Chapter 1

Perspectives on the
Epistemological Bases for
Qualitative Research
Carla Willig
Copyright American Psychological Association. Not for further distribution.

This chapter reviews and clarifies the various ways in (see the section Epistemological Frameworks). In the
which qualitative researchers approach the creation of last section, we discuss their implications for the eval-
knowledge. Qualitative research can take many forms. uation of qualitative research (see Evaluation).
Within the general rubric of qualitative research, we
can find a wide range of activities that are driven by
What Is Qualitative Research?
different goals, deploy different research strategies,
and generate different kinds of insights. This means Qualitative research is primarily concerned with
that although all qualitative research shares some meaning. Qualitative researchers are interested in
important attributes (and these will be identified in subjectivity and experience. They want to under-
the next section), it also is characterized by funda- stand better what their research participants’ experi-
mental differences in epistemological orientation. In ences are like, what they mean to them, how they
other words, qualitative researchers can take a range talk about them, and how they make sense of them.
of different positions in relation to questions about Qualitative researchers try to capture the quality and
the nature and status of any knowledge claims that texture of their research participants’ experiences
may be made on the basis of their research. This chap- and aim to understand the implications and conse-
ter maps out the range of epistemological positions quences of those experiences, for participants and
available to qualitative researchers and discusses the for other people. Qualitative research addresses the
implications for the way in which qualitative research following types of questions:
is conducted and evaluated.
The chapter is structured as follows: In the first sec- ■■ What does something feel like? For example,
tion, we are reminded of the nature and purpose of a qualitative researcher might want to find out
qualitative research in general. We identify what it is like to be the only man in an all-female
the most important characteristics of qualitative workplace.
research, those which are shared by all forms of quali- ■■ How is something experienced? For example, we
tative research (see the section What Is Qualitative may want to conduct qualitative research into
Research?). In the second section, we discuss the dif- the experience of being made redundant.
ferent strands within the qualitative research endeavor. ■■ How do people talk about something and with
Here, we focus on the different types of knowledge what consequences? For example, we may ana-
that can be generated on the basis of different lyze naturally occurring conversations about
approaches to qualitative enquiry (see the section Dif- housework and explore subject positions avail-
ferences Among Qualitative Approaches). In the third able to men and women within this.
section, we introduce the various epistemological ■■ How do people make sense of an experience?
frameworks that underpin these different approaches How do they construct its meaning? What does

DOI: 10.1037/13619-002
APA Handbook of Research Methods in Psychology: Vol. 1. Foundations, Planning, Measures, and Psychometrics, H. Cooper (Editor-in-Chief)
5
Copyright © 2012 by the American Psychological Association. All rights reserved.
 Carla Willig

this allow them to do or not to do? To feel or not extent to which they deploy poetic language or a
to feel? For example, a qualitative study could prose style. Qualitative research findings, how-
explore the ways in which people who have been ever, tend not to be represented by numbers or
injured in a road traffic accident make sense equations, they do not involve statistical calcula-
of this experience and how this allows them to tions, and they do not draw conclusions about
position themselves in relation to the accident. probabilities of occurrences or covariations of
■■ How does a particular (social or psychological) phenomena within a population.
event unfold? How do participants experience ■■ Views meaning in context. Qualitative research-
the event? What may be its consequences? For ers are concerned with how individual research
them or for others? For example, we may want participants make sense of specific experiences
to find out how the end of an intimate relation- within particular contexts. This means that any
ship comes about, how those involved experi- meanings identified are specific to the context
ence such an ending, what breaking up means to within which they are constructed and deployed
Copyright American Psychological Association. Not for further distribution.

