FOUNDATIONS OF QUALITATIVE

METHODS
KEY POINTS IN THIS CHAPTER

T Qualitative research uses language as its raw material.

T It aims to study people’s thoughts, experiences, feelings,
or use of language in depth and detail.

T The main advantage of qualitative methods is that they

allow a rich description.

T There are two main traditions, phenomenology and

constructionism, although there is considerable
diversity within, and overlap between, these tradi- tions.

T Phenomenologists
and experiences.
attempt to understand
the person’s perceptions

T Constructionists focus on how language is used in social

interactions, and how discourse is affected by
culture, history, and social structure.

T It is possible to specify criteria for evaluating

qualitative research studies.

The raw material for qualitative research is ordinary language, as

opposed to the numbers that are the raw material for quantitative
research. The language may be obtained in many ways. It may be the
participant’s own descriptions of him or herself, recorded during a
qualitative interview. Or it could be words transcribed from a
conversation, such as that between a client and a therapist during a
session of therapy. Or it could be something printed, such as a newspaper
article or the operational policy statement of a hospital’s management
committee. It could also take the form of the researcher’s field notes of
the participants’ behavior, as written down after a qualitative observation
session. Whatever source it may come from, linguistic data can give the
researcher rich, deep, and complex information, sometimes referred to as
‘‘thick description’’ (Geertz, 1973).

1
 FOUNDATIONS OF QUALITATIVE METHODS
73
This can be used to understand people’s feelings, thoughts, ways of
under- standing the world, or ways of communicating with others.
A simplified illustration of the difference between the quantitative and
the qualitative approach is shown in the differing responses to the
question ‘‘How are you feeling today?’’ A quantitatively oriented
researcher might ask the participant to respond on a seven-point scale,
ranging from 1 = ‘‘Very unhappy’’ to 7 = ‘‘Very happy,’’ and receive an
answer of 5, signifying ‘‘Somewhat happy.’’ A qualitative researcher
might ask the same person the same question, ‘‘How are you feeling
today?’’, but request an open-ended answer, which could run something
like ‘‘Not too bad, although my knee is hurting me a little, and I’ve just
had an argument with my boyfriend. On the other hand, I think I might
be up for promotion at work, so I’m excited about that.’’ In other words,
the quantitative approach yields data which are relatively simple to
process, but are limited in depth and hide ambiguities; the qualitative
approach yields a potentially large quantity of rich, complex data
which may be difficult and time consuming to analyze.
However, the difference between quantitative and qualitative
approaches to research is about more than the difference between
numbers and words; it is also about epistemology, the theory of what
knowledge consists of (see Chapter 2). As we noted in Chapter 4,
quantitative research is largely based on the philosophy of positivism.
Qualitative researchers usually reject positivism, often quite
vehemently, instead preferring non-realist epistemological positions
based on developing understanding rather than on testing hypotheses
(Bryman, 1988). We will describe these positions more fully below.
It is worth noting in passing that there is one potential source of
confusion over the meaning of the word ‘‘qualitative,’’ as it also has a
second distinct meaning in research terminology. In quantitative
research, the term ‘‘qualitative data’’ is used to refer to nominal scale
data, to distinguish it from ordinal or interval scale data (see Chapter 4
on psychometric theory). Thus, census categories measuring ethnic
background (white European, black African, etc.) may be referred to as
a ‘‘qualitative’’ variable (because they have no ordering property), even
though the data are analyzed by quantitative methods. However, in this
book we will reserve the term ‘‘qualitative’’ for data that are collected
by open-ended questions or by observations that yield verbal
descriptions. Simple yes–no responses or nominal categories will be
considered as a form of quantitative data, since they are narrowly
delimited. The qualitative–quantitative distinction, as we are using it,
boils down to whether the data are collected and analyzed as words or
numbers (including counts, proportions, multiple choice, and yes–no
responses).

Advantages
The main advantages of using qualitative methods are:
● They avoid the simplifications imposed by quantification, since
some things cannot be easily expressed numerically. That is, they
enable more complex
73
 FOUNDATIONS OF QUALITATIVE METHODS
74
aspects of experience to be studied and impose fewer restrictions on the
data or the underlying theoretical models than quantitative approaches.
● They allow the researcher to address research questions that do not
easily lend themselves to quantification, such as the nature of
individual experiences of a psychological condition (e.g., eating
disorders) or event (e.g., being a victim of crime).
● They enable the individual to be studied in depth and detail.
● The raw data are usually vivid and easy to grasp: good qualitative
research reports make the participants come alive for the reader. In
general, the reports of qualitative studies are often more readable
than those of quantitative studies (except that some qualitative
researchers, especially those with postmodernist or existential-
phenomenological leanings, tend to write in an impenetrable jargon
all of their own).
● Qualitative methods are good for hypothesis generation, and for
exploratory, discovery-oriented research. They permit a more flexible
approach, allowing the researcher to modify his or her protocol in
mid-stream. The data collection is not constrained by pre-existing
hypotheses.
● Qualitative self-report methods usually give more freedom to the
participant than structured quantitative methods. For example, open-
ended questions give interviewees a chance to respond in their own
words and in their own way.
● Since the data collection procedures are less constrained, the
researchers may end up in the interesting position of finding things
that they were not originally looking for or expecting.

Historical Background
Qualitative methods can be traced back to the ancient Greek historians.
For example, Herodotus, who is often called the father of history, traveled
widely in the ancient world and recounted in his Histories the stories he
had heard from the people he met. His successors down the ages
recorded their observations of people that they encountered in their
travels. These kinds of observations eventually became formalized in the
discipline of anthropology.
In their modern form, qualitative methods were first used in ethnographic
fieldwork. In the early decades of the 20th century, the founders of
cultural anthropology, such as Malinowski and Boas, conducted
ethnographic observa- tions on cultural groups that were remote from
their own: Malinowski in the Trobriand Islands in Papua New Guinea and
Boas with the Kwakiutl tribe in the Pacific Northwest Coast of North
America. They spent many months living with and observing the cultures
they were studying. In the 1920s and 1930s, sociologists adapted these
methods to study subcultures within their own society. For example, the
‘‘Chicago school’’ of sociology tended to focus on people at the fringes
of society, such as criminals and youth gangs. A classic

