demand using personal or environmental resources. Coping strategies are the actual steps taken by the individual to deal with the demand and, in Lazarus’s scheme, are categorized into problem-focused and emotion-focused strategies. The former address the problem itself, while the latter deal with the emotions generated by the problem. These categories are further subdivided into eight specific strategies or ‘ways of coping’ such as confrontive coping, distancing, self-control, and seeking social support. Stressful experiences are those where demands on resources are not met, either in prospect or following ineffective attempts at coping.
The complexity and sophistication of Lazarus’s theory emerge in two ways. Firstly, the demand–primary appraisal–secondary appraisal–coping–outcome dynamic is not a simple linear sequence. These processes interweave and involve transformations such as reappraisals and anticipatory coping manoeuvres. Secondly, the ways in which these dynamics unfold depend on a range of idiosyncratic factors such as personality attributes (including available and preferred ways of coping), beliefs, and commitments concerning the situation, personal skills and capabilities, and the availability of environmental resources and constraints. In more recent discussions, Lazarus (1991) has elaborated the theory still further, most notably highlighting the key role of emotions. Such a brief description does scant justice to this theory, but at least its complexity should be apparent.
Lazarus’s theory has generated enormous debate and has informed extensive research in the health field (Bartlett 1998). However, it is striking that virtually none of this research constitutes a test of the theory as such. Instead, investigators continue to choose particular variables (appraisals of life events, particular coping strategies such as denial, stress-prone personality traits such as type A behaviour, or social support) and attempt to link them individually to health and disease states. In so doing, they lose the transactional force of the theory as well as many of its other strengths.
Lazarus (1990) has called for more longitudinal research to expose the links between stress, coping, and health. But, as Bartlett (1998) argues, the problems of testing the theory run much deeper. The psychological methods needed to capture the dynamic complexities of this or any similar theory have yet to be developed. Ironically the research designs and particular methods used by most health psychologists are best suited to testing the type of theories of which they are most critical—piecemeal static models of limited scope. From one perspective this is deeply problematic, but the tension between theory and method can also be seen as a force which will drive the development of both.
Most of the methodological strategies adopted by psychologists working in public health research are not distinctive to their discipline. The use of cross-sectional and prospective surveys, programme evaluations, randomized control trials, probability sampling strategies, interviews and questionnaires, and statistical analysis is common to a variety of researchers in public health. In fact, this sharing of methods across disciplines is increasing as public health researchers attempt to incorporate elusive psychological and social phenomena such as stress, lifestyle, and quality of life (Daly et al. 1997). Within health psychology, the growing interest in qualitative methods noted above is producing a convergence of approaches with those developed by sociologists and anthropologists (Murray and Chamberlain 1999).
This extensive sharing of methods makes it unnecessary and inappropriate to repeat discussions contained in this and other chapters. Instead, this section highlights and illustrates some methodological strategies that are particularly characteristic of psychologists’ investigations of health. These encompass using laboratory experiments to test causal hypotheses about health-related behaviour, enhancing the quality of psychological measurement, and using multivariate statistical techniques, especially those designed to test complex causal models of health determinants and consequences.
Since its inception in the late nineteenth century, psychology has made extensive use of laboratory experiments to evaluate causal hypotheses about behaviour and mental life. The logic of experimentation is the same as that for randomized controlled trials, but the laboratory setting permits greater control of extraneous variables. In essence, an independent variable is manipulated under controlled conditions to test whether a dependent variable varies in a way consistent with a causal hypothesis. Control strategies are primarily aimed at preventing the operation of confounding variables. These mask the true relationship between the independent and dependent variables by virtue of their relationships with both. Confounds may be controlled by holding them constant or by the use of some balancing strategy such as matching or random allocation of participants to experimental conditions. Such control strategies help to ensure that the effects of any manipulations of the independent variable are not confounded with variations in experimental procedure, or with differences in individual characteristics such as age, personality, and experience. Experimental designs can become very complex according to the number of independent and dependent variables, and the ways in which participants are exposed to experimental conditions (Keppel 1991; Kirk 1995). However, the fundamental logic of manipulation and control remains throughout.
An example from the cardiovascular disease literature shows these ideas in action. The reactivity hypothesis suggests that risk of coronary heart disease and hypertension may be increased in people who consistently over-react physiologically to psychological stress (Turner 1994). Many experimental studies have tried to elucidate the nature of the stressors that might elicit reactivity and the characteristics of reactors. In one such study (Brown and Smith 1992) husband and wife couples were randomly allocated to one of two laboratory tasks: a general discussion or a competitive argument in which the ‘winner’ received the chance of a monetary reward. During these activities blood pressure, hostile behaviours, and anger experiences were recorded in a standardized form. Results showed that during the argument, but not the discussion, husbands exhibited notable increases in blood pressure and anger, whereas wives did not. It was concluded that striving for interpersonal control may induce blood-pressure reactivity, but that this process is moderated by spouse status. The authors then discussed the implications of the results for psychological interventions to reduce reactivity and disease risk.
Since the laboratory experiment is generally seen as a powerful way to demonstrate causal processes, it is tempting to accept results and interpretations uncritically, especially when the design and analyses are complex. However, the application of the strategies of manipulation and control to human experience raises many questions and difficulties, both technical and ethical. There is a considerable literature on the subtle ways in which experimenters and participants can unwittingly introduce biases that can confound results (Rosenthal 1994). Even when these are controlled, the question remains as to whether all participants construe their experiences in the same way and in the way intended by the experimenter (Harré 1993). Even more fundamental is the ‘identity problem': the question of whether complex psychological processes such as interpersonal control can be induced or simulated in a laboratory in ways which retain their meanings in everyday life (Greenwood 1989). Without some reassurance as to the ecological validity of psychological laboratory results, implications for the understanding and enhancement of health cannot be legitimately drawn. Despite the many challenges, psychological laboratory investigations have an important role to play in elucidating psychological aspects of health, especially when the results are incorporated with those from other research strategies. In the case of the reactivity hypothesis, for example, the importance of psychological control is becoming apparent from a range of epidemiological, ambulatory, and laboratory studies of blood-pressure reactivity (Steptoe 1998).
Enhancing the quality of psychological measurement
Psychologists make use of a wide range of measurement instruments to investigate public health concerns. Johnston and Johnston (1998) have recently provided a comprehensive overview of these measures, and other more detailed guides are also available (McDowell and Newell 1987; Bowling 1991, 1995; Cohen et al. 1995; Johnston et al. 1995). As Johnston and Johnston (1998) note, measurements are taken in a variety of modes: self-reports in interviews and questionnaires, observations of behaviour, physiological assessment, and health-care records. The attributes measured are extremely diverse and include appraisals of stressors, health, and illness, health-related behaviours such as exercise and adherence to medication, emotional precursors and consequences of illness, personality factors, coping strategies, communication processes between patients and health-care providers, and social resources.
Whatever the nature of the measuring instruments, their quality is assessed using well-established principles of psychometric theory (Nunnally and Bernstein 1994; Anastasi and Urbina 1997). These distinguish between various types of reliability (consistency) and validity (accuracy) and provide a statistical basis for assessing the quality of life measures. For example, the WHO Quality of Life Assessment Instrument-100 measures 24 facets of quality of life using four questionnaire items for each. A recent report on the reliability and validity of the WHO Quality of Life-100 (WHO Quality of Life Group 1998) provides correlational data which demonstrate the internal consistency or homogeneity of the 24 subscales. It also addresses the factorial validity of the 24 facet scores, showing that they cluster into four domains rather than the six which had been suggested initially. Discriminant validity of the items is demonstrated by their ability to distinguish between healthy and unhealthy respondents, and construct validity of the facets and domains is shown by the pattern of correlations between these scores and those on an overall quality of life subscale.