them, and how it may shape their views of future by the participants. For example, to understand
relationships. what it means to somebody to get married, we
need to know something about the individual’s
Qualitative research does not, and cannot,
life history and their social and cultural context
answer questions about relationships between vari-
as well as their situation at the time of the inter-
ables or about cause-and-effect relationships. Quali-
view. Qualitative research, therefore, tends not to
tative research is concerned with the description and
draw conclusions about what something might
interpretation of research participants’ experiences.
mean in general. Indeed, from a qualitative per-
It tends to prioritize depth of understanding over
spective, it is questionable whether such general-
breadth of coverage, and as such, the knowledge it
ized meanings do, in fact, exist.
generates tends to be localized and context specific.
■■ Incorporates researcher reflexivity. Qualitative
Qualitative researchers do not aim to generalize
researchers’ concern with meaning and inter-
their findings to general populations and they do
pretation means that they need to pay particular
not aim to develop predictive models of human
attention to the ways in which their own beliefs,
behavior. Instead, qualitative researchers tend to
assumptions, and experiences may shape (both
work in a bottom-up fashion, exploring in depth rela-
limit and facilitate) their reading of qualitative
tively small amounts of data (e.g., a small number of
data. For example, whether the researcher has
semistructured interviews, an individual case, or a
personal experience of the phenomenon under
set of documents relating to a specific event), work-
investigation is important and the nature of the
ing through the data line by line. As a result, any
experience (or indeed its absence) needs to be
insights generated on the basis of qualitative analysis
thought about as it inevitably will frame the
tend to be context specific and are not generalizable
researcher’s approach to the topic. Researcher
to general populations.
reflexivity ought to be an integral part of any
Common features of qualitative research include
qualitative study because meaning is always given
the following:
to data and never simply identified or discovered
■■ Presents findings in everyday language. Because within it.
qualitative research aims to capture and convey ■■ Studies the real world. Qualitative research is
the meanings research participants attribute to concerned with participants’ life experiences,
their experiences and actions, research findings which means that ideally qualitative data ought
tend to take the form of verbal accounts. Such to be collected in situ, that is, where and when
qualitative accounts may vary in the extent to the experiences of interest actually take place.
which they are descriptive or interpretative, in Such naturally occurring data include tape
the extent to which they utilize expert discourse recordings of conversations in real-life contexts,
(such as psychological terminology), and in the such as homes, workplaces, or over the telephone,

6
 Perspectives on the Epistemological Bases for Qualitative Research

as well as video recordings of social interactions research methodology are incompatible with one
such as those at football matches, pubs, or clubs. another. The various formal philosophical and epis-
Because collecting naturally occurring data is not temological positions available to qualitative
always ethically or practically possible, however, researchers are mapped out in the section Epistemo-
a lot of qualitative data takes the form of tran- logical Frameworks. In this section, we prepare the
scripts of semistructured interviews with people ground by identifying the major points of tension
who have agreed to talk about their experiences. around which the family of qualitative research
Either way, whether in situ or in the form of organizes itself. These points of tension include
description and reflection after the event, qualita- (a) the role of theory, (b) description versus inter-
tive data always are concerned with real life, that pretation, (c) realism versus relativism, and
is, with events and experiences that take place (d) politics.
irrespective of whether the researcher studies
them. Experimentation has no place in qualita- The Role of Theory
Copyright American Psychological Association. Not for further distribution.

tive research (unless the aim is to study the expe- As indicated in the section What Is Qualitative
rience of taking part in an experiment). Research? although most qualitative research adopts
■■ Is primarily inductive. Unlike hypothetico- an inductive model of knowledge generation, some
deductive research, qualitative research does not qualitative approaches also include a deductive ele-
set out to test hypotheses derived from exist- ment. For example, grounded theory methodology
ing theories. On the contrary, most qualitative involves a process of testing emerging theoretical
research deliberately brackets the researcher’s formulations against incoming data, thus moving
theoretical knowledge to allow novel insights between developing and testing theory as the
and understandings to emerge from the data. research progresses toward saturation. For example,
As such, most qualitative research aspires to a researcher may want to understand what caused a
an inductive model of knowledge generation. fight between rival fans at a football match. The
Exceptions to this do exist, however, and these researcher begins the research with no assumptions
are in the section Differences Among Qualita- about what happened and she or he begins by inter-
tive Approaches. Also, most if not all qualitative viewing bystanders, witnesses, and participants in
researchers recognize that pure induction is an the fight. Preliminary analysis of the data generates a
impossibility given the role of the researcher in hypothesis about what triggered the event and the
the research process and that without some kind researcher returns to the field and conducts further
of theoretical lens data collection and analysis interviews with particular individuals to test the
cannot take place. The challenge to the qualita- hypothesis and to develop it into a coherent account
tive researcher is to enable the data set to speak of how the fight came about. In this case, the theory
for itself (as far as possible) and to surprise the that is being tested is the emergent theory that has
researcher rather than to simply confirm or been conceived on the basis of an inductive process
refute his or her expectations. and does not involve the application of preexisting
theoretical perspectives.
Alternatively, approaches such as psychoanalytic
Differences Among Qualitative
case studies draw on existing theoretical frameworks
Approaches
(such as Freudian or Kleinian theories) to account
Drisko (1997) developed Glaser’s (1992) analogy of for the manifest content of the data. For example,
qualitative research as a “family of approaches” by the researcher may attribute theory-driven meanings
suggesting that “in this family there are some close to an interviewee’s behaviors during the interview
relations, some distant relations, some extended kin, and conclude that the interviewee’s long pauses,
some odd cousins, and a few nasty divorces.” Differ- hesitations, and incomplete sentences signify resis-
ences between qualitative approaches to research tance to acknowledging underlying feelings, such as
can go deep and some varieties of qualitative anger or anxiety. In these cases, theory is imported

