Competing Paradigms and Health Research

Full text

Competing Paradigms and
Health Research
ALEX BROOM AND EVAN WILLIS
INTRODUCTION
From newspaper and television documentaries to web sources and acade-
mic journals, research on health abounds. As indicated in Chapter 1, the
broad area of ‘health research’ uses a plurality of research methodologies
and related methods, from documentary analysis and unstructured inter-
views to randomized controlled trials and experimental methods. However,
there are differences between particular methodological approaches that are
systematic rather than idiosyncratic. In this chapter we make use of the soci-
ological concept of ‘paradigms’ to illustrate the way in which research
methodologies are embedded in particular political and ideological posi-
tions. This chapter aims to provide a critical overview of what we might
broadly call research paradigms. We explore the nature of a paradigm,
examining different types of paradigms in health research, and the implica-
tions of using a particular paradigm for knowledge production.
Furthermore, by giving the reader a critical understanding of the nature of
particular research methodologies and associated methods, we aim to pro-
vide a means to establish the most effective way of answering research ques-
tions about health and health care.
2
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 16

Competing Paradigms and Health Research
17
In order to do this, we need to have a clear understanding of the epistemo-
logical foundations and philosophical roots of particular forms or styles of
research. Epistemology in this context refers to the nature of knowledge or how
we come to know certain things about the world. In the context of health care,
an understanding is necessary of how certain research methods and certain
approaches to data collection have emerged from vastly different traditions,
and, importantly, how they produce different understandings of the social
world.
Within this discussion we examine broad traditions in health research, from
positivism to interpretivism, exploring the implications of these traditions and
the various methodological approaches derived from them for health research.
We seek not to make an argument about which methodological paradigm is
best, but rather to provide the reader with a critical understanding of how the
methodologies presented in the following chapters have emerged from, and
contribute to, the reproduction of particular understandings of the social and
natural world. We propose the view, also reflected in subsequent chapters, that
certain phenomena, and thus certain research questions, are best explored by
drawing on one particular paradigm rather than another. We present the reader
with a critical view of each, and how this can be used to inform research design
and analysis.
In this chapter, we draw on a particular research study to illustrate the dif-
ference. The research study examined the impact of the Internet on health care
services, using men and medical specialists’ experiences of prostate cancer as a
case study. It was conducted by one of the authors for purposes of a higher
degree under the supervision of the other (see Broom 2004).
DEFINING TERMS: WHAT IS A PARADIGM?
A paradigm can be defined as an overarching philosophical or ideological stance,
a system of beliefs about the nature of the world, and ultimately, when applied
in the research setting, the assumptive base from which we go about producing
knowledge (Rubin and Rubin 2005). In the context of health care research, a
researcher’s paradigmatic positioning relates to their understanding of the nature
of knowledge (their epistemological standpoint) and of reality (their ontological
standpoint). An interpretivist researcher will maintain that knowledge is socially
constructed and reality is ultimately subjective. A positivist paradigm, however,
will maintain that reality is fixed and that objective knowledge can be produced
through rigorous methodology. Methodologies utilized by the former are
referred to as qualitative and those by the latter as quantitative.
Health researchers vary greatly in their epistemological and ontological
standpoints, and these paradigmatic differences have an important influence
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 17