Distinguishing and assessing multiple types of reliability and validity can provide a comprehensive evaluation of measurement quality, but such information can be difficult to integrate for a particular measure. The multitrait–multimethod strategy (Campbell and Fiske 1959) has helped investigators to combine data on some types of validity. The problem of integrating reliability data, which may reflect measurement error due to different times of administration, different observers, or heterogeneous items, has also been addressed by Cronbach’s generalizability theory (Cronbach et al. 1972; Shavelson et al. 1989). This allows the investigator to quantify different sources of error within one analytic framework, and then to estimate how the quality of the measuring instrument would be affected by manipulation of particular sources. Streiner and Norman (1995) provide a helpful and detailed introduction to this approach in a health context using the Objective Structured Clinical Examination as an example. Despite being available for some years in the psychological literature, neither the multitrait–multimethod technique nor generalizability theory have made much impact in public health research. They are demanding in terms of both study design and analysis. However, the gains they offer in developing and assessing effective measuring instruments are more than sufficient to warrant their wider use.
Enhancing the quality of psychological measurement can also be achieved by elucidating the psychological and social processes that occur during the measurement process. How well does the structure of a scale requiring a respondent to rate their intention to exercise mesh with their cognitive schemas? What are the best ways to recover memories of past events such as seeking medical care? Does the use of a standardized interview on health problems disrupt normal social processes and undermine the participant’s ability to provide the interviewer with accurate information? These and similar questions have received increasing attention from psychologists and other social scientists in recent years (Hipler et al. 1987; Tanur 1991; Abraham and Hampson 1996).
Ostrom (1987), for example, raises the worrying concern that continuous rating scales do not map onto respondents’ cognitive representations since, according to cognitive theory, the latter have an all-or-none categorical form. Accordingly, respondents are obliged to decompose the scale into categories which may be highly idiosyncratic. Thus, respondents’ apparently simple and comparable ratings may hide individual cognitive processes that make the aggregation of ratings highly questionable. Turning to memory, Loftus et al. (1991) report a series of studies on patients’ recall of visits to health professionals where they found that respondents systematically underestimated their total number of visits, but overestimated the frequency with which they had undergone a specific procedure. They use these findings to reflect on the relative merits of asking questions with different time frames, and of guiding respondents’ recall backwards or forwards along a time line. As a final example, Suchman and Jordan (1991) provide an incisive exploration of the tensions between the interview seen as a standardized instrument of assessment and as a conversation, partly using data from the National Health Interview Survey in the United States. They explore the conventional proposition that the validity of interview data requires shared understandings among the researcher, the interviewer, and the interviewee. However, they argue that these shared meanings depend on the availability of conversational resources that ‘routinely mediate uncertainties of relevance and interpretation’ (Suchman and Jordan 1991). These resources are inevitably suppressed to some degree, they argue, by the scripted nature of the interview, and accordingly validity is threatened. Researchers are clearly aware of this problem in general, and make use of devices such as clarifying definitions and probes to ensure that the meanings of questions and responses are standardized. However, the concern remains that this approach is guided by an overly simple view of the social dynamics of interviews, especially their capacity for suppressing responses.
It should be stressed that cognitive and social research on the processes underlying psychological measurement in the health context has not progressed very far. Although there is no shortage of relevant theory and findings within cognitive and social psychology, research which explicitly brings them to bear on measurement issues is limited and has so far raised more questions than answers or clear recommendations. It seems reasonable to suggest that this type of research is a potentially important contribution that psychologists can make to the enhancement of measurement and thus to the overall quality of research in public health.
Multivariate statistical techniques
Multivariate statistical techniques play a central role in psychological research in all fields including health. A variety of excellent texts are available (Grimm and Yarnold 1995; Tabachnick and Fidell 1996), and psychologists contribute to the development of the techniques in journals such as Psychological Methods and the Journal of Multivariate Analysis. Techniques such as multiple and logistic regression, discriminant analysis, and log-linear analysis enable the estimation and testing of hypothesized effects while statistically holding constant any measured confounding variables. The particular technique selected depends partly on the assumptions that the analyst is prepared to make about their data, especially the level of measurement which has been achieved. Although statistical control of confounds is less effective than experimental control, it provides a powerful analytic tool when experimentation is impossible or inappropriate: a common situation in psychological analyses of health.
A notable development in this area is the increasing application of structural equation modelling using programs such as LISREL, AMOS, and EQS (Hayduk 1996; Tabachnick and Fidell 1996). Structural equation models are an extension of path analysis (Klem 1995), which is in turn an elaboration of regression analysis. They enable the analyst to estimate and test complex networks of relationships including causal chains, reciprocal relationships, and feedback loops. A particular model can be assessed using one powerful and efficient analysis, as opposed to the piecemeal decomposition of relationships required by earlier techniques. Furthermore, structural equation models allow the analyst to incorporate multiple measures (indicators) of any construct (latent variable), and to build estimates of measurement error into the analysis. Thus a structural equation model analysis is able to test simultaneously a causal model and a measurement model using extensions of regression and factor analysis. Again, this contrasts with traditional approaches which treat statistical evaluation of measures and of causal relationships as two separate steps.
Structural equation models are very useful for the evaluation of competing models since they provide helpful indices of fit between the theoretical model and the data. For example, Hall et al. (1998) used LISREL to compare two models that sought to explain why sicker patients report less satisfaction with their medical care. The simpler model envisaged the dissatisfaction to be directly attributable to patients’ generalized negativity, while the more complex model hypothesized that the dissatisfaction effect was mediated by negative interactions with doctors. Health status and doctor–patient communication were assessed with multiple measures and satisfaction with a single measure in two samples. The analyses showed that generally the simple model provided a better fit to the data, though there was some slight evidence of mediation in one of the samples. This example illustrates the simultaneous analysis of a causal and measurement model, but structural equation models can also be used to analyse one or the other in isolation. The structural equation model analysis of a measurement model is known as confirmatory factor analysis. For example, in their evaluation of the WHO Quality of Life-100 described above, the WHO Quality of Life Group (1998) conducted a confirmatory factor analysis using the EQS program to test whether the facets clustered into one, four, or six domains. As noted above, the four-domain model provided the best fit to the data according to this structural equation model analysis.
The analysis of structural equation models can be daunting to the novice, though the availability of good introductory texts and the increasing friendliness of computer programs is helping to reduce this problem. Regardless of this, the technique will inevitably become widely used in public health research since multiple measures are often needed to capture the complexity of health-related phenomena, and theoretical accounts of their relationships are becoming increasingly complex. Complexity in the theoretical and measurement domains will have to be matched by a corresponding sophistication of statistical analysis.
Areas of psychological investigation
In the last 20 to 30 years, psychologists have applied their theories and methods to a wide range of problems in the public health arena. The contents of any standard text in health psychology (for example DiMatteo 1991; Taylor 1995) reveal investigations of the following:
why people do or do not engage in healthy behaviours such as exercise and safe sex
the psychological precursors and consequences of many diseases such as cardiovascular diseases, cancer, and diabetes
patients’ interpretation of physical symptoms
patients’ engagement with health professionals
how patients cope with pain and chronic conditions
the psychological problems faced by health professionals
the psychological aspects of terminal illness and bereavement.