7
 Carla Willig

from outside of the study into the research. Another meaning recollection” (descriptive) and “hermeneu-
example of deliberate and purposeful importing of tics of suspicion” (interpretative; see Langdridge,
theory into qualitative research is provided by criti- 2007, Chapter 4, on Ricoeur and hermeneutics;
cal approaches, such as Marxist or feminist analyses, see also Giorgi, 1992, for a discussion of the differ-
whereby a preestablished perspective is applied to ences between interpretative science and descriptive
interpret the data (see Drisko, 1997). Imported the- science).
oretical perspectives supply a lens through which
the data can be read, thus generating insights into Realism Versus Relativism
particular dimensions of experience that have been Qualitative researchers need to think carefully about
identified as being of interest to the researcher or as the status of the products of their research and the
being important for social or political reasons long sorts of claims they wish to make on the basis of
before the data have been collected. them. They need to ask themselves to what extent
their research aims to shed light on reality (i.e., on
Copyright American Psychological Association. Not for further distribution.

Description Versus Interpretation how things are in the world) and to what extent
Qualitative approaches also vary in the extent to they are simply trying to offer reflections that may
which they aspire to move beyond the data and to (or may not) be of use to others who are trying to
interpret what is being presented. That is to say, make sense of their own and others’ experiences. In
they vary in the extent to which they take data “at other words, does their research aim to hold up a
face value.” Some qualitative approaches, such as mirror to reflect reality as it is or is the purpose of
descriptive phenomenology, stay close to research their research to provide a space within which to
participants’ accounts of their experience as the aim engage with and reflect on a particular experiential
of such research is to capture, clarify, and represent phenomenon? Discussions about realism and rela-
the quality and texture of those experiences. Here, tivism in qualitative research are complicated by the
analyzing data means paying close attention to what fact that both the status of the data (as realist or rela-
is being said by the participant, grasping and distill- tivist) and the status of the analysis of the data (as
ing its meaning, and systematically representing it to realist or relativist) need to be established. It is
others. In descriptive approaches to qualitative important to recognize that these are two distinct
research, meaning is found in the text itself or, as but equally important considerations that easily can
Kendall and Murray (2005) put it, “the meaning of get confused or conflated.
any story is embodied in that story” (p. 749). To start with the status of the data, qualitative
Other approaches, such as interpretative phenom- researchers can take a realist position that takes
enology, aspire to go further and to give meaning to data (such as research participants’ accounts) at
participants’ experiences beyond that which the par- face value and treats them akin to witness state-
ticipants may be able or willing to attribute to it. In ments, that is to say, as a description of events that
other words, even without the application of a partic- actually took place in the real world. From such a
ular theory to the data (see the section What Is Qual- position, the researcher would take great care to
itative Research?), it is possible to extract meanings ensure that the data collected are accurate and
that are not immediately obvious to even the person truthful by ensuring that the conditions under
who has produced the account (i.e., the research par- which accounts are produced are favorable (e.g.,
ticipant). For example, existential themes such as that participants feel safe and nondefensive, and
fear of death or fear of meaninglessness may be that nothing will prevent them from opening up
expressed only indirectly and by way of analogy in and telling the truth). Alternatively, the researcher
the research participant’s account, yet an interpreta- can adopt a relativist position in relation to the sta-
tive analysis may conclude that they underpin and, tus of the data, which means that research partici-
indeed, give a deeper meaning to the account. pants’ accounts are of interest not because they
These two positions (descriptive vs. interpreta- inform the researcher about what is actually going
tive) are sometimes referred to as “hermeneutics of on in the world (e.g., what really happened to the