on their study objective and design, and thus on the type of knowledge they
produce from their research. Nonetheless, researchers may actually draw on a
number of different paradigmatic positions to achieve the best possible out-
come in terms of knowledge production. In fact, the separation of particular
ontological and epistemological positions can come across as a somewhat arti-
ficial dichotomy, in that health researchers are increasingly pragmatic. They
choose the best means to answer a research question and use a mixed-method
approach, rather than being explicitly philosophically driven. However,
although researchers may embrace disparate methodologies and methods
within a single research project, there is a tendency for the actual analysis,
including the conceptualization of results, to be derived from one distinct
philosophical standpoint and to take a positivist or constructivist/interpretivist
approach.
Nevertheless, it is necessary to confront the issue of the extent to which quan-
titative and qualitative paradigms of research are metaphysically incommensu-
rable. By this we mean the extent to which they are logically inconsistent or
mutually exclusive. For example, ontologically speaking, social reality cannot
be both objectivist in nature and also a social construction. It is either one or the
other. This is another way of saying that these two paradigms of research are in
fact incommensurable and that health research cannot be simultaneously quan-
titative and qualitative. However, it is possible to have separate quantitative
and qualitative components, or, as we sometimes say, there may be different
‘arms’ to a research project.
Traditionally, health researchers have tended to embrace one particular par-
adigm. The commonest split has been between the positivists and the construc-
tivists. It is these schools of thought that will be examined closely in this
chapter. As suggested by Wolfe (1997), this positivist/constructivist split can be
seen as the ‘two faces of the social sciences’. As such, most of the methods out-
lined in this book can broadly be seen as philosophically congruent to one of
these traditions. In saying this, we argue that positivism and constructivism
should be viewed as ‘ideal types’ in the Weberian sense. An ideal type, accord-
ing to Weber, is formed by the one-sided accentuation of one or more points of
view. Hence although the distinction between positivism and constructivism is
useful and important, in practice the boundaries between different paradigms
are blurred (De Vaus 1995).
As reflected in the structure of this book, a paradigmatic difference exists
between what we call qualitative and quantitative research methods. This is not
a matter of study design such as the use of observational methods (for instance,
case–control study or survey) versus experimental methods (for example, the
randomized controlled trial). Rather, the distinction between qualitative and
quantitative approaches lies in the assumptive base that underpins the research
design and thus data collection. A general distinction is made between these
Conducting Health Research
18
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 18

two methodological traditions based broadly on either an inductive or a
deductive approach to research. In the case of induction, the researcher gathers
qualitative data (such as on patients’ experiences of pain) and then establishes
patterns drawing on theory, thereby, for example, establishing how things may
be improved for those patients in a health care setting. However, a deductive,
quantitative approach involves the researcher testing an existing theory or
hypothesis. A hypothesis might, for instance, be that patients who are given
pain killers two days after treatment will report no different experiences than
those who are not. The research then aims to prove or disprove the hypothesis.
Quantitative and qualitative methodologies are different strategies used by
health researchers with quite different visions of the social world. These are the
product of a number of very important Western philosophical traditions that
have been developing for centuries. As such, social scientists have for long chal-
lenged one another – as well as scientists in the natural sciences. Some argue
that quantification using a positivist paradigm is essential for an objective and
rigorous investigation. Others argue from an interpretivist approach that no
description could be complete without a qualitative understanding of the sub-
jective meanings of social actors involved in social interactions (Glassner and
Moreno 1989).
Our view is that such arguments are ideological and neither quantitative nor
qualitative methods should be viewed as superior. Rather, each serves a partic-
ular purpose. Methodologies are not neutral tools and health researchers need
to be able to think critically about the ways in which certain methods produce
certain types of knowledge. In the twentieth century health research was dom-
inated by positivist approaches to knowledge production epitomized by the
randomized controlled trial. This involves randomization of patients into at
least two different intervention groups (normally usual care versus experimen-
tal treatment) discussed in detail in Chapter 12. However, in recent decades
there has been greater scepticism about the positivist methods associated with
modern medicine. The aim of the rest of this chapter is to highlight the politi-
cal and ideological nature of health research by outlining the philosophical
underpinnings and paradigmatic bases of the research methods discussed in
the following chapters.
POSITIVIST PARADIGM
Roots of positivism
Positivism is a somewhat ambiguous and loaded philosophical idea, but it con-
tinues to form the paradigmatic basis for much health research today. The term
positivism has represented a means to critique the natural sciences and
Competing Paradigms and Health Research
19
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 19