In turn these investigations increasingly inform primary, secondary, and tertiary interventions aimed at improving public health. This section provides discussion of just two examples of psychological investigations: psychological precursors of coronary heart disease and psychological determinants of condom usage. These illustrate psychological approaches to predicting disease and promoting health, respectively. They also exemplify different psychological foci: personality and the social environment in the former and cognitive processes in the latter. Conversely, the two examples have in common an extensive research base so that it is possible to discuss patterns of findings rather than isolated studies.
Psychological risk factors for coronary heart disease
The aetiology of coronary heart disease has been a major focus of attention for psychologists because of the disease’s major impact on morbidity and mortality in developed countries, and the clear implication of antecedent behavioural factors. A comprehensive review of psychological risk factors for coronary heart disease was undertaken in the mid-1990s as part of a larger review of behavioural research in cardiovascular, lung, and blood health and disease by the National Heart, Lung, and Blood Institute Task Force (NHLBI Task Force 1998). The report distinguishes between lifestyle factors, such as smoking, exercise, and diet; individual characteristics such as personality and physiological reactivity to stress; and psychosocial variables such as socio-economic status, social support, and occupational stress. Since the role of lifestyle factors in cardiovascular diseases is discussed elsewhere (Chapter 9.1), this section illustrates psychological research with examples from the latter two categories.
Research on individual psychological characteristics and coronary heart disease has been dominated by the type A behaviour pattern: a profile of excessive competitiveness, sense of urgency, ambition, and hostility. There have been many accounts of the history of the type A behaviour pattern and its association with coronary heart disease (Strube 1991) which tend to highlight similar themes. The first is the complexity of the pattern originally described by the cardiologists Freidman and Rosenman in the 1960s (Freidman and Rosenman 1974), and the rapidity with which this was simplified by other investigators. Whereas the type A behaviour pattern was conceived as a complex pattern of person–environment interaction, it quickly became treated as a personality trait. This development was not unrelated to the proliferation of self-report measures of the pattern which were unable to evaluate the key behavioural components (for example, vocal mannerisms, posture, and conversational style) which Friedman and Rosenman assessed with their structured interview. As a result, analyses of the links between the type A behaviour pattern and coronary heart disease have been beset with problems of shifting definitions and of measures which are not highly correlated with each other.
A second theme is the changing pattern of results from epidemiological studies. Two major prospective studies in the 1970s—the Western Collaborative Study and the Framingham Study—reported that the type A behaviour pattern doubled the risk of coronary heart disease independently of established risk factors. However, major studies in the following decade, notably the Multiple Risk Factor Intervention Trial, failed to find any association between the type A behaviour pattern and coronary heart disease using various different measures of the pattern. Despite these mixed findings, meta-analyses have demonstrated a statistically reliable association (Booth-Kewley and Friedman 1987; Matthews 1988). Explanations for the variability in results include the use of different measures and study designs, the health and socio-economic status of the study populations, and the increasing prevalence of the pattern. It is also possible that the effect of the type A behaviour pattern may depend on the specific coronary heart disease endpoint. Spicer et al. (1996) and others have suggested, for example, that the type A behaviour pattern may increase the risk of non-fatal coronary events, but decrease the risk of fatal events.
Another possible explanation has led to the main theme in current research on coronary-prone behaviour. This is the proposal that only some components of the type A behaviour pattern confer coronary risk, and that the most potent of these is hostility. Research in this area has broadened beyond the type A behaviour pattern emphasis on hostile behaviour to include affective aspects such as anger, and cognitive aspects such as cynicism and mistrust of others. An impressive array of evidence now shows a consistent link between hostility, broadly defined, and coronary heart disease (Siegman and Smith 1994; Miller et al. 1996). This epidemiological linkage is given further credibility by evidence showing that hostility is reliably associated with variables which might mediate coronary risk such as smoking, lipid levels, poor social support, and physiological hyper-reactivity (Siegman and Smith 1994). Hostility levels might also provide a partial explanation of the well-established inverse association between socio-economic status and risk of coronary heart disease.
Although research on type A behaviour pattern/hostility has dominated the study of coronary-prone behaviour, other psychological characteristics have also received attention. Booth-Kewley and Friedman’s (1987) meta-analysis highlighted a strong association between depression and coronary risk. The National Heart, Lung, and Blood Institute Task Force report (NHLBI Task Force 1998) also gives prominence to depression as a coronary risk factor, noting further that it increases the risk of future coronary heart disease events in myocardial infarction patients, regardless of their severity status. Of related interest is a body of literature on ‘vital exhaustion’, which the report describes as a ‘mental state characterized by unusual fatigue, a feeling of being dejected or defeated, and increased irritability’. This state is interesting, not only because of its demonstrated association with coronary risk, but also because of its resonances with the findings on depression and anger. Theoretical attempts to link type A behaviour, anger, and depression have been made (Price 1983), but the theory as a whole has received little empirical attention.
Psychological risk factors for coronary heart disease have also been sought in the environment. Two noteworthy examples of this type of research are investigations of job strain and poor social support. Job strain is generated by occupations that are highly demanding but offer little latitude for decision-making and control in general. Schnall et al. (1994) have reviewed studies on the association between job strain and coronary heart disease, and found supporting evidence. That said, the association is not always present, most notably in a more recent prospective study of patients undergoing diagnostic testing for coronary heart disease where job strain was not related to coronary heart disease prevalence or to subsequent morbidity or mortality (Hlatky et al. 1995).
An extensive literature has developed on the health effects of social isolation and poor social support (House et al. 1988). Isolation is evaluated in terms of the degree of social contact and the individual’s experience via partners, friends, relatives, and organizations. In contrast, support refers to individuals’ perceptions of what social resources they can call on when necessary. Many different sorts of support have been distinguished, such as material, emotional, and informational support, and a variety of support measures have been developed. The review by House et al. (1988) provides compelling evidence of a reliable association between poor social resources and health, especially between social isolation and all-cause mortality. More recently Shumaker and Czajkowski (1994) have brought together the evidence relating to poor social resources and cardiovascular disease and again documented consistent associations. The effects of social resources on health may be most obviously explained in terms of social influences on health-related behaviours such as diet, health checks, and adherence to medications. In addition, there is growing evidence to suggest that social resources may also benefit physiological functions such as blood pressure (Uchino et al. 1996).
The foregoing is no more than a limited sample of the enormous literature on psychological risk factors for coronary heart disease. Only a few major epidemiological topics have been highlighted, and hardly any mention has been made of the numerous experimental and animal studies aimed at elucidating risk mediation, or of the intervention literature. Nonetheless, even the limited research reviewed constitutes a substantial contribution. After their extremely comprehensive review, the National Heart, Lung, and Blood Institute Task Force prefaced its 1998 report with the conclusions:
Behavioral research has contributed significantly to the understanding of disease risk and the progression and clinical manifestations of disease. Effective modification of individuals’ behavior and/or psychosocial environments can potentially reduce disease risk, ameliorate the burdens of illness, and promote recovery and rehabilitation. (NHLBI Task Force 1998)
Psychological determinants of condom usage
An earlier section introduced the theme of health behaviours and the use of social cognitive models to uncover their psychological determinants. This section provides a brief account of two models that have superseded the health belief model discussed above, an account which draws in particular on the reviews of Conner and Norman (1996, 1998). More specifically, it summarizes some results from research that have applied these models to the problem of condom usage. This particular health behaviour is important in the context of both unwanted pregnancy and sexually transmitted diseases such as HIV. As Conner and Norman (1998) note, condom usage is still worryingly low despite extensive public health messages. For example, they cite figures from the General Household Surveys in the United Kingdom indicating that usage among 16- to 24-year-olds increased from 6 per cent in 1983 to only 12 per cent in 1991.