74
 FOUNDATIONS OF QUALITATIVE METHODS
75
example of this genre is Whyte’s (1943) Street corner society, which
was based on fieldwork with an Italian-American youth gang in Boston,
Massachusetts. Ethnographic methods started out being used to study
the ‘‘weird and wonderful’’ (from a Eurocentric viewpoint), e.g., Pacific
Island tribal cultures, and have been brought progressively closer to
bear on the investigators’ own culture, culminating in such
contemporary specialties as medical anthropology, which use
anthropological methods to study health and illness in our own culture
(Helman, 1990).
Some ethnographic work is located on the rather fuzzy boundary between
social science and journalism. A good example is Blythe’s (1979) The view
in winter, a moving account of people describing how they experience
being old. The distinction is that journalism seeks to report accurately
and produce an engaging story, whereas social science brings a body of
theory to bear on the subject matter, or seeks to develop theory
from the data, and it articulates its assumptions and procedures in
order to enable replication.
In clinical research, qualitative methods were first used in case histories
(see Chapter 9), for instance, Breuer and Freud’s (1895/1955) first
cases, which began the psychoanalytic tradition, and Watson and
Rayner’s (1920) study of ‘‘Little Albert,’’ which helped establish the
behavioral tradition. There is also a tradition of participant observation
methods in mental health research, though they are more often
conducted by sociologists than by psychologists. Classic examples of
participant observation studies are Goffman’s (1961) Asylums and
Rosenhan’s (1973) ‘‘Sane in insane places’’ study.
The two main qualitative data collection methods currently used in
clinical psychology research are in-depth interviewing (see Chapter 6)
and qualitative observation (see Chapter 7). Common data analytic
methods are grounded theory, various types of phenomenological
analysis, and discourse analysis (general principles of qualitative data
analysis are covered in Chapter 12).

Traditions of qualitative research

Qualitative research is unfortunately not immune to the usual kinds of
factions that bedevil most academic enterprises. Although qualitative
researchers are united by their wish to move beyond the perceived
limitations of the quantitative approach, they dispute the underlying
epistemology and philosophy of science that characterizes their
endeavors. Broadly speaking, we will consider qualitative approaches
under two headings: the phenomenological tradition and the social
constructionist tradition. However, other writers would find this
dichotomy simplistic. For example, Willig (2001) depicts the approaches
as being arranged on a continuum, ranging from realist to relativist.
Much like in qualitative analysis generally, there is no one best way to
organize the material, just various possible constructions.
It is also worth noting that, whichever way they are classified,
there is considerable diversity within, and overlap between, the
various qualitative
75
 FOUNDATIONS OF QUALITATIVE METHODS
76
approaches. Also, some approaches have different variants under the
same label: for example, there is more than one version of grounded
theory and of discourse analysis.
Phenomenologists attempt to understand the person’s thoughts, feelings,
perceptions, and interpretations of the world. Social constructionists, and
the postmodernists with whom they are often allied, look at language as a
social product in itself, questioning many of the familiar concepts, such as
reality, truth or the person, that are taken for granted in other branches
of the discipline. The next two sections of this chapter describe each of
these traditions in more detail.

PHENOMENOLOGICAL APPROACHES

Central tenets of phenomenology:

● The objects of study are people’s experiences, ‘‘life worlds,’’
and under- lying assumptions.
● Understanding is the true end of science.
● Multiple valid perspectives are possible.

The word ‘‘phenomenology’’ is itself a bit of a mouthful, and some of the

underlying theory is couched in off-putting existentialist jargon. However,
phenomenology is simply the study of phenomena (singular:
phenomenon), and ‘‘phenomenon’’ is simply a fancy word for perception
(that is, what appears to us). In any case, the essence of phenomenology
is relatively simple: it is the systematic study of people’s experiences and
ways of viewing the world.
Sometimes the approach is known as ‘‘phenomenological-hermeneutic,’’
to stress its interpretive aspect. (‘‘Hermeneutic’’ is a fancy word for
interpretive, and can be used interchangeably with it.) However, there
is a potential source of confusion here, as there is a brand of
phenomenological research known as ‘‘Interpretative Phenomenological
Analysis’’ (Smith et al., 1999), which we will discuss below. (Interpretative
is given here in its British spelling, because the approach originated in the
UK.) Here, we will use the term ‘‘phenomen- ological’’ in its general
sense, to also encompass phenomenological-hermeneutic methods.

Philosophical Background
Phenomenological approaches in psychology derive from the
phenomenological movement in philosophy, which developed in the late-
19th and early-20th centuries. It in turn is descended from the rationalist,
idealist philosophical tradition of Plato and Kant. Husserl was its
founder; Brentano, Heidegger,

76
 PHENOMENOLOGICAL APPROACHES
77
Merleau-Ponty, and Sartre were key figures in its development
(Jennings, 1986; Spinelli, 1989). Their ideas were introduced into
psychology by Giorgi, Laing, May, and others (e.g., Giorgi, 1975; Laing,
1959; May et al., 1958), and were a major influence on the client-
centered, humanistic, and existential approaches to psychological
therapy.

Assumptions
We can distinguish four central assumptions of phenomenology. First,
percep- tion is regarded as the primary psychological activity, since our
perceptions give rise to what we do, think, and feel. Because of this,
perceived meaning is more important than objective reality, facts, or
events.
Second, understanding is regarded as being the true end of science (in
contrast, for example, to the aim of prediction and control that more
traditional hypothetico-deductive approaches espouse). The goal is to
produce explana- tions of the person’s experiences and actions in terms
of intentions, purposes, and meanings.
A third key assumption is that of multiple perspectives, also known as
‘‘epistemological pluralism.’’ Each person’s perspective has its own
validity (i.e., it is how they see things); therefore, multiple, differing
perspectives are equally valid and of interest for study. These multiple
perspectives constitute different life-worlds (in German, umwelten). For
example, the same aging oak tree is radically different when perceived
by the forester, the lost child, the fox, or the wood beetle. These life-
worlds are the object of study for the phenomenologist (Pollio, 1982).
Fourth, individuals’ perceptions of their life-worlds are based on implicit
assumptions or presuppositions, which phenomenologists also try to
understand. That is, what we perceive is built on multiple assumptions
about ourselves, others, and the world. These assumptions are the
taken-for-granted, unques- tioned context for our actions and
perceptions. For example, if an acquaintance greets you with ‘‘How are
you?’’, you are not usually expected to give an accurate or detailed
answer; in fact, to anyone but a close friend, it would seem quite odd to
do so—the underlying assumption is that we respond with a brief,
positive answer. Although we accept these underlying assumptions, we
are not generally aware of them and do not question them. In other
words, they are believed to be ‘‘known to all’’ and part of what
‘‘everybody knows that everybody knows’’ (Garfinkel, 1967).