researchers espousing a scientific model within the social sciences. By default,
a positivist assumes that reality is concrete and objectivity is achievable –
therefore the notion of science being ideologically driven is supplanted by the
notion of science as transcending ideology – a rather circular logic. Rather than
defining a clear set of practices, positivism represents a broad tradition of
thought that assumes that reality is constant. It can be measured, and thus the
concepts and methods employed in the natural sciences can be applied to form
a ‘science of man’ (Giddens 1987).
Of course, the immediate dilemma is that what constitutes ‘science’ within
the natural sciences is not an entirely coherent and consistent entity, with a
number of internal differences in the conceptual base of each method. Thus, the
notion of ‘positivism’ within the social sciences as the imposition of a natural
‘science’ model imposed on the social world is at best problematic. Rather,
positivism in the social sciences should be seen as a tradition followed by cer-
tain social scientists who aim to follow the basic premises of scientific enquiry.
This includes the assumption that the researcher is able to collect and interpret
social facts objectively; produce laws and models of behaviour from these
social facts (Bryman 2001); and view ‘scientifically’ produced data as ultimately
neutral and unbiased (Rubin and Rubin 2005).
Positivism has its roots in a long history of Western philosophy and the
agenda for early Enlightenment thinkers was to ‘enlighten’ the masses. The
belief was that human reason could be used to combat ignorance, superstition
and tyranny, and, ultimately, be used to build a better world. Gradually scien-
tific method and quantification would come to be seen as the means to improve
both the natural and the social worlds (Glassner and Moreno 1989).
The actual notion of positivism per se was first coined by Auguste Comte.
Comte saw sociology as the very culmination of positivism: a science of man
that would complete the historical evolution of the hierarchy of the scientific
disciplines (Giddens 1987). Positivism was thus cemented in sociology
through Comte’s idea of transforming society on the basis of science. Herbert
Spencer’s organic theory of society built on this. Spencer conceived of society
as an organism and held that human society reflects the same evolutionary
principles in its development as biological organisms. The dominance of pos-
itivism in the social sciences came to a peak in the post-war United States with
the structural functionalist work of Talcott Parsons and Robert Merton,
only then to wane in the late twentieth century with the emergence of post-
structuralist and postmodernist paradigmatic – a shift discussed later in this
chapter.
These philosophical ideas have evolved considerably since Comte and
Spencer and still form the basis for much health research today. Of course, there
are now much more effective and systematic tools that are used to achieve the
so-called ‘truth’ or produce a ‘scientific fact’. However, the basic premise of a
Conducting Health Research
20
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 20

reality that can be objectively measured, given the right instruments and
conditions, still dominates health research in the clinical and social sciences.
Positivism in health research: From randomized
controlled trials to social surveys
As we will see in the following chapters on quantitative methods, there are
many different methods for advancing our knowledge of health and illness
within a positivist framework. Trial designs, and particularly randomized con-
trolled trials, are the most commonly used designs in medical research,
although medical researchers are increasingly incorporating social science
methods in their studies. Retrospective cohort and cross-sectional designs are
also based on a positivist assumptive base and are used to map out such things
as aetiology (the characteristics of disease), prognosis (the predicted course of
a disease) and prevalence of disease (the extent of spread of a disease within a
population), as well as other factors that may influence health outcomes or
health behaviour.
These types of approaches have emerged from, and contribute to, the repro-
duction of the positivist paradigm, focusing on establishing objective scientific
facts about disease and the body, through scientific method and quantification.
Based on the biomedical model of disease, such studies approach the patient as
a physical/mechanistic entity that can be measured, controlled and ultimately
manipulated. Assumptions about the objectivity, neutrality and generalizability of
data place most biomedical research methods squarely within the realms of a
positivist paradigm.
Likewise, quantitative methods commonly used by social scientists – such as
structured interviews, surveys and self-completion questionnaires – also seek to
test hypotheses by quantifying human behaviour (Bryman 2001), with the objec-
tive of creating models that can predict behaviour (De Vaus 1995). An example of
this was the pilot research the authors undertook on online communities. This
embraced a positivist paradigm in that it sought to establish and quantify pat-
terns in men’s experiences of online support groups as opposed to face-to-face
support groups. As can be seen below, the knowledge produced through this
paradigm is very different from the findings that emerged from more qualitative
frames of reference. We will return to this theme later in the chapter.
Although randomized controlled trials and quantitative surveys may have
very different foci (for example, disease reoccurrence versus feelings of auton-
omy), and may utilize different tools for testing their hypotheses (for instance,
blood tests versus anxiety scales), they are ultimately commensurable within
this broad paradigm. The methods and features associated with this paradigm
are as follows:
Competing Paradigms and Health Research
21
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 21