One of the most influential social cognitive models is the theory of planned behaviour (Ajzen 1991), which has been applied to a wide range of behaviours inside and outside the health field. According to the theory, a behaviour such as using a condom is most directly determined by how strongly the individual intends to carry out the behaviour, and by the degree to which they perceive the act as within their control. Intentions themselves are influenced not only by perceived control, but also by attitudes and so-called ‘subjective norms’. In this context, attitudes refer to what the person believes about the outcomes of using a condom in conjunction with the value they attach to those outcomes. Subjective norms refer to the person’s beliefs about whether significant others would prefer him or her to engage in the behaviour, and the strength of the person’s motive to comply with this preference. Other more indirect behavioural determinants are also proposed, but the key factors are attitudes, subjective norms, and perceived control, which mainly influence behaviour via intentions. According to this model, people’s decision to use a condom is thus strongly determined by their motives, beliefs, and evaluations with respect to their capacity to act, the outcomes of their action, and the views of salient others.
Another social cognitive model of interest is derived from protection motivation theory (Rogers 1983). This envisages the decision to use a condom as immediately determined by the person’s motivation to protect his or her health, which is influenced by appraisals of both the threats and rewards associated with not engaging in the behaviour, and the issues surrounding the actual performance of the behaviour. As Conner and Norman (1998) note, these appraisals are conceptually similar to Lazarus’s notions of primary and secondary appraisal that were discussed in the earlier section on theory. Of particular interest in the latter set of appraisals is the notion of efficacy. This encompasses response efficacy, the sense that a behaviour will be effective in protecting health, and self-efficacy, the belief that the person is capable of carrying out the behaviour. People’s beliefs about efficacy seem to play a critical role in determining their health behaviours. Conner and Norman (1998) show how the construct plays a key role in a variety of social cognitive models including protection motivation theory, and review evidence from correlational and intervention studies which support its importance.
Turning to the evidence on relationships between social cognitions and condom usage, Sheeran et al. (1999) have recently published a meta-analysis of 121 studies that have used social cognitive models, notably those summarized above, or at least elements of them. These studies are of heterosexuals from a variety of countries, just over a third of whom are university students. The analysis provides average correlations (weighted by sample size) between social cognitive model elements and condom usage. Interestingly, there appeared to be few differences between the pattern of correlations found in cross-sectional studies and in prospective designs. Since the latter studies, although only 15 per cent of the total, provide the most rigorous test of social cognitive model hypotheses, the following figures are taken from this subset. As the theory of planned behaviour suggests, the most powerful correlate is the intention to use a condom as indexed by an average correlation of 0.46. This is particularly striking when compared with the correlation of 0.39 for previous use. It is a truism that past behaviour tends to be a strong predictor of its recurrence, but intention appears to be a stronger determinant. The next-strongest correlations were for attitudes towards condoms and so-called descriptive norms (beliefs about the views and behaviours of others regarding condom use): r = 0.33 in both cases. Self-efficacy is then correlated 0.24 with usage, at about the same level as age (–0.23).
There are other statistically significant though weak correlations between cognitive factors and condom usage. The correlations presented here are sufficient, however, to support the effectiveness of the social cognitive model approach. It is worth highlighting that in the full analysis the effects of intentions, attitudes, norms, and self-efficacy are notably stronger than those of knowledge about the risks of not using condoms, or of gender or socio-economic status. This evidential emphasis on attitudes, motives, and beliefs has clear implications for health-promotion programmes and especially the need to move beyond the provision of information (Bennett and Murphy 1997).
In the words of the Health Psychology 2000 report (Marks et al. 1998): ‘The mission of professional health psychology is to promote and maintain wellbeing through the application of psychological theory, methods and research, taking into account the economic, political, social, and cultural context’. As with most mission statements, it is hard to disagree with such lofty and desirable ambitions. In the context of public health, psychologists encounter various enablements and constraints in trying to achieve this mission. Returning to the biopsychosocial model, psychologists are well placed to play various ‘bridging’ roles across the three levels of analysis. Their social psychological accounts of attitudes and decision-making in individuals are needed to inform population-level interventions aimed at enhancing health. Psychophysiology, which has a long history in psychology, helps to explain how individual behaviour can result in disease processes. Furthermore, psychological analyses are also needed to help to explain the impact of social forces on biological processes. In attempting to work within this biopsychosocial perspective, psychologists have a wide range of theories and methods to draw on, as exemplified in this chapter. Moreover, since they are interested in behaviour and behaviour change, they are used to seeking effective explanations and interventions.
The main constraints that psychologists face in developing their contributions to public health are implied by the phrase above ‘taking into account the economic, political, social and cultural context’. Most health psychologists are not trained or used to thinking in these terms. Some progress may be made by extending the training of health psychologists into other disciplines and by their participation in multidisciplinary teams (Marks et al. 1998). But ‘taking into account’ ultimately requires theoretical work that spans the multiple levels of analysis found in the public health domain. So, although psychologists have the potential to help build the bridges described above, they face the major theoretical challenge of locating their explanations of individual behaviour in its multiple contexts. An increasing number of psychologists, notably qualitative researchers, are engaging with this fundamental problem, and reflecting on their ideological role in the health arena (Murray and Chamberlain 1999). In their short history health psychology and behavioural medicine have already made extensive contributions to public health. This will no doubt continue, but the potential for even more fundamental contributions remains to be developed.
Abberley, P. (1992). Counting us out: a discussion of the OPCS disability surveys. Disability, Handicap and Society, 7, 139–55.
Aboud, F.E. (1998). Health psychology in global perspective. Sage, Thousand Oaks, CA.
Abraham, C.S. and Hampson, S.E. (1996). Special issue on controversy and method in the interpretation of verbal reports in health psychology research. Psychology and Health, 11.
Ajzen, I. (1991). The theory of planned behaviour. Organizational Behaviour and Human Decision Processes, 50, 179–211.
Albrecht, G.L. and Devlieger, P.J. (1999). The disability paradox: high quality life against all odds. Social Science and Medicine, 48, 977–88.
Alderson, P. (1998) The importance of theories in health care. British Medical Journal, 317, 1007–10.
Amin, K. and Oppenheim, C. (1992). Poverty in black and white. Child Poverty Action Group/Runnymede Trust, London.
Anastasi, A. and Urbina, S. (1997). Psychological testing (7th edn). Prentice-Hall, Englewood Cliffs, NJ.
Annandale, E. (1998). The sociology of health and medicine: a critical introduction. Polity Press, Cambridge.
Arber, S. (1993). Designing samples. In Researching social life (ed. N. Gilbert). Sage, London.
Ashton, J. and Seymour, H. (1988). The new public health. Open University Press, Buckingham.
Atkinson, P. (1981). The clinical experience. The construction and reconstruction of medical reality. Gower, London.
Balarajan, R. and Bulusu, L. (1990). Mortality among immigrants in England and Wales 1979–1983. In Mortality and geography: a review in the mid-1980s (ed. M. Britton), Vol. 9. OPCS, London.
Barbour, R. (1999). Are focus groups an appropriate tool for studying organisational change? In Developing focus group research (ed. R. Barbour and J. Kitzinger). Sage, London.