A key set of underlying assumptions is known as the ‘‘natural attitude’’

or ‘‘mundane reason’’ (Pollner, 1987). This is made up of the
unquestioned belief that things are what they appear to be, and that all
sane persons share the same world. In fact, in everyday life, it is
considered strange or deviant to talk about many of these
presuppositions, so that their very obviousness at the same time hides
them or prevents them from being noticed.

77
 PHENOMENOLOGICAL APPROACHES
78
Table 5.1 Types of phenomenological research

Type of research Sample references

Client-centered research Rogers, 1985; Shlien, 1970
Consensual qualitative research Hill et al., 1997
Ethnography Fetterman, 1989
Empirical phenomenology (Duquesne school) Giorgi, 1975; Wertz,
1983 Field and participant observation Taylor & Bogdan,
1998
Grounded theory Henwood & Pidgeon, 1992;
Strauss & Corbin, 1998
Hermeneutic-interpretive research Packer & Addison, 1989; Terwee,
1990 Interpretative phenomenological analysis Smith et al., 1999
Life history or biographical research Denzin, 1989; Taylor & Bogdan,
1998 Protocol analysis Ericsson & Simon, 1993

Types of Phenomenological Research

The explosion of interest in phenomenological research in the past 20
years has given rise to a bewildering number of variations, the more
important of which are listed in Table 5.1. However, it is worth bearing in
mind that, in practice, they have much more in common than separates
them. We will briefly describe four approaches that are commonly
encountered by clinical psychologists: empirical phenomenology,
grounded theory, hermeneutic approaches, and interpretative
phenomenological analysis.

Empirical Phenomenology
The oldest systematic approach to phenomenological psychology is
associated with Duquesne University (Pittsburgh, USA), and published in
the Journal of Phenomenological Psychology (see also Duquesne studies
of phenomenological psychology, 1971 to 1983). Giorgi, Wertz, and
Fischer are three of the better known proponents. The approach stresses
in-depth analysis, often at first of single cases, aiming to describe the
main defining features of an experience (e.g., that of being criminally
victimized), and the different variations that the experience may have in
the population (analogous to the statistical ideas of mean and standard
deviation). Carrying out research in this tradition involves the practice of
the phenomenological method, a rigorous procedure that is based on two
key processes: bracketing and describing.
Bracketing is an attempt to set aside one’s assumptions and
expectations, as far as is humanly possible. However, because one’s
underlying assumptions are often hidden, it requires a special act of
reflection to identify them. This act has been described in several
different ways. The most common is ‘‘bracketing the natural attitude’’ (or
‘‘bracketing’’ for short). It involves a process of stepping back from the
phenomenon in order to see it as if from the outside, as if we were the
proverbial observer from Mars. Bracketing involves a special kind of
turning

78
 PHENOMENOLOGICAL APPROACHES
79
away from the natural attitude, in which the researcher does not accept
a description as a statement about the world, but simply as a statement
about an experience of the world.
In the clinical context, bracketing is one aspect of the process of
empathy in such exploratory psychotherapies as client-centered and
experiential therapy. When a client says that she is ‘‘trapped’’ in a
situation, the therapist is not interested in determining whether this is
factually the case; what is important is that the client feels trapped
(Rogers, 1975). In contrast, beginning therapists generally prefer to
stay ‘‘within the natural attitude’’ by trying to talk the client out of such
presumed irrational beliefs, often questioning the facts of the situation.
One important component of empathy is letting go of one’s own
presuppositions, in order to understand what the client is trying to say.
A similar idea is found in the ideal therapist state of ‘‘evenly hovering,
free-floating attention’’ referred to in the psychoanalytic literature (e.g.,
Greenson, 1967).
A naive approach to bracketing might be to mentally steel oneself and
promise to give up one’s biases. However, a more fruitful alternative is
to begin by carefully investigating one’s assumptions. At the beginning
of a study, the researcher can conduct a thought experiment of carrying
out the study in imagination in order to identify expectations of what it
might find. This thought experiment might also be repeated at the end
of the study in order to identify additional expectations that only
became clear in the course of the study. These expectations take the
place of hypotheses in traditional research, but they are not the same.
In phenomenological research, expectations are not given a place of
honor at the end of the introduction, instead, they are figuratively
locked in a drawer until the study is over. Phenomenological research is
perhaps most exciting when it uncovers understandings that are
unexpected or even startling.
The second step in the empirical phenomenological method is
describing. Several principles are involved (see, for example, Spinelli,
1989). First, good descriptions focus on concrete or specific
impressions, as opposed to the abstract or general. Second, they avoid
evaluative terms such as ‘‘good or bad’’ and their many synonyms and
euphemisms (e.g., ‘‘ineffective,’’ ‘‘helpful’’), except where these are
part of the experience itself. Third, they tend to avoid explanations,
particularly early in the research. The task is to discover meaning,
not invent it. This means avoiding ‘‘why’’ questions or anything that
encourages the informant to speculate on causes or reasons: such
questions encourage intellectualization and interfere with the slow,
careful process of attending to concrete experience.