• Methods utilizing a positivist paradigm:
— Epidemiological/analytical design strategies (for example, randomized
controlled trials, before and after studies, cohort/incidence studies, cross-
sectional studies)
— Survey research
— Secondary document analysis (for instance, content analysis)
— Structured interviewing
— Systematic reviews (meta-analysis).
• Features of a positivist paradigm:
— Determinism: Phenomena can be predicted from a knowledge of scientific laws.
— Objectivity: The researcher is separate and detached from the patient/partici-
pant, maintaining objectivity, assisted by scientific evidence.
— Quantification: Information is derived from what can be quantified.
— Reliability: Through randomization we can know that our findings support
or negate the hypothesis in the study and can be extrapolated to a larger
population.
— Generalizability: Data are reliable and unbiased and we can therefore general-
ize our findings to the rest of the population.
There are considerable advantages to using methods that are based on a posi-
tivist paradigm, and which seek to use the scientific method to answer important
questions about the effectiveness of health care. For example, randomization is
an effective means of reducing bias in the assessment of how effective an inter-
vention is as it guarantees, assuming it is carried out properly, that treatment
assignment will not be based on patients’ prognostic factors. This prevents the
investigators from assigning, consciously or unconsciously, patients with a
greater likelihood of responding well to a treatment that they hope will be supe-
rior. It increases confidence that the study has identified the most effective treat-
ment. Blinding the researchers and their clients is also a method of eliminating
bias (Altman 2000). Randomized trials that have not used appropriate levels of
blinding tend to show significantly larger treatment effects than blinded studies
(Schulz et al. 1995). Thus, although in the process of randomization the patient is
effectively reduced to a number, and thus to some degree their individuality is
lost, this process has the benefit, amongst other things, of increasing our knowl-
edge of the effects of a drug on an organ, or the effect of a psycho-social interven-
tion when a course of treatment is followed.
Another benefit of a randomized controlled trial design is that consistent and
relatively concrete outcome measures are used (as, for instance, pain, recovery
time or anxiety). In this way, we know that the study is assessing, albeit in a
restricted way, the actual effects of the intervention or the disease process, and
not merely recording people’s subjective feelings about how they have
improved over the course of treatment. We can thus be confident, if we ever get
Conducting Health Research
22
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 22

ill and need treatment, that we are not merely being treated according to an
individual doctor’s preference, or on the basis of the outcomes from a group of
a typical patients, but as a result of a well-conducted trial. In a similar way,
surveys and structured interviews can be extremely useful for assessing such
issues as patients’ views on the quality of treatment information provided and
their experiences of treatment for a particular illness.
Despite the potential benefits of drawing on a positivist paradigm, and uti-
lizing study designs such as clinical trials and surveys, there are also poten-
tially negative implications of which health researchers should be aware.
A critique of positivist approaches to social research
While trial designs and other positivistic research methods have much to offer in
health research, their promotion – which is approached with great zeal in some
quarters – can be misplaced. Many questions about health care are not amenable
to randomized controlled trials. For example, a randomized trial could not
research the question of the number of people who die every year as a result of
waiting too long with severe chest pain before calling an ambulance for ethical
and practical reasons. A randomized controlled trial design cannot answer the
question of why – it cannot explore what is going on in a person’s life and rela-
tionships that may have influenced them to delay seeking help. Herein lies one
of the weaknesses of utilizing a positivist, quantitative methodology.
To be more specific, there has been considerable concern, within both acade-
mic and lay communities, about the epistemological assumptions that inform
the process of categorization in quantitative research. For example, some have
argued that positivist quantitative research designs give insufficient attention
to a person’s lived experience (Rubin and Rubin 2005; Bryman 2001).
Sociologists are increasingly arguing that people’s experiences are complex,
subjective and embedded in specific social and historical contexts (Silverman
2001). Take for example the research undertaken by one of the authors on
Australian men with prostate cancer and their use of the Internet (see Broom
2005a; 2005b). Results from the pilot survey showed the numbers of men who
were more comfortable expressing their concerns and fears in an online envi-
ronment. Although useful, this did not give any indication of why this was the
case. We needed to know the systems of meaning and cultural understandings
that actually shaped these men’s perceptions of face-to-face versus online sup-
port groups that might influence their use. This was only achievable with a
qualitative methodology.
The difficulty is that it is not possible to dissect and categorize social life
objectively, through blinding and randomization procedures (Rubin and Rubin
2005). Furthermore, many qualitative social scientists reject the claims of
Competing Paradigms and Health Research
23
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 23