8
 Perspectives on the Epistemological Bases for Qualitative Research

participant), but rather because they tell the Indeed, feminist scholars were instrumental in
researcher something about how the participants introducing and promoting qualitative research
are constructing meaning in their lives. In such a methods in psychology. Because qualitative
case, the researcher is not concerned with the truth research tends to be bottom-up (allowing the
value of what participants are telling him or her; voices of research participants to be heard) and
instead, the aim of the research is to generate rich because it tends to be inductive (avoiding the
and detailed accounts that will enable the imposition of existing concepts and categories),
researcher to gain a better understanding of the qualitative research can be used as part of an
participant’s meaning-making activities. empowerment agenda. Qualitative research also
Moving on to the status of the analysis, again, can be practiced in an egalitarian, participatory,
two broad positions are available to the researcher: and collaborative way (such as in action research
a realist position that aspires to the production of or some types of ethnography in which the
accurate and valid knowledge about what is going research participants set the agenda and shape the
Copyright American Psychological Association. Not for further distribution.

on, either in the social world, in terms of (a) direction of the research), thus allowing the
events that are taking place in this world (this is in researcher to challenge established power rela-
line with the realist position on the status of the tions between (expert) researchers and (naïve)
data) or (b) actions that research participants are research participants.
taking when they construct meaning (this is in line More interpretative versions of qualitative
with the relativist position on the status of the research (see the section Differences Among Quali-
data). In both cases, the researcher’s (metaphori- tative Approaches) adopt a more conventional
cal) task is to hold up a mirror to accurately reflect “knowing” stance, embracing the role of an expert
what is going on either in the world out there, or who, as a result of familiarity with the relevant psy-
inside the mind of the research participant. This chological literature, may be able to understand the
means that it is possible to adopt a realist position participants better than they can understand them-
(i.e., holding up the mirror) in relation to relativist selves. For example, Hollway and Jefferson’s (2000)
data (i.e., the research participant’s constructions). approach to qualitative analysis was based on the
Such a position claims that the researcher can premise that people “may not know why they expe-
accurately and truthfully represent the partici- rience or feel things in the way that they do [and]
pant’s subjective world (i.e., their constructions of are motivated, largely unconsciously, to disguise the
meaning). Alternatively, the research can adopt a meaning of at least some of their feelings and
relativist position in relation to the analysis. This actions” (p. 26). Thus, qualitative researchers have a
would mean abandoning any truth claims regard- range of options regarding the political orientation
ing the analytic insights produced, arguing instead of their research activities. Although qualitative
that what is being offered is the researcher’s read- research often is associated with a liberal, egalitarian
ing of the data, which tells us just as much (or social agenda, not all qualitative research adopts this
more) about the researcher (and his or her meaning- perspective.
making activities) as it does about the participants
or indeed about the social world. It could be
Epistemological Frameworks
argued that a very fine line exists between this type
of research and the sorts of activities that an artist The previous section demonstrated that qualitative
may engage in. researchers can adopt a wide range of positions
regarding the meaning and status of the kind of
Politics knowledge their research generates (or, indeed,
Qualitative research can have an explicitly politi- regarding the extent to which the production of
cal dimension in that some qualitative researchers knowledge is possible or desirable in the first place).
are motivated by a desire to give voice to other- Epistemological positions are characterized by a set
wise underrepresented or oppressed social groups. of assumptions about knowledge and knowing that

9
 Carla Willig

provide answers to the question “What and how can herself a series of questions (see also Willig, 2008,
we know?” Paradoxically, although we tend to think Chapter 1), such as the following:
about research as being about finding answers to
■■ What kind of knowledge do I aim to create?
questions through some form of systematic process
■■ What are the assumptions that I make about the
of empirical enquiry, the starting point of any
(material, social, and psychological) world(s)
research project is, in fact, a set of assumptions that
that I study?
themselves are not based on anything other than
■■ How do I conceptualize the role of the researcher
philosophical reflection. This is inevitable, and it
in the research process? What is the relation-
is important that researchers are aware of, clear
ship between myself and the knowledge I aim to
about, and prepared to acknowledge and own their
generate?
epistemological position. This is not always easy
because the most fundamental assumptions we The remainder of this section looks at the range
make about the world are often unacknowledged of possible answers to these three questions and pro-
Copyright American Psychological Association. Not for further distribution.