Grounded Theory
Grounded theory is probably the most common form of qualitative
analysis used today. It was developed by two North American
sociologists, Glaser and Strauss, in their 1967 book, The discovery of
grounded theory: Strategies for qualitative research. As the title
suggests, Glaser and Strauss were attempting to articulate how
qualitative data could be used not just to provide rich descriptions,
79
 PHENOMENOLOGICAL APPROACHES
but also to 80

80
 PHENOMENOLOGICAL APPROACHES
81
generate theory. Grounded theory can be used with a range of qualitative
material, such as semi-structured interviews, focus groups, participant
observation, and diaries.
The term ‘‘grounded theory’’ is potentially confusing, as it refers both to a
method—a set of systematic procedures for analyzing data—and also to
the outcome or product of the analysis, which is theory ‘‘grounded’’ in the
data. The basic process involves identifying categories at a low level of
abstraction and then building up to more abstract theoretical concepts.
The end point is often one or more core categories, which capture the
essence of the phenomenon (see Chapter 12). This process of analysis
occurs concurrently with the process of data collection, and the
developing theory guides the sampling strategy (‘‘theoretical sampling’’:
see Chapter 10).
The original Glaser and Strauss (1967) volume was more theoretical and
polemical rather than practical; it was aimed at challenging the prevailing
quantitative paradigm in American sociology. The practical implications
for researchers, i.e., the steps in actually carrying out a grounded theory
study, are developed in Glaser (1978) and Strauss and Corbin (1998).
The grounded theory method was taken up by psychologists in the 1980s
and 1990s. Articles by Rennie et al. (1988) and by Henwood and Pidgeon
(1992) were aimed at introducing the grounded theory approach to an
audience of psychologists. Rennie and Brewer’s (1987) study entitled
‘‘A grounded theory of thesis blocking’’ (i.e., writer’s block among
research students) may well be of personal interest to some readers of
this text! As more psychologists have taken up the invitation of Rennie et
al., and of Henwood and Pidgeon, grounded theory has become a popular
approach to qualitative research.
One example of the method in clinical psychology is Bolger’s (1999) study
of the phenomenon of emotional pain. The participants were women in a
therapy group for adult children of alcoholics; they were interviewed on
several occasions following group therapy sessions in which they had
explored painful life experiences. The interviews focused on how pain was
experienced and what was significant in that experience for them. The
core category that emerged from the analysis was labeled the ‘‘broken
self,’’ characterized by four sub-categories of woundedness,
disconnection, loss of self, and awareness of self.
Another, well-known, example of a grounded theory study, in a more
popularized book format, is Charmaz’s (1991) analysis of the experience
of living with chronic illness (see box).

Hermeneutic Approaches
A somewhat more flexible approach to qualitative research is represented
by approaches that bill themselves as hermeneutic (e.g., Packer &
Addison, 1989). Researchers who describe themselves in this way find
grounded theory and empirical phenomenology too restrictive and use a
wider range of methods. They argue that it is important to go
beyond the surface meaning of research

81
 SOCIAL CONSTRUCTIONIST APPROACHES
81

Grounded theory example: Charmaz (1991)

The sociologist Kathy Charmaz conducted in-depth qualitative
interviews with people who had a chronic illness. The results,
written up in her book Good days, bad days (1991), give
compelling accounts of the impact of chronic illness on people’s
lives. In accordance with the grounded theory approach, she also
used the data to construct a theory of how the person’s experience
of time changes, and how this impacts on their sense of self.

protocols, in order to identify the implicit or even unconscious meanings

embedded in texts.

A recent example is a study by Walsh et al. (1999), which used a

hermeneutic approach to explore ‘‘good moments’’ within a videotaped
psychotherapy session. They identified the differing values of
professionals and students at various stages of training about what
constituted good psychotherapy.

Interpretative Phenomenological Analysis

A final example of the phenomenological approach, which often appeals
to newcomers to qualitative research, is Interpretative
Phenomenological Analysis (Smith et al., 1999). This is an accessible
and clearly presented procedure for conducting qualitative research, set
out with a minimum of theoretical baggage. Interpretative
Phenomenological Analysis is basically a systematic and practical
approach to analyzing phenomenological data. It articulates the steps
involved in conducting an investigation: e.g., how to generate
meaningful lower order and higher order categories from the data.
Smith et al.’s (1999) chapter sets out the basis of their method,
illustrating its steps using examples of data drawn from clinical and
health psychology. Hill and colleagues (1997) have also described a
comparable highly structured, easy-to-follow method of qualitative
analysis, which they refer to as Consensual Qualitative Research.

SOCIAL CONSTRUCTIONIST APPROACHES

Social constructionists (constructionists, for short) are interested in how

language is used to order and manage the world. In contrast to
phenomenologists, constructionists do not see language as necessarily
reflecting the individual’s underlying thoughts and feelings; rather they
are interested in how people use language to structure things, or to get
things done. For example, constructionist researchers have examined
psychiatric diagnostic systems from the point of view of how diagnosis
may be used by mental health professionals to impose a parti- cular
view of the world on people’s experience (e.g. Georgaca, 2000; Harper,
1994).

81
 SOCIAL CONSTRUCTIONIST APPROACHES
82

Central features of social constructionism:

● Part of the postmodernist and poststructuralist movements.
● Non-realist.
● ‘‘Radical pluralism.’’
● Often focuses on language in text or speech.
● Stresses the indeterminacy (ambiguity) of language and meaning.
● Interested in language as social action.
● Does not assume that language reflects cognition.
● Emphasizes the reflexivity (circular nature) of psychological
theory.

Background to Social Constructionism

The basis of the constructionist approaches is an opposition to the realist
approach to social science, in particular as articulated by adherents of
positivism (see Chapters 2 and 4). Social constructionists reject, or at
least dispense with, the assumption of an underlying, independent reality
(Gergen, 1985; Madill et al., 2000; Willig, 2001). They may speak in terms
of multiple realities—that each individual constructs their own personal
reality. This rejection of realism is to some extent shared by the
phenomenologists, although the constructionist position tends to be more
forcefully expressed and may be more thorough going: phenomenologists
do not explicitly reject realism, they just accept that different people may
have different concepts of what reality is.
However, as we have mentioned above, there is a diversity of views
within many qualitative approaches, and social constructionism is no
exception. There is a radical version of constructionism that completely
rejects any notion of reality. Thus Guba and Lincoln (1989) write that their
constructivist paradigm ‘‘denies the existence of an objective reality,
asserting instead that realities are social constructions of the mind, and
that there exist as many such constructions as there are individuals
(although clearly many constructions will be shared)’’ (p. 43). Such
radical constructionists also do not wish to ‘‘privilege’’ one world view
over any other. Thus they see traditional scientific methods as one
possible way of understanding the world, but would not necessarily
regard them as being any more valid than other systems of belief, such
as shamanism or astrology. However, they would accept that scientists’
own criteria for validity are meaningful within the scientists’ own domain
of discourse.
Social constructionists pay close attention to language, spoken and
written. However, they analyze language in a different way to researchers
working within