natural scientists and quantitative social sciences to objectivity; the
mechanisms they use for measuring validity; and their desire for generalizabil-
ity. They argue that it is impossible to talk about understand or communicate
social phenomena without employing a particular language or conceptual
scheme and this is not neutral (De Vaus 1995). It can be argued that all repre-
sentations, like data from a survey, clinical trial or interview-based study, are
socially and culturally constructed (Rubin and Rubin 2005). Researchers choose
the questions and interpret the results. Ultimately, therefore, all knowledge
produced by research is subjective, interpreted, political and, in part, ideologi-
cal driven. This leads to the consideration of an alternative frame of reference –
that of interpretivism or, as it is sometimes called, the constructivist tradition.
INTERPRETIVIST PARADIGM
The constructivist traditions: Dislodging
the positivist canon
Although positivist approaches within medicine and the social sciences have
maintained their dominance for a significant part of the twentieth century, a
strong qualitative or interpretive tradition emerged in the social sciences in the
1960s and 1970s. Grounded theory, symbolic interactionalism and phenome-
nology, and more recently poststructuralism and postmodernism are part of
the interpretivist or constructivist tradition. This is associated with such
authors as Glaser and Strauss (1967), Berger and Luckmann (1967), Geertz
(1973), Lofland and Lofland (1984) and Rubin and Rubin (2005).
Originating at the Chicago School in the 1920s and 1930s, and gathering
momentum in the late 1960s, factions within the social sciences were beginning
to diverge from the positivist and deterministic approaches of the classical
social scientists, refocusing social science on the ways in which meanings are
constructed, negotiated and managed by different individuals and groups
within various social and historical contexts. Before this movement, research
methods were largely quantitative and firmly based on a positivist paradigm.
The interpretive paradigm precipitated the development of the tools we now
know as qualitative methods.
The qualitative methodologies, emerging from methods such as in-depth
interviewing and observation, sought to establish an understanding of people’s
lives, experiences and the subjective meanings that could explain the process of
decision making and action. Rather than seeking to measure or categorize
behaviour or attitudes, interpretivist researchers focused on the understand-
ings of research respondents, pursuing an analysis based on the constructivist
Conducting Health Research
24
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 24

ontological position that individuals actively negotiate meaning. As ontology
refers to the study of the nature of reality, a constructivist ontological view is
that reality is constructed rather than ‘set in stone’. It is not objectively measur-
able, and, furthermore, individuals construct their reality by associating ‘mean-
ing’ with certain events or actions (Bryman 2001).
By approaching an issue from this position, a researcher can establish regu-
lar patterns and irregularities in the meanings associated with particular life
events. For example, there may be social patterning in terms of whether or not
particular signs and symptoms are interpreted as an illness, whether profes-
sional help is sought, and how the chance of a cure is perceived. As Rubin and
Rubin (2005) suggest, interpretive social research is about figuring out what
events mean to research subjects: how people adapt and how they view what
has happened to them and around them (see also Bryman 2001). It is identify-
ing this complexity and subjectivity that Ezzy (2002) and others argue should
underpin qualitative research projects.
Interpretivist approaches to health research
Researchers who have positioned themselves within an interpretivist, construc-
tivist paradigm tend to use qualitative methods, such as in-depth interviews,
focus groups and ethnographic observation. These types of methods are consis-
tent with the interpretivist paradigm, in that they aim to record the types of
data that will enable the researcher to reflect on subjective meanings and inter-
pretations; the social and culturally embedded nature of individual experi-
ences; and the relationship between the researcher and the researched (Rubin
and Rubin 2005). What we see is a shift in focus from a position where the
researcher seeks to observe patterns in group behaviour to a position that seeks
to understand individual experiences of interactions, events and social
processes and identify patterns in these subjective experiences. The methods
and features associated with this paradigm are as follows:
• Methods utilizing an interpretivist/constructivist paradigm:
— In-depth, semi-structured or unstructured interviews
— Observation (participatory or non-participatory)
— Focus groups
— Secondary discourse analysis.
• Features of an interpretivist/constructivist paradigm:
— Interpretivist: Seeks ‘understanding’ with a focus on subjective meanings and
interpretation.
— Naturalistic: Data are collected in the setting of everyday life.
Competing Paradigms and Health Research
25
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 25