Barbour, R. and Kitzinger, J. (ed.) (1999). Developing focus group research. Sage, London.
Barnes, C. (1991). Disabled people in Britain and discrimination. Hurst, London.
Bartlett, D. (1998). Stress: perspectives and processes. Open University Press, Buckingham.
Baumann, Z. (1987). Legislators and interpreters: on modernity, postmodernity and intellectuals. Polity Press, Cambridge.
Beck, U. (1992). Risk society: towards a new modernity. Sage, London.
Becker, M.H. (1974). The health belief model and sick role behavior. Health Education Monographs, 2, 409–19.
Beeken, S. and Waterston, T. (1992). Health service support for breastfeeding—are we practising what we preach? British Medical Journal, 305, 285–7.
Bennett, P. and Murphy, S. (1997). Psychology and health promotion. Open University Press, Buckingham.
Berger, A. (1998). Why doesn’t audit work? British Medical Journal, 316, 875–6.
Bhaskhar, R. (1989). A philosophical critique of the contemporary human sciences. Harvester Wheatsheaf, Hemel Hempstead.
Bhopal, R. (1997). Is research into ethnicity and health racist, unsound or important science? British Medical Journal, 314, 1751–6.
Bickenbach, J.E., Chatterji, S., Badley, E.M., and Ustun, T.B. (1999). Models of disablement, universalism and the international classification of impairments, disabilities and handicaps. Social Science and Medicine, 48, 1173–87.
Black, N. (1994). Why we need qualitative research (editorial). Journal of Epidemiology and Community Health, 48, 425–6.
Blaxter, M. (1976). The meaning of disability. Heinemann, London.
Bloor, M. and Horobin, G. (1975). Conflict and conflict resolution in doctor–patient interactions. In A sociology of medical practice (ed. C. Cox and A. Mead). Collier Macmillan, London.
Booth-Kewley, S. and Friedman, H.S. (1987). Psychological predictors of heart disease: a quantitative review. Psychological Bulletin, 101, 343–62.
Bosk, C.L. (1981). Forgive and remember. Managing medical failure. University of Chicago Press, London.
Bowler, I. (1993). ‘They’re not the same as us': midwives stereotypes of South Asian descent maternity patients. Sociology of Health and Illness, 15, 157–78.
Bowling, A. (1991). Measuring health. Open University Press, Buckingham.
Bowling, A. (1995). Measuring disease. Open University Press, Buckingham.
Bradby, H. (1997). Health, eating and heart attacks: Glaswegian Punjabi women’s thinking about everyday food. In Food, health and identity (ed. P. Caplan), pp. 213–33. Routledge, London.
Brown, P.C. and Smith, T.W. (1992). Social influence, marriage and the heart: cardiovascular consequences of interpersonal control in husbands and wives. Health Psychology, 11, 88–96.
Buchanan, D. (1992). An uneasy alliance: combining qualitative and quantitative research methods. Health Education Quarterly, 19, 117–35.
Bury, M. (1986). Social constructionism and the development of medical sociology. Sociology of Health and Illness, 8, 137–69.
Bury, M. (1996). Defining and researching disability: challenges and responses. In Exploring the divide: illness and disability (ed. C. Barnes and G. Mercer). Disability Press, Leeds.
Bury, M. (ed.) (1997). Chronic illness and disability. In Health and illness in a changing society. Routledge, London.
Calnan, M. (1984). The politics of health: the case of smoking control. Journal of Social Policy, 13, 279.
Campbell, D.T. and Fiske, D.W. (1959). Convergent and discriminant validation by the multitrait-multimethod matrix. Psychological Bulletin, 56, 81–105.
Campbell, J. and Oliver, M. (1996). Disability politics: understanding our past, changing our future. Routledge, London.
Chewning, B. and Sleath, B. (1996). Medication decision-making and management: a client-centred model. Social Science and Medicine, 42, 389–98.
Cheyne, H., Turnbull, D., Lunan, C.B., Reid., M., and Greer, I.A. (1995). Working alongside a mid-wife led unit; what do obstetricians think? British Journal of Obstetrics and Gynaecology, 102, 485–7.
Cohen, S., Kessler, R.C., and Gordon, L.U. (1995). Measuring stress: a guide for health and social scientists. Oxford University Press.
Conner, M. and Norman, P. (ed.) (1996). Predicting health behaviour. Open University Press, Buckingham.
Conner, M. and Norman, P. (1998). Health behavior. In Comprehensive clinical psychology (ed. D. Johnston and M. Johnston), Vol. 8, Health psychology, pp. 1–37. Elsevier, Oxford.
Cornford, C. and Morgan, M. (1999) Elderly peoples’ beliefs about influenza vaccination. British Journal of General Practice, 49, 519–21.
Cornwell, J. (1984). Hard earned lives. Tavistock Press, London.
Cronbach, L.J., Gleser, G.C., Nanda, H., and Rajaratnam, N. (1972). The dependability of behavioral measurements: theory of generalizability of scores and profiles. Wiley, New York.
Cunningham-Burley, S., Kerr, A., and Pavis, S. (1999). Theorising subjects and subject matter in focus group research. In Developing focus group research (ed. R. Barbour, and J. Kitzinger), pp. 186–99. Sage, New York.
Currer, C. (1986). Concepts of mental well- and ill-being: the case of Pathan mothers in Britain. In Concepts of health and illness (ed. C. Currer and M. Stacey). Berg, Leamington Spa.
Daly, J., Kellehear, A., and Gliksman, M. (1997). The public health researcher: a methodological guide. Oxford University Press.
Davis, F. (1963). Passage through crisis: polio victims and their families. Bobbs-Merrill, Indianapolis, IN.
Davison, C., Davey Smith, G., and Frankel, S. (1991). Lay epidemiology and the prevention paradox: the implications of coronary candidacy for health education. Sociology of Health and Illness, 13, 1–19.
Department of Health (1993). Changing childbirth, Parts 1 and 2. Report of the Expert Maternity Group. HMSO, London.
DiMatteo, M.R. (1991). The psychology of health, illness, and medical care. Wadsworth, Belmont, CA.
Donovan, J.L. and Blake, D.R. (1992). Patient non-compliance: deviance or reasoned decision-making? Social Science and Medicine, 34, 507–13.
Douglas, J. (1992). Black womens’ health matters: putting black women on the research agenda. In Women’s health matters (ed. H. Roberts), pp. 33–45. Routledge, London.
Douglas, J. (1998). Meeting the health needs of women from black and ethnic minority communities. In Women and health services (ed. L. Doyal). Open University Press, Buckingham.
Dowler, E. and Calvert, C. (1995). Nutrition and diet in lone-parent families in London. Family Policy Studies Centre, London.
Doyal, L. (1981). The political economy of health. Pluto Press, London.
East, L. and Robinson, J. (1994). Changes in process; bringing about change in health care through action research. Journal of Clinical Nursing, 3, 57–61.
Engel, G.L. (1977). The need for a new medical model: a challenge for biomedicine. Science, 196, 129–36.
Fenton, S. and Sadiq-Sangster, A. (1996). Culture, relativism and the expression of mental distress: South Asian women in Britain. Sociology of Health and Illness, 2, 66–85.
Fenwick, N., Morgan, M., McKenzie, C., and Wolf, C. (1998). General practitioners’ attitudes to the development of midwifery group practices. British Journal of General Practice, 48, 1395–8.
Fine, G. (1993). The sad demise, mysterious disappearance, and glorious triumph of symbolic interactionism. Annual Review of Sociology, 19, 61–87.