82
 SOCIAL CONSTRUCTIONIST APPROACHES
83
traditional realist approaches, who are usually concerned with whether
statements expressed in language are true or not. Thus if a psychiatrist
says, ‘‘This patient is paranoid’’, a realist approach would be to see
whether the statement was accurate—i.e., is the patient paranoid or
not? In constructionism the focus shifts towards looking at how people
construct their arguments and what work their constructions do: for
example, what rhetorical devices does the psychiatrist use to convince
us of the validity of her position, that the patient is indeed paranoid
(see Harper, 1994)?
The type of focus on language also distinguishes the constructionist
approach from the phenomenological one. Phenomenologists and social
constructionists may share the assumption that objective reality is not
of primary concern. Furthermore, phenomenologists also use spoken
language, in the form of qualitative interviews, as their primary medium
of research. However, the phenomenologist is using the language to
understand the thoughts and feelings of the participants—to try and
understand their inner world. For radical constructionists, this act of
understanding, too, is a social construction, leaving us with only the
process of construction to study, especially as this plays out in

Social Constructionism versus Constructivism—What’s the difference?

The terms ‘‘constructionism’’ and ‘‘constructivism’’ are often used
interchangeably, but they are not identical. Constructionism usually
refers to the view that the concepts we use—e.g., madness or
masculinity—are socially determined, that is, they don’t refer to an
independent reality but may vary across cultures or over time.
Constructivism is a more psychological concept; it refers to the
process by which individuals arrive at the constructs they use. One
important example of constructivist thinking is Kelly’s (1955)
personal construct theory, which looks at the central constructs each
individual uses in order to make sense of their world.
Another example of constructivism can be found in contemporary
cognitive therapy. Historically, cognitive therapists have viewed the
external world as less important than how clients make sense of
the world. The ancient Stoic philosopher Epictetus is often quoted:
‘‘Men are not disturbed by things, but by the view which they take of
them.’’ However, as Neimeyer (1993) points out, most cognitive
therapists follow in the realist tradition, in that thoughts are viewed
as rational (and therefore healthy) if they correspond to reality. This
realist stance is similar to that adopted by traditional psychiatry, in
which the prime criterion for psychosis is loss of contact with reality.
However, some contemporary cognitive-behavioral therapists have
begun to move toward a constructivist position, which is, as we have
noted, more internally consistent with its philosophical roots.

83
 SOCIAL CONSTRUCTIONIST APPROACHES
84
language use. As Reicher (2000) puts it ‘‘. . . language is a form of social
action which we use in order to create our social world. The focus is on
how apparent descriptions serve to manage our social relations.
Psychological categories such as beliefs, desires, and even experience,
are only of interest in so far as participants themselves put them to use in
their discourse’’ (pp. 3–4).
The other feature of many constructionist theories is that they stress the
reflexivity (from ‘‘reflection,’’ a function of mirrors) of psychological
theorizing. By this is meant that psychologists are doing the theorizing,
but the psychologists themselves, as human beings, are also the object of
the theory. Psychological research is thus a circular process, which is
what the reflexivity metaphor is attempting to capture. Some
constructionists maintain that the existence of reflexivity fatally
undermines any claims of psychology to be an objective science.

Postmodernism
Social constructionism is closely aligned with the body of thought known
as postmodernism, or sometimes poststructuralism (the two terms
overlap, but are not synonymous). This can be rather difficult territory.
One major difficulty is that it is often hard to pin down exactly what many
of the authors writing in this tradition are actually saying, as their prose
style is often opaque, and the ratio of useful ideas to verbiage can seem
frustratingly low. However, postmodern thought is currently fashionable
and much discussed within several fields of study, so it is important to
come to grips with it.
A second difficulty, however, is that the term ‘‘postmodernism’’ itself is
hard to define. It refers to a rather loose collection of ideas that have
found expression in a number of different fields, e.g., literary theory,
sociology, and architecture. The key figures are all French: the literary
theorist Derrida, the historian Foucault, the psychoanalyst Lacan. Some of
its central themes are:
● A rejection of ‘‘grand theories’’ that provide overarching
explanations, such as psychoanalysis or Marxism; instead micro- or
composite theories are favored. This is coupled with a questioning of
the personal and social interests that lie behind scientific theories,
particularly where those theories seem to serve the interests of those
in power (Prilleltensky, 1997).
● An intellectual playfulness, that borrows from many different
traditions within the same piece of work. For example, postmodernist
architecture often ‘‘quotes’’ from earlier traditions, inserting a Gothic
turret here, a Georgian window there. This is exemplified in the
image of the qualitative researcher as ‘‘bricoleur,’’ a sort of handy-
person who uses whatever is at hand to construct things that are
useful but not elegant (Levi-Strauss, 1958/1963; McLeod, 2001).
● A focus on language. Lyotard (1979) borrows Wittgenstein’s phrase,
‘‘language games,’’ to capture both the aspect of playfulness and the
idea that language is governed by rules.
● The indeterminacy (ambiguity) of language. Post-modernists stress
that all communication carries multiple meanings, so that
84
 SOCIAL CONSTRUCTIONIST APPROACHES
understanding is always 85

85
 SOCIAL CONSTRUCTIONIST APPROACHES
86
an act of interpretation, and what one reader makes of a text may differ
from another reader’s understanding. Eagleton (1983) gives the
example of a sign seen in the London Underground: ‘‘Dogs must be
carried on escalators.’’ Does this mean one cannot step on the
escalator unless one is carrying a dog?
Constructionism is not identical to postmodernism, and it was first
articulated before postmodernism became popular (e.g., Berger &
Luckmann, 1966). However, the postmodernist viewpoint has been
adopted by many social constructionists (e.g., Gergen, 1994, 2001), and
it provides a useful framework in which to understand their ideas.