— Subjectivity: Research practice and knowledge production is not objective or
neutral — rather it is gendered and partial.
— Complexity: Not so concerned with inference but rather with the depth of
analysis.
— Political: The position of value neutrality is viewed as misleading as it makes the
focus on the research seem independent from social relations.
— Validity: High on validity as it draws on the understandings of research subjects, but
not necessarily generalizable as it relies on the interpretation of the researcher.
Overall, the advantage of a flexible, descriptive and qualitative approach is that
researchers are less likely to become stuck in conventional ways of thinking or
particular sets of assumptions – as the qualitative chapters in this book high-
light. Rather than merely testing pre-existing ideas, they can make observations
that demand the creation of new ideas and categories that might not emerge in
quantitative designs (Ezzy 2002; Strauss and Corbin 1998). Good qualitative
researchers will adjust their approach in response to data which may contradict
their initial assumptions or theories. Within a positivist paradigm, by insisting
on a strict protocol for the study design and parameters, based on fixed
hypotheses, only those events considered relevant are recorded. Conversely,
qualitative studies may begin to uncover the complexity of social processes
(Strauss and Corbin 1998).
A qualitative, interpretive approach allows acknowledgement of conflict,
ongoing struggle, tension and subjectivity, as well as the situated and co-
produced nature of accounts (Rubin and Rubin 2005). Interpretive research is
about subjectivity and complexity. It seeks not necessarily to count or reduce,
but to represent rich, subjective experience in such a way as to reflect on con-
sistencies and parallels, while retaining the various nuances of the data. As
Ezzy (2002) argues, qualitative research is biased to a degree, but all research
contains a degree of bias and is thus inherently political. To suggest that some-
thing can be biased is by default to suggest that there is an unbiased ‘truth’ that
we could access. The view taken here is that all research is driven by political
interests and theoretical models, and qualitative research is no different.
As suggested earlier, qualitative researchers tend to espouse a constructivist
ontological view of the world. As a result they are focused less on generalizabil-
ity or external validity and more on reliability and internal validation – that is,
the degree to which the data accurately represent attitudes, perceptions and
views of the population being studied. Rather than establishing universal
truths about the world, a qualitative study is about gaining an understanding
of how some differently positioned actors talk about their experiences and the
meanings they associate with particular events, actions and claims. This is a
valuable approach because so often the cost of attempting to generalize is that
we do not see and investigate those aspects of a process that do not fit our pre-
suppositions about a particular phenomenon.
Conducting Health Research
26
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 26