Foucault, M. (1976). The birth of the clinic: an archaeology of medical perception. Tavistock Press, London.
Freidman, M. and Rosenman, R.H. (1974). Type A behaviour and your heart. Wildwood House, London.
Freidson, E. (1970). The profession of medicine. University of Chicago Press.
Freidson, E. (1994). Professionalism reborn. Polity Press, Cambridge.
French, S. (1993). Disability, impairment or something in-between? In Disabling barriers—enabling environments (ed. J. Swain, V. Finkelstein, S. French, and M. Oliver) Sage, London.
French, N.K. and Chopra, R.V. (1999). Parent perspectives on the roles of paraprofessionals. Journal of the Association for Persons with Severe Handicaps, 24, 259–72.
Giddens, A. (1984). The constitution of society. Polity Press, Cambridge.
Giddens, A. (1989). Sociology. Polity Press, London.
Giddens, A. (1991). Modernity and self-identity. Polity Press, Cambridge.
Giddens, A. (1994). Beyond left and right: the future of radical politics. Polity Press, Cambridge.
Glazer, N. and Strauss, A. (1965). Awareness of dying. Aldine, Chicago, IL.
Glazer, N. and Strauss, A. (1967). The discovery of grounded theory. Aldine, Chicago, IL.
Goffman, E. (1961). Asylums. Penguin, Harmondsworth.
Greenwood, J.D. (1989). Explanation and experiment in social psychological science. Springer-Verlag, New York.
Grimm, L.G. and Yarnold, P.R. (ed.) (1995). Reading and understanding multivariate statistics. American Psychological Association, Washington, DC.
Hall, J.A., Milburn, M.A., Roter, D.L., and Daltroy, L.H. (1998). Why are sicker patients less satisfied with their medical care? Tests of two explanatory models. Health Psychology, 17, 70–5.
Hamilton, D. (1994). Traditions, preferences and postures in applied qualitative research. In Handbook of qualitative research (ed. N. Denzin and Y. Lincoln). Sage, London.
Hammersley, M. (1990). Reading ethnographic research. Longman, London.
Hammersley, M. and Atkinson, P. (1997). Ethnography, principles in practice. Routledge, London.
Harré, R. (1993). Social being (2nd edn). Basil Blackwell, Oxford.
Harrison, S. and Pollitt, C. (1995). Controlling health professionals; the future of work and organisation within the NHS (2nd edn). Open University Press, Buckingham.
Hart, E. and Bond, M. (1995). Action research for health and social care. Open University Press, Buckingham.
Hart, G. and Flowers, P. (1996). Recent developments in the sociology of HIV risk behaviour. Risk, Decision and Policy, 1, 153–65.
Hayduk, L.A. (1996). LISREL issues, debates and strategies. Johns Hopkins University Press, Baltimore, MD.
Haynes, S.G., Feinleib, M., and Kannel, W.B. (1980). The relationship of psychosocial factors to coronary heart disease in the Framingham study. American Journal of Epidemiology, 111, 37–58.
Higgs, P. and Scambler, G. (1998) Explaining health inequalities: how useful are concepts of social class? In Modernity, medicine and health: medical sociology towards 2000 (ed. G. Scambler and P Higgs). Routledge, London.
Hipler, H-J., Schwarz, N., and Sudman, S. (ed.) (1987). Social information processing and survey methodology. Springer-Verlag, New York.
Hlatky, M.A., Lam, L.C., Lee, K.L., et al. (1995). Job strain and the prevalence and outcome of coronary artery disease. Circulation, 92, 327–33.
Hoffer, C.R. and Schuler E.A. (1948). Measurement of health needs and health care. American Sociological Review, 13, 719–24.
Homans, H. (1983). ‘A question of balance': Asian and British women’s perceptions of food during pregnancy. In The sociology of food and eating—essays in the social significance of food (ed. A. Murcott). Gower, Aldershot.
House, J.S., Landis, K.R., and Umberson, D. (1988). Social relationships and health. Science, 241, 540–5.
Illich, I. (1977). Limits to medicine. Medical nemesis: the expropriation of health. Penguin, Harmondsworth.
Jagger, A. and Rothenberg, P. (1984). Feminist frameworks. McGraw-Hill, New York.
Jansen, M.A. and Weinman, J. (ed.) (1991). The international development of health psychology. Harwood, Philadelphia, PA.
Jefferys, M., Nullard, J.B., Hyman, M., and Warren, M.D. (1969). A set of tests for measuring motor impairment in prevalence studies. Journal of Chronic Diseases, 28, 303–9.
Johnson, T. (1972). Professions and power. Macmillan, London.
Johnston, M. and Johnston, D.W. (1998). Assessment and measurement issues. In Comprehensive clinical psychology, (ed. D. Johnston and M. Johnston), Vol. 8, Health psychology, pp. 113–35. Elsevier, Oxford.
Johnston, M., Wright, S., and Weinman, J. (1995). Measures in health psychology: a user’s portfolio. NFER-Nelson, Windsor.
Jones, T. (1992). Britain’s ethnic minorities. Policy Studies Institute, London.
Kai, J. (1996). What worries parents when their pre-school children are acutely ill, and why: a qualitative study. British Medical Journal, 313, 983–6.
Kaplan, R.M., Sallis, J.F., and Patterson, T.L. (1990). Health and human behavior. McGraw-Hill, New York.
Kelleher, D. (1988). Coming to terms with diabetes: coping strategies and non-compliance. In Living with chronic illness (ed. R. Anderson and M. Bury), pp. 136–55. Unwin-Hyman, London.
Kelly, M.P. (1992). Colitis. Routledge, London.
Kelly, M.P. and Field, D. (1998). Conceptualising chronic illness. In Sociological perspectives on health, illness and health care (ed. D. Field and S. Taylor). Blackwell Science, Oxford.
Keppel, G. (1991). Design and analysis: a researcher’s handbook (3rd edn). Prentice-Hall, Englewood Cliffs, NJ.
Kirk, R.E. (1995). Experimental design: procedures for the behavioral sciences (3rd edn). Brooks/Cole, Pacific Grove, CA.
Kitzinger, S. (1992). Birth and violence against women: generating hypotheses from women’s accounts of unhappiness after childbirth. In Women’s health matters (ed. H. Roberts). Routledge, London.
Kitzinger, J. (1994). The methodology of focus groups: the importance of interaction between research participants. Sociology of Health and Illness, 116, 103–21.
Kitzinger, J. and Farquer, C. (1999). The analytical potential of ‘sensitive moments’ in focus group discussions. In Developing focus group research (ed. R. Barbour and J. Kitzinger). Sage, London.
Klem, L. (1995). Path analysis. In Reading and understanding multivariate statistics (ed. L.G. Grimm and P.R. Yarnold), pp. 65–97. American Psychological Association, Washington, D.C.
Lansbury, H. (2000). Chronic pain management: a qualitative study of elderly people’s preferred coping strategies and barriers to management. Disability and Rehabilitation, 22, 2–14.
Lanza, M.L. and Ericsson, A. (2000). Consumer contributions in developing clinical practice guidelines. Journal of Nursing Care Quality, 14, 33–40.
Lazarus, R.S. (1990). Stress, coping and illness. In Personality and disease (ed. H.S. Freeman), pp. 97–120. Wiley, New York.
Lazarus, R.S. (1991). Emotion and adaptation. Oxford University Press, New York.
Lazarus, R.S. and Folkman, S. (1984). Stress, appraisal and coping. Springer, New York.