Critiques of Postmodernism
The postmodernist position (and also the extreme versions of
constructionism) have been fiercely criticized, in a debate which has
generated much heat, and perhaps a little light. The main lines of
argument are:
● The lack of interest in people’s mental states is open to the same
criticisms that were earlier leveled at the methodological
behaviorists, who adopted an identical stance for completely
different reasons (i.e., to become, as they saw it, more scientific). A
psychology that sidesteps the role of inner experiencing is severely
limited, and ultimately presents an alienating view of the person.
● Likewise, the non-realist emphasis can make research into an ivory
tower exercise. Reality is important, especially unpleasant reality.
Studying rape, child abuse, racism, or genocide purely from a
discursive viewpoint can easily seem to diminish their importance or
even appear to deny their existence or the need to prevent them.
● The underlying model of the person is that of a fragmented,
unintegrated self; it is not the model that many psychological
therapists, who are trying to help their clients feel more whole, would
endorse. Likewise, the person can be viewed as a manipulator of
language, whose goal is to manage the impression they make or to
get others to act in a certain way. This is not an image of human
beings that can support the enterprise of helping people lead more
fulfilling, meaningful lives.
● The language that postmodernists employ is often riddled with
impenetrable jargon, which seems designed to convey an impression
of erudition and profundity. Sokal and Bricmont (1999) exposed the
ridiculousness of much postmodernist writing, especially its use of
scientific metaphors, in their critique, Intellectual impostures.

Conclusion
In summary, the strong points of the constructionist position are that
they remind us to look closely and critically at how language is used to
construct reality and to accomplish practical purposes. What ‘‘position’’
is the speaker or writer trying to adopt, how is their language being
used to bolster this position, and what is

86
 SOCIAL CONSTRUCTIONIST APPROACHES
87
Table 5.2 Main types of social constructionist

approaches Type of research Sample

references
Conversation analysis Sacks (1995); Schegloff (1999)
Critical approaches Lather (1991); Reason and Rowan (1981)
Deconstructionism Slife and Williams (1995)
Discourse analysis Potter and Wetherell (1987)
Radical feminist research Belenky et al. (1986)
Social representations Farr and Moscovici (1984)

finally achieved by this? It also stresses the theory-dependent nature of

scientific observation: a view that the constructionists share with Popper
(see Chapter 2), who reached this position from a completely different
philosophical standpoint. Finally, it stresses the social nature of
psychological concepts. Instead of treating a concept such as ‘‘racism’’ or
‘‘mental illness’’ as an individual trait, constructionists urge us to look at
the term in its wider social and political context.

Types of Social Constructionist Research

Like the phenomenological branch of qualitative research, several
different strands of constructionist research can be identified (see Table
5.2). The most popular with psychologists is discourse analysis; other
approaches include conversation analysis, social representations theory,
critical and feminist approaches, and deconstructionism.

Discourse Analysis

Example of discourse analysis:

Madill and Barkham (1997) examined the transcripts of a single
case of time-limited psychodynamic therapy. They showed how,
during the course of therapy, the client took on three different
subject positions—which they labeled as the dutiful daughter, the
bad mother, and the damaged child— and the discourses that she
drew upon which exemplify each of these. For instance, they argued
that the dutiful daughter position ‘‘draws on 18th and 19th century
discourses of female subjectivity. During this period, subject
positions were provided for women based primarily upon their
domesticity . . .’’ (p. 242). Thus they were able to analyze the
client’s talk within the context of its historical and social
antecedents.

There are many kinds of discourse analysis (see van Dijk, 1997a,
1997b); it is an interdisciplinary field spanning psychology, sociology,
communication,
87
 SOCIAL CONSTRUCTIONIST APPROACHES
88
linguistics, and literature. Within psychology, the most popular
approach is that articulated by Potter and Wetherell (1987). In general,
discourse analysis involves rigorously examining texts, in order to
analyze the repertoires of discourse that a speaker is drawing upon, and
the kinds of ‘‘subject positions’’ that the speaker is adopting (see box
for an example).

Conversation Analysis
Although sometimes grouped with discourse analysis, conversation
analysis has its own tradition and its own particular methods. An
outgrowth of the work of sociologists Garfinkel and Goffman, it was
developed by Harvey Sacks (1995) as a rigorous method for identifying
the common conversational sequences and strategies used by people to
accomplish practical purposes. Conversation analysis attempts to
study how speakers perceive each other’s utterances, based on how
they respond to each other. Although it attempts to develop general
models of the strategies people use to accomplish practical work in
conversation, it emphasizes the ad hoc, contextually embedded nature
of ‘‘talk- in-interaction’’ (Schegloff, 1999). Over the past 25 years,
conversation analysis has built up a large repertoire of provisional
understandings of everyday and professional speech, including many
investigations of psychotherapy (see Mondada, 1998, for a review).
One interesting clinical application of conversation analysis is Rapley
and Antaki’s (1996) British study of how psychologists administered a
standardized quality-of-life instrument to people with learning
disabilities (mental retardation in US terminology). They demonstrated
how the interviewers’ questioning strategies and reformulations shaped
the answers that respondents produced. Their analysis challenged the
traditional notion of acquiescence in this population (see Chapter 6) and
suggested a more complex picture of the conversational strategies used
by both interviewers and respondents.

Social Representations Theory

Example of research based on social representations theory:

Joffe (1996) examined how HIV/AIDS was discussed within various
cultural groups. She adopted an interesting sampling strategy, inter-
viewing black and white, heterosexual and homosexual participants
in both Britain and South Africa. Using ideas drawn from
psychoanalytic theory, she found that people in the various groups
tended to represent HIV/AIDS in a way that enabled them to blame
the members of an ‘‘outgroup’’ for the problem.

Social representations theory is influential within social psychology, but

has not yet been widely taken up outside that sub-field. Originally
articulated by

88
 SOCIAL CONSTRUCTIONIST APPROACHES
89
Moscovici and promulgated by Farr (e.g., Farr & Moscovici, 1984), social
representations research involves the examination of how new ideas,
especially scientific ideas, are perceived and assimilated within the
culture at large. For instance, social representations researchers might
examine the metaphors that people use to understand psychological
disorders and their treatment.