A critique of interpretivist approaches
However, there are a number of problems with interpretivist or constructivist
approaches when applied to health research, some of which are paradigmatic
and some of which relate to methods. As Bryman (2001) notes, first, there is no
real consensus among qualitative researchers on appropriate methods for data
collection and analysis. Second, qualitative researchers do sometimes attempt to
make unjustified generalizations from the accounts of a small number of individ-
uals. It could be argued that there is little value in funding research that cannot
be generalized to the larger population. Furthermore, qualitative researchers
often adopt convenience sampling strategies, resulting in samples that could be
seen to be ‘biased’ through researcher assumptions and respondent self-selection
strategies (Bryman 2001). It has also been argued that informants’ accounts are
not so much uncovered as created by the researcher (Silverman 2001). If the
method adopted is to facilitate an open discussion, for example through unstruc-
tured interviews or focus groups, how can the researcher determine which aspect
of an account is what a person actually thinks, and which part simply reflects
what the researcher prompted the research subject to say?
There are also issues that arise when the constructivist ontological position is
taken in researching health and disease. What can be more ‘real’ than cancer,
pain or diabetes? How can it be argued that these phenomena are socially and
cultural determined rather than purely physiological conditions? Of course, we
come back to the ‘ideal type’mentioned earlier. In practice sociologists of health
and illness will usually take the middle ground, sometimes called the ‘soft con-
structivist position’, and acknowledge that there are both subjective and purely
physical aspects to disease and treatment processes.
In terms of the qualitative arm of the Internet study, the objective was to explore
the experiences of a specific group of men regarding their use of online communi-
ties for social support for prostate cancer. Rather than claiming to map out the expe-
riences of all men with prostate cancer, this study aimed to provide a set of
indications as to the possible experiences of other men and the potential benefits of
the online support groups as sources of information and support. As such, an in-
depth exploratory approach to data collection was taken to document the complex,
subjective experiences of the respondents, instead of merely reflecting on, for
instance, the frequency of Internet usage, existing sources of information and the
type of information and support retrieved. The study focused on understanding the
ways in which the online support groups had impacted on the lives of the respon-
dents, building theory from their accounts rather than imposing it on them. What
emerged was a process by which men with prostate cancer attempt to negotiate cul-
tural constructions of masculinity and prostate disease, along with particular under-
standings about the nature of cyberspace. Such subjective processes were analyzed
by using positivist designs through quantification and abstraction.
Competing Paradigms and Health Research
27
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 27

CONCLUSION
The argument made in this chapter is that health researchers should take
a pragmatic approach to research design and data collection. From a
methodological perspective, we advocate what could be called a ‘horses
for courses’ approach. However, before such an approach can be taken,
researchers must gain a critical understanding of the different kinds of
paradigms that inform health research. They should have an appreciation
of the complex epistemological and ontological differences between qual-
itative and quantitative research methods. Previously quantitative, posi-
tivist research methodologies have been seen by many stakeholders
within the health care community as the only useful way of generating
knowledge. Indeed, it is still the case that research funding is more likely
to be forthcoming for grant applications demonstrating a quantitative
rather than a qualitative methodological approach. However, qualitative
research has grown in importance and indeed legitimacy, and increas-
ingly qualitative approaches are being used to inform health care policy
and practice. Moreover, some questions can only be answered – or at least
will be answered more effectively – by a qualitative design. Ultimately,
therefore, we argue that the appropriateness and usefulness of a particu-
lar paradigm is inextricably tied to the nature of the research question.
This point is tested out in the exercise for the reader below on research
into prostate cancer.
Exercise: Different methodological approaches
to researching the use of the Internet in the
case of men with prostate cancer
Based on what you have read in this chapter, what different sorts of
information are yielded by the data sources set out in Boxes 2.1 and 2.2
based on the work conducted by the authors in Australia in the study of
men with prostate cancer? Which do you think are the most important?
Give reasons for your answer.
Conducting Health Research
28
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 28

Box 2.2 A qualitative in-depth interview study and example of output
Selected interview themes:
Patients’ reasons for decisions over whether or not to use online support groups.
Patients’ perceptions regarding the benefits and limitations of online support
Example of data:
Andrew: ‘One of the things you find is an amazing openness and frankness about these sort of
matters that I’m sure men if they were meeting face-to-face would not talk about … we’re
doing it through this medium [the Internet] and we can be a lot more frank … There’s the
anonymity, there’s the disembodiment … you’re able to project in a way that isn’t having any
comeback on you.’ (Six months’ post-treatment, organ-confined disease, Internet user/online
support, 40—50 years)
David: ‘Some men don’t want to be face-to-face. Maybe they’re frightened of it; maybe they
don’t want to travel the distances. Maybe they’re scared of being ridiculed or something … all
sorts of reasons like that. Maybe they’re a bit anxious about having the problem [prostate
cancer] and not wanting to share it with other people. I think that’s men for you. Some will
find it easier to talk online.’ (Three years’ post-treatment, organ-confined disease, Internet
user/online support, 61—70 years)
Competing Paradigms and Health Research
29
Box 2.1 A quantitative pilot survey design and example of output
Selected interview themes:
1 = strongly disagree 2 = disagree 3 = neutral 4 = agree
5 = strongly agree
1 It would be easier to share my personal experiences
in an anonymous environment like an online 1 2 3 4 5
(Internet) support group.
2 In face-to-face support groups, the threat of
embarrassment stops some men from sharing 1 2 3 4 5
concerns about fears, emotional distress,
symptoms or complications of treatments.
Example of data:
• Of the 50 men surveyed 46 per cent agreed that it would be easier to share experiences
online.
• Approximately 30 per cent of respondents agreed that face-to-face support groups can
deter self-expression.
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 29