Lewin, K. (1946). Action research and minority problems. Journal of Social Issues, 2, 34–46.
Locker, D. (1983). Disability and disadvantage: the consequences of chronic illness. Tavistock Press, London.
Lofland, J. and Lofland, L. (1995). Analysing social settings. Wadsworth, London.
Loftus, E.L., Smith, K.D., Klinger, M.R., and Fiedler, J. (1991). Memory and mismemory for health events. In Questions about questions: inquiries into the cognitive bases of surveys (ed. J.M. Tanur), pp. 102–37. Russell Sage Foundation, New York.
Lonsdale, S. (1990). Women and disability: the experience of physical disability among women. Macmillan, London.
McColl, E., Jacoby, A., Thomas, L., et al. (1998). Designing and using patient and staff questionnaires. In Health services research methods (ed. N. Black, J. Brazier, R. Fitzpatrick, and B. Reeves). BMJ Books, London.
Macdonald, L. (1988). The experience of stigma: living with rectal cancer. In Living with chronic illness: the experience of patients and their families (ed. R. Anderson and M. Bury). Unwin-Hyman, London.
McDowell, I. and Newell, C. (1987). Measuring health: a guide to rating scales and questionnaires. Oxford University Press.
McKeganey, N. and Barnard, M. (1992). AIDS, drugs and sexual risk. Open University Press, Buckingham.
McKinlay, J. and Stoeckle, J. (1988). Corporatisation and the social transformation of doctoring. International Journal of Health Services, 18, 191–205.
McLaren, N. (1998). A critical review of the biopsychosocial model. Australian and New Zealand Journal of Psychiatry, 32, 86–92.
Marks, D.F., Bruecher-Albers, C., Donker, F.J., et al. (1998). Health Psychology 2000: the development of professional health psychology; EFPPA Task Force on Health Psychology final report. Journal of Health Psychology, 3, 149–60.
Marmot, M., Adelstein, A., and Bulusu, L. (1984). Immigrant mortality in England and Wales 1970–1978. OPCS Studies on Population and Medical Subjects, No. 47, HMSO, London.
Martin, D. and McQueen, D. (ed.) (1989). Readings for a new public health. Edinburgh University Press.
Martin, J., Mettzer, H., and Elliot, D. (1988). The prevalence of disability among adults. HMSO, London.
Matarazzo, J.D. (1980). Behavioral health and behavioral medicine: frontiers for a new health psychology. American Psychologist, 35, 807–17.
Matthews, K.A. (1988). Coronary heart disease and type A behaviors: update on and alternative to the Booth-Kewley and Friedman (1987) quantitative review. Psychological Bulletin, 104, 373–80.
Mead, G.H. (1934). Mind, self and society from the standpoint of social behaviourism. University of Chicago Press.
Mechanic, D. (1962). Sources of power of lower participants in complex organisations. Administration Science Quarterly, 7, 349–64.
Melia, K. (1997). Producing ‘plausible stories': interviewing student nurses. In Context and method in qualitative research (ed. G. Miller and R. Dingwall). Sage, London.
Meyer, J. (2000). Using qualitative methods in health related action research. British Medical Journal, 320, 178–81.
Miller, T.Q., Smith, T.W., Turner, C.W., Guijarro, M.L., and Hallet, A.J. (1996). A meta-analytic review of research on hostility and physical health. Psychological Bulletin, 119, 322–48.
Morgan, M., Patrick, D.L., and Charlton, J.R. (1984). Social networks and psychosocial support among disabled people. Social Science and Medicine, 19, 489–99.
Murphy, E., Dingwall, R., Greatbatch, D., Parker, S., and Watkson, P. (1998). Qualitative research methods in health technology assessment; a review of the literature. Health Technology Assessment, 2.
Murray, M. and Chamberlain, K. (1998). Special issue. Qualitative research. Journal of Health Psychology, 3.
Murray, M. and Chamberlain, K. (1999). Qualitative health psychology: theories and methods. Sage, London.
Navarro, V. (1982). The labor process and health. International Journal of Health Services, 12, 5.
Nazroo, J.Y. (1997). The health of Britain’s ethnic minorities. Findings from a national survey. Policy Studies Institute, London.
Nettleton, S. (1995). The sociology of health and illness. Polity Press, Cambridge.
Neuberger, J. and Tallis, R. (1999). Do we need a new word for patients? British Medical Journal, 318, 1756–7.
NHLBI (National Heart, Lung, and Blood Institute) Task Force (1998). Behavioral research in cardiovascular, lung, and blood health. National Institutes of Health, Bethesda, MD.
Nunnally, J.C. and Bernstein, I.H. (1994). Psychometric theory (3rd edn). McGraw-Hill, New York.
Oakley, A. (1980). Women confined: towards a ociology of childbirth. Martin Robertson, Oxford.
Oakley, A. (1981). Interviewing women; a contradiction in terms. In Doing feminist research (ed. H. Roberts). Routlege and Kegan Paul, London.
Ogden, J. (1997). The rhetoric and reality of psychosocial theories of health: a challenge to biomedicine? Journal of Health Psychology, 2, 21–9.
Oliver, M. (1996). Understanding disability: from theory to practice. Macmillan, London.
Ostrom, T.M. (1987). Bipolar survey items: an information processing perspective. In Social information processing and survey methodology (ed. H.-J. Hipler, N. Schwarz, and S. Sudman), pp. 71–85. Springer-Verlag, New York.
Owens, P. and Petch, H. (1995). Professionals and management. In Interprofessional issues in community and primary health care (ed. P. Owens, J. Carrier, and J. Horder), pp. 37–55. Macmillan, London.
Owens, P., Carrier, J., and Horder, J. (1995). Interprofessional issues in community and primary health care. Macmillan, London.
Parsons, T. (1951). The social system. Free Press, New York.
Patrick, D.L. and Peach, H. (1989). Disablement in the community. Oxford University Press.
Pavis, S. and Cunningham-Burley, S. (1999). Male youth street culture; understanding the context of health-related behaviours. Health Education Research: Theory and Practice, 14, 583–96.
Pearce I.H. and Crocker L.H. (1949). The Peckham experiment. Allen and Unwin, London.
Pearson, M. (1989). Sociology of race and health. In Ethnic factors in health and disease (ed. J. Cruickshank and D. Beevers). Wright, Sevenoaks.
Pinder, R. (1990). The management of chronic illness: patient and doctor perspectives on Parkinson’s disease. Macmillan, London.
Pollack, M., Poucheler, G., and Pierret, J. (1992). AIDS: a problem for sociological research. Sage, London.
Popay, J., Rogers, A., and Williams, G. (1998a). Rationale and standards for the systematic review of qualitative literature in health services research. Qualitative Health Research, 8, 341–51.
Popay, J., Williams, G., Thomas, C., and Gatrell, T. (1998b). Theorising inequalities in health: the place of lay knowledge. Sociology of Health and Illness, 20, 619–44.
Pope, C. (1991). Trouble in store: some thoughts on the management of waiting lists. Sociology of Health and Illness, 13, 194–212.
Pope, C. and Mays, N. (ed.) (2000). Qualitative research in health care (2nd edn). BMJ Books, London.
Powles, J. (1973). On the limitations of modern medicine. Science, Medicine and Management, 1, 1–30.
Price, V.A. (1983). Type A behavior pattern: a model for research and practice. Academic Press, New York.