Critical Approaches
Critical researchers attempt not only to understand but also to
emancipate their research participants; they try to minimize the
distinction between researcher and ‘‘subject’’ in order to create research
in which participant and researcher are co-researchers who interact as
equals. They argue that the act of studying another human being
establishes an alienated relationship between them: the researcher treats
the ‘‘subject’’ like an object.
This viewpoint is forcefully articulated by Reason and Rowan (1981). Their
approach is sometimes described as the human inquiry approach and
sometimes as new paradigm research (they use both labels in the title of
their book). As McLeod (2001) notes, human inquiry research emphasizes
disciplined subjec- tivity, respect for the whole person, linking research to
action, and creative forms of presenting the results of research (e.g., in
the form of art or poetry). An example is an investigation that Wilshaw
(1996) carried out in collaboration with a support group for individuals
who had been sexually abused as children. In this study, he acted as a co-
participant in the support group and as a research consultant; members
of the group selected the research questions, methods, and means of
presenting the results (which consisted in part of group members’
paintings, depicting the impact that their abuse had had on them).
Other critical research approaches include neo-Marxist approaches, as
well as participatory action research and Freirian research (Lather, 1991).

Radical Feminist Approaches

There are several different strands of thinking within feminist psychology
(Campbell & Wasco, 2000). Some feminists conduct their research largely
within traditional quantitative approaches, others use more
phenomenological methods, and some (e.g., Belenky et al., 1986;
Carlson, 1972; Riger, 1992; Wilkinson, 1986) make a similar argument to
the critical researchers. These latter see traditional paradigms, where the
researcher is in charge of the relationship, as replicating patriarchal
power relationships. Their critique is not only aimed at quantification and
experimental manipulation, but also at more traditional forms of
qualitative interviewing (Oakley, 1981).
Radical feminist researchers argue that to empower women one must
listen directly to what they are saying and respond personally without
hiding behind the facade of the objective researcher. Several authors
(e.g., Carlson, 1972; Wilkinson, 1986) draw on P. Bakan’s (1966)
distinction between ‘‘agentic’’ and ‘‘communal’’ approaches. Agentic
research, involving separating, ordering and

89
 WAYS OF EVALUATING QUALITATIVE STUDIES
89
controlling, is seen as a masculine activity; communal research,
involving sensitivity and personal participation, is seen as feminine.
However, in a feminist critique of this stance, Peplau and Conrad (1989)
argue that attempts to identify a distinctive set of feminist research
methods are mistaken, and that feminist researchers should avail
themselves of the full range of research methods, including quantitative
and experimental approaches.

Deconstructionism
Finally, deconstructionist researchers engage in self-critique, embracing
a postmodern view of the research process. They see the major task of
researchers as being ‘‘deconstruction’’ of the cultural, social, or
epistemological assumptions of their work and that of others. They
embrace radical pluralism, and attempt to speak or give air to multiple
voices while eschewing any attempt to bring these voices together into
a single message. In essence, they attempt to mirror fragmented,
postmodern, multicultural society in their research. For example, a
deconstructionist researcher such as Lather (1991) might present her
findings as a kind of research collage.
Perhaps most importantly, deconstruction is an essential component of
the process of evaluating research, in which one attempts to identify
the implicit assumptions that drive a research study. Slife and Williams
(1995) provide an excellent introduction to this approach. In our view,
deconstructionism is less useful as a primary research method than as a
method for reflecting on and critiquing research. This takes us to our
final topic.

WAYS OF EVALUATING QUALITATIVE STUDIES

As must be obvious from the above discussion, the traditional

psychometric criteria of reliability and validity do not easily carry over
to qualitative approaches. The concepts of face and content validity
can be used without much stretching, and a case can be argued for
adapting some of the other concepts. However, it appears that a more
fruitful approach is to articulate specific criteria for evaluating
qualitative studies. Several scholars have recently attempted to do this
(e.g., Elliott, Fischer, & Rennie, 1999; Stiles, 1993; Yardley, 2000).
We are partial to the Elliott, Fischer et al. (1999) version, not only
because Elliott is a co-author of this text, but also because their
guidelines, although having a broad applicability, were mainly
developed and published within a clinical psychology context. Elliott et
al. are attempting to help journal reviewers and editors evaluate
qualitative studies that have been submitted for publication, but their
framework is relevant to any readers of qualitative studies, as well as to
researchers themselves. They describe some common guidelines
shared by both quantitative and qualitative approaches, for example
respect for participants and use of appropriate methods, and then
guidelines specific to qualitative approaches (see Table 5.3). They
describe each one and then give examples of

89
 WAYS OF EVALUATING QUALITATIVE STUDIES
90
Table 5.3 Summary of Elliott et al.’s (1999) evolving guidelines

A. Publishability guidelines shared by both qualitative and quantitative

approaches
1. Explicit scientific context and purpose
2. Appropriate methods
3. Respect for participants
4. Specification of methods
5. Appropriate discussion
6. Clarity of presentation
7. Contribution to knowledge
B. Publishability guidelines especially pertinent to qualitative research
1. Owning one’s perspective
2. Situating the sample
3. Grounding in examples
4. Providing credibility checks
5. Coherence
6. Accomplishing general versus specific research tasks
7. Resonating with readers
Note. This table has been reproduced with permission from the British Journal of Clinical
Psychology
# The British Psychological Society.

good and bad practice under each. In summary, their guidelines for
qualitative studies are:
Owning one’s perspective. The authors describe their theoretical
orientations and biases, in order to help readers evaluate the researchers’
interpretation of the data. For example, they would state if they were
coming to the research from a psychoanalytic, or from a feminist,
perspective.
Situating the sample. The authors describe the research participants so
that readers can judge how widely the findings might apply.
Grounding in examples. The authors provide enough examples of their
raw data to illustrate the analytic procedures used and to allow the reader
to evaluate their findings. They also stay close to the data; any
speculations that exceed the data are clearly labeled as such.
Providing credibility checks. The researchers use methods for checking
the credibility of the results, for example, analytic auditing (e.g., using
multiple researchers or an additional person who checks the results
against the data), triangulation (examining the phenomenon from
multiple, varied perspectives) and testimonial validity (checking the
results with the original informants or similar others).
Coherence. The interpretation of the data is coherent and integrated, but
at the same time it does not oversimplify the data.
Accomplishing general versus specific research tasks. If the research
aims to achieve a general understanding, then the appropriate range of
people or situations is sampled. If it aims to achieve a specific
understanding of a particular case, that case is described thoroughly
enough for the reader to gain a full understanding.