RECOMMENDED FURTHER READING
Brown, B., Crawford, P. and Hicks, C. (2003) Evidence-based Research: Dilemmas and
Debates in Health Care. Maidenhead: Open University Press.
This book focuses on charting the philosophical background and debates surrounding
the use of health research methods, including the importance of paradigms.
Daly, J., MacDonald, I. and Willis, E. (eds) (1992) Researching Health Care: Dilemmas,
Designs and Disciplines. London: Routledge.
This is an older, but now classic, reference with relevant contributions by many leading
researchers in the field internationally.
Rice, P. and Ezzy, D. (1999) Qualitative Research Methods: A Health Focus. Melbourne:
Oxford University Press.
This is an excellent introductory textbook on the use of qualitative methods in research-
ing health.
REFERENCES
Altman, D. (2000) ‘Blinding in clinical trials and other studies’, British Medical Journal,
321: 504.
Berger, P. and Luckmann, T. (1967) The Social Construction of Reality: A Treatise in the
Sociology of Knowledge. London: Allen Lane.
Broom, A. (2004) ‘Virtually he@lthy: the impact of Internet use, by Australian men with
prostate cancer, on patient/medical specialist interaction and disease experiences’.
PhD thesis, La Trobe University, Australia.
Broom, A. (2005a) ‘The eMale: prostate cancer, masculinity and online support as a chal-
lenge to medical expertise’, Journal of Sociology, 41(1): 87–104.
Broom, A. (2005b) ‘Virtually he@lthy: a study into the impact of Internet use on disease
experience and the doctor/patient relationship’, Qualitative Health Research, 15(3):
325–45.
Bryman, A. (2001) Social Research Method. Oxford: Oxford University Press.
De Vaus, D.A. (1995) Surveys in Social Research. Crows Nest, NSW: Allen & Unwin.
Ezzy, D. (2002) Qualitative Analysis: Practice and Innovation. Crows Nest, NSW: Allen &
Unwin.
Geertz, C. (1973) The Interpretation of Cultures: Selected Essays. New York: Basic Books.
Giddens, A. (1987) Positivism and Sociology. London: Heinemann.
Glaser, B. and Strauss, A. (1967) The Discovery of Grounded Theory: Strategies for Qualitative
Research. Chicago: Aldine.
Glassner, B. and Moreno, J. (1989) ‘Introduction: quantification and enlightenment’, in
B. Glassner and J. Moreno (eds), The Qualitative-Quantitative Distinction in the Social
Sciences. Boston: Kluwer Academic.
Lofland, J. and Lofland, L. (1984) Analyzing Social Settings: A Guide to Qualitative
Observation and Analysis, 2nd edition. Belmont, CA: Wadsworth.
Conducting Health Research
30
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 30

Rubin, H. and Rubin, I. (2005) Qualitative Interviewing: The Art of Hearing Data, 2nd
edition. London: Sage.
Schulz, F., Chalmers, I., Hayes, R. and Airman, D. (1995) ‘Empirical evidence of bias:
dimensions of methodological quality associated with estimates of treatment effects
in controlled trials’, Journal of the American Medical Association, 273: 408–12.
Silverman, D. (2001) Interpreting Qualitative Data. London: Sage.
Strauss, A. and Corbin, J. (1998) Basics of Qualitative Research, 2nd edition. London: Sage.
Wolfe, A. (1997) ‘The two faces of social science’, in K. Erickson (ed.), Sociological Visions.
Oxford: Rowman and Littlefield.
Competing Paradigms and Health Research
31
Saks-3519-ch 02.qxd 1/12/2007 7:52 PM Page 31