Robinson, I. (1988). Reconstructing lives: renegotiating the meaning of multiple sclerosis. In Living with chronic illness (ed. R. Anderson and M. Bury), pp. 43–66. Unwin-Hyman, London.
Rogers R.W. (1983). Cognitive and physiological processes in fear appeals and attitude change: a revised theory of protection motivation. In Social psychophysiology: a sourcebook (ed. J.T. Cacioppo and R.A. Petty), pp. 153–76. Guilford, New York.
Rosenthal, R. (1994). Interpersonal expectancy effects: a 30-year perspective. Current Directions in Psychological Science, 3, 176–9.
Savage, M. and Miles, A. (1994). The remaking of the British working class. Routlege, London.
Scambler, G. (1989). Epilepsy. Routledge, London.
Scambler, G. and Hopkins, R. (1986). Being epileptic: coming to terms with stigma. Sociology of Health and Illness, 8, 26–43.
Schnall, P.L., Landsbergis, P.A., and Baker, D. (1994). Job strain and cardiovascular disease. Annual Review of Public Health, 15, 381–411.
Schneiderman, N. and Tapp, J.T. (ed.) (1985). Behavioral medicine: the biomedical approach. Lawrence Erlbaum, Hillsdale, NJ.
Schutz, A. (1962). Collected papers, Vol. 1. Martinus Nijhoff, The Hague.
Schutz, A. (1964). Collected papers, Vol. 2. Martinus Nijhoff, The Hague.
Schwalbe, M.L. and Staples, C.L. (1986). Class position, work experience and health. International Journal of Health Services, 16, 583.
Schwartz, M., Landis, S.E., Rowe, J.E, Janes, C.L., and Pullman, N. (2000). Using focus groups to assess primary care patients’ satisfaction. Evaluation and the Health Professions, 23, 58–71.
Scott, R. (1969). The making of blind men. Russell Sage Foundation, New York.
Seale, C. (1999). Quality in qualitative research. Qualitative Enquiry, 5, 465–78.
Shavelson, R.J., Webb, N.M., and Rowley, G.L. (1989). Generalizability theory. American Psychologist, 44, 922–32.
Sheeran, P. and Abraham, C. (1996). The health belief model. In Predicting health behaviour (ed. M. Conner and P. Norman), pp. 23–61. Open University Press, Buckingham.
Sheeran, P., Abraham, C., and Orbell, S. (1999). Psychosocial correlates of heterosexual condom use: a meta-analysis. Psychological Bulletin, 125, 90–132.
Shumaker, S.A. and Czajkowski, S.M. (ed.) (1994). Social support and cardiovascular disease. Plenum, New York.
Shuttleworth, S. (1990). Female circulation: medical discourse and popular advertising in the mid-Victorian era. In Body/politics: women and discourses of science (ed. M. Jacobus, E.F. Keller, and S. Shuttleworth). Routlege, London.
Siegman, A.W. and Smith, T.W. (1994). Anger, hostility, and the heart. Lawrence Erlbaum, Hillsdale, NJ.
Smaje, C. (1995). Health, ‘race’ and ethnicity; making sense of the evidence. King’s Fund Institute, London.
Spicer, J. (1997). Systems analysis of stress and coping: a testing proposition. Journal of Health Psychology, 2, 167–70.
Spicer, J. and Chamberlain, K. (1996). Developing psychosocial theory in health psychology: problems and prospects. Journal of Health Psychology, 1, 161–71.
Spicer, J., Jackson, R., and Scragg, R. (1996). Type A behaviour, social contact and coronary death. Psychology and Health, 11, 733–43.
Steptoe, A. (1998). Psychophysiological bases of disease. In Comprehensive clinical psychology (ed. D. Johnston and M. Johnston). Vol. 8, Health psychology, pp. 39–78. Elsevier, Oxford.
Strauss, A., et al. (1963). The hospital and its negotiated order. In The hospital in modern society (ed. E. Freidson). Free Press, New York.
Streiner, D.L. and Norman, G.R. (1995). Health measurement scales: a practical guide to their development and use (2nd edn). Oxford University Press.
Strube, M.J. (ed.) (1991). Type A behavior. Sage, Newbury Park, CA.
Suchman, L. and Jordan, B. (1991). Validity and the collaborative construction of meaning in face-to-face surveys. In Questions about questions: inquiries into the cognitive bases of surveys (ed. J.M. Tanur), pp. 241–67. Russell Sage Foundation, New York.
Tabachnick, B.G. and Fidell, L.S. (1996). Using multivariate statistics (3rd edn). Harper Collins, New York.
Tanur, J.M. (ed.) (1991). Questions about questions: inquiries into the cognitive bases of surveys. Russell Sage Foundation, New York.
Taylor, S.E. (1995). Health psychology (3rd edn). McGraw-Hill, New York.
Titchen, A. and Binnie, A. (1993). Research partnershsips: collaborative action research in nursing. Journal of Advanced Nursing, 18, 858–65.
Tonkiss, F. (1998). Analysing discourse. In Researching society and culture (ed. C. Seale). Sage, London.
Tucker, J., Hall, M.H., Howie, P.W.M, et al. (1996). Should obstetricians see women with normal pregnancies? A multi-centre randomised controlled trial of routine antenatal care by general practitioners and midwives compared with shared care led by obstetricians. British Medical Journal, 312, 554–9.
Turnbull, D., Holmes, A., Shields, N., et al. (1996). Randomised, controlled trial of efficacy of midwife-managed care. Lancet, 348, 213–18.
Turner, B.S. (1987). Medical power and social knowledge. Sage, London.
Turner, B.S. (1992). Regulating bodies; essays in medical sociology. Routledge, London.
Turner, J.R. (1994). Cardiovascular reactivity and stress. Plenum, New York.
Uchino, B.N., Cacioppo, J.T., and Kiecolt-Glaser, J.K. (1996). The relationship between social support and physiological processes: a review with emphasis on underlying mechanisms and implications for health. Psychological Bulletin, 119, 488–531.
Voysey, M. (1975). A constant burden: the reconstitution of family life. Routledge and Kegan Paul, London.
Walters, A.J. (1995). A hermaneutic study of the experiences of relatively of critically ill patients. Journal of Advanced Nursing, 22, 998–1005.
WHO (World Health Organization) (1980). International classification of impairments, disabilities and handicaps. WHO, Geneva.
WHO (World Health Organization) Quality of Life Group (1998). The World Health Organization quality of life assessment (WHOQOL): development and general psychometric properties. Social Science and Medicine, 46, 1569–85.
Wilcocks, S. (1999). Clinical management and cultural diversity: the cultural context of doctor involvement in the managerial process. Health Services Management Research, 12, 212–16.
Williams, R.G.A. (1983). Concepts of health and illness: an analysis of lay logic. Sociology, 17, 182–205.
Williams, G. (1984). The genesis of chronic illness: narrative reconstruction. Sociology of Health and Illness, 6, 175–200.
Williams, S. and Calnan, M. (1994). Perspectives on prevention: the views of General Practitioners. Sociology of Health and Illness, 16, 372–93.
Williams, R.D. and Clark, A.J. (2000). A qualitative study of women’s hysterectomy experience. Journal of Women’s Health and Gender-based Medicine, 9, S15–25.
Wing, J.K. and Brown, G.W. (1970). Institutionalism and schizophrenia. Oxford University Press.
Zola, I. (1975). Culture and symptoms: an analysis of patients’ presenting complaints. In A sociology of medical practice (ed. C. Cox and A. Mead). Collier-Macmillan, London.