90
 CONCLUSIONS
91
Resonating with the reader. From the point of view of the reader, the
results are not only believable but seem to capture or make sense of
the phenomenon, enabling the reader to understand the phenomenon
more fully.

CONCLUSIONS

Qualitative methods have now become much more fully accepted within
psychology, and the heat seems to be dying out of the old quantitative
versus qualitative debate. Research methodologists are now focusing
their attention on when best to use either a quantitative or a qualitative
approach, what is the appropriate qualitative method for any given
research question, and how best to appraise qualitative studies.
However, given that we have now described the fundamentals of both
quantitative and qualitative approaches, it is worth briefly considering
how researchers might decide between them.

How do you choose between a Qualitative and a

Quantitative Approach?
We espouse the notion of methodological pluralism: that different
research methods are appropriate for different types of research
question (see Chapter 3). For example, qualitative methods are good for
descriptive questions within a discovery-oriented framework—e.g.,
when you are trying to learn about a phenomenon that has not been
previously researched. Quantitative methods are good for delimited
questions of covariation and comparison, e.g., looking for relationships
between variables.
On the other hand, all methods have weaknesses or limitations, so if
possible it is better to use multiple methods of measuring important
variables, an approach known as triangulation. In other words, it is
unwise to rely solely on one perspective, source or approach (Campbell
& Fiske, 1959; Cronbach & Meehl, 1955; Patton, 2002), because all of
these have their limitations. For example, in psychotherapy outcome
research it is useful to assess client change from the perspective of the
client, the therapist, and a clinical interviewer. Moreover, a qualitative
study focusing on how change occurred would complement the
quantitative data (McLeod, 2001).
Clinical psychology may be gradually entering a more pluralist phase for
pragmatic reasons. A variety of publications have urged psychologists
to adopt a qualitative approach to research (e.g., Elliott, Fischer, &
Rennie, 1999; Henwood & Pidgeon, 1992; Patton, 2002; Richardson,
1996; Smith, in press). However, the acid test—whether qualitative
studies get published in prestigious journals—still reveals a strong
quantitative bias in the field. There is still a residual attitude that
qualitative methods are second class: the saying of Rutherford, the
eminent physicist, that ‘‘qualitative is bad quantitative’’ (quoted in
Stewart, 1989: 219) expresses this viewpoint succinctly. However, one
sign that a pluralist attitude may be taking root is the interest among
the newer generation of researchers. Qualitative methods seem to
appeal particularly to graduate students in applied
91
 CONCLUSIONS
92
social sciences, because they allow much closer contact with clinical
phenomena. In the institutions we are familiar with, an increasing number
of dissertations and theses now employ qualitative methods, perhaps so
much so that there is a danger in some places that traditional
quantitative skills are no longer being acquired. We believe that clinical
psychologists should be competent in both quantitative and qualitative
methods.

Combining Qualitative and Quantitative Methods

It is possible to do research that combines both quantitative and
qualitative methods; the two approaches can often complement each
other. This can take several forms:
1. Beginning research in a new area with qualitative studies, either
pilot research or more elaborate qualitative investigations.
2. Building quantitative studies on earlier qualitative research.
3. Using qualitative methods such as interviews and focus groups to
develop quantitative measures.
4. Using qualitative data to elucidate or explore quantitative findings,
either as an adjunct to a primarily quantitative study or as a follow-
up investigation.
5. Using quantitative data to elucidate qualitative findings, i.e., the
reverse of point 4, often found in sociology articles.
6. Using both kinds of data in a complementary fashion in the same
study (e.g., case studies by Elliott, 1984; Parry et al., 1986).
7. Carrying out separate qualitative and quantitative studies of the
same participants, either to address different questions, or to
address the same question from different angles (e.g., Madill &
Barkham, 1997; Patton, 2002).
As we hope to have made clear, choosing the approach depends largely
on the question you are trying to answer. The next two chapters examine
practical issues in selecting and constructing measures, covering the two
major approaches to psychological measurement: self-report and
observation, looking at each from both qualitative and quantitative points
of view.
We hope that this chapter has given readers a taste of the range of
available qualitative methods, and an understanding of their underlying
philosophies, particularly the distinction between phenomenological and
constructionist methods. This chapter has been mostly theoretically
oriented; Chapter 12 will look in more detail at practical issues in
analyzing qualitative data.

CHAPTER SUMMARY

This chapter has looked at the varieties of qualitative research.

Qualitative approaches use language as their raw material, in order
to examine the
92
 CONCLUSIONS
93

participants’ thoughts, feelings, behavior, or linguistic strategies. Their main

advantage is that they allow a rich description of the phenomena in depth and
detail, sometimes called ‘‘thick description’’ (Geertz, 1973). There are two
central traditions in qualitative research: phenomenology and constructionism.
Phenom- enologists attempt to understand the person’s perceptions and
experiences— their inner world. Constructionists, on the other hand, focus on
how language is used in social interactions. The criteria of reliability and
validity do not translate easily to qualitative research, but it is nevertheless
possible to specify criteria for how qualitative research studies can be
evaluated.

FURTHER READING
Many treatments of qualitative methods have recently been published.
Richardson (1996), Smith (in press), and Willig (2001) give accessible
treatments of the theory and practice of the commonly used approaches.
Some of the old stalwarts, such as Lincoln and Guba (1985) and Patton (2002),
still hold up, and Taylor and Bogdan’s (1998) sociologically oriented text
includes some illustrative studies, although their current edition omits the
fascinating and moving single case account (originally published in Bogdan &
Taylor, 1976) of an articulate ‘‘mentally retarded’’ man, ‘‘Ed Murphy’’, who
describes his life in state institutions. Potter and Wetherell (1987) is a core
reference for the social- psychological approach to discourse analysis. For an
extensive but accessible treatment of the quantitative versus qualitative
debate, see Bryman (1988) and Polkinghorne (1983). Since many qualitative
approaches have their roots in literary theory, it is also worth reading about
them in that context. Eagleton (1983) gives an excellent exposition and
critique of, among other things, phenomenology, hermeneutics, and
poststructuralism as applied to the analysis of literary texts.

93