and implicit; that is, we take them for granted. This vides examples of research designs informed by the
section maps out the range of epistemological posi- epistemological positions indicated by such answers.
tions available to qualitative researchers and dis- Positions and their concomitant designs will be
cusses their relationships with one another. It also grouped into three broad approaches that are char-
suggests ways in which researchers can identify and acterized by the type of knowledge they aim to cre-
clarify their own assumptions. ate: (a) realist knowledge, (b) phenomenological
Perhaps the easiest way for a researcher to access knowledge, and (c) social constructionist knowl-
the assumptions she or he makes is to ask him- or edge (see Figure 1.1 for a summary).

Three types of knowledge

Realist Phenomenological Social Constructionist

direct critical descriptive interpretative radical moderate

Researcher as:

detective counselor architect

Aims to:

discover/uncover reality understand deconstruct

The (social/psychological) world is:

intelligible & rule-bound experientially diverse socially constructed

Figure 1.1.  Three types of knowledge.

10
 Perspectives on the Epistemological Bases for Qualitative Research

Realist Knowledge (realist versions of) ethnography and grounded the-

Qualitative researchers can use qualitative methods ory methodology as well as such varieties of inter-
of data collection and analysis to obtain a rich, accu- pretative analysis as psychoanalytic approaches (but
rate, detailed, and comprehensive picture of (some these methods also can be used from within a less
aspects of) the social world or of human psychology. realist epistemological framework, which is dis-
The type of knowledge sought in this case aspires to cussed in the section Varieties of Realist
capture and reflect as truthfully as possible some- Knowledge).
thing that is happening in the real world and that
exists independently of the researcher’s, and indeed Phenomenological Knowledge
the research participants’, views or knowledge about Alternatively, qualitative research can aim to pro-
it. The sorts of things a researcher who aspires to duce knowledge about the subjective experience of
generate this type of (realist) knowledge might research participants (rather than about the social or
study include social processes (e.g., what happens psychological patterns or processes that underpin,
Copyright American Psychological Association. Not for further distribution.

when a new member joins an established reading structure, or shape such subjective experiences, as
group or what happens when an organization imple- realist knowledge does). In this case, the researcher
ments a new equal opportunities policy?) and psy- aspires to capture something that exists in the world
chological mechanisms or processes (e.g., how a (namely, the participants’ feelings, thoughts, and
person who suffers from panic attacks plans a journey perceptions—that is, their experiences); however,
on public transport or how people who lost a parent no claim is being made regarding its relationship
at an early age approach intimate relationships). with other facets of the world or indeed regarding
The assumption underpinning this type of the accuracy of the participants’ accounts of their
research is that certain processes or patterns of a experiences (e.g., whether a phenomenological
social or psychological nature characterize or shape account of an embodied experience such as anger or
the behavior or the thinking of research partici- anxiety matches up with objective physiological
pants, and these can be identified and conveyed by measures such as blood pressure or galvanic skin
the researcher. This means that the researcher response). Such research aims to understand experi-
assumes that the (material, social, psychological) ence (rather than to discover what is “really” going
world she or he investigates potentially can be on). In other words, it does not matter whether
understood, provided that the researcher is skilled what a research participant describes is an accurate
enough to uncover the patterns, regularities, struc- reflection of what happened to him or her or a fan-
tures, or laws of behavior that characterize it and tasy; instead, the type of knowledge the researcher is
that generate the social or psychological phenomena trying to obtain is phenomenological knowledge—
we witness (and that constitute one’s data). The that is, knowledge of the quality and texture of the
researcher can succeed or fail in this process, which participant’s experience. For example, a researcher
means that the researcher aspires to generate valid might want to find out what it is like to be living
and reliable knowledge about a social or psychologi- with a diagnosis of psychosis or how a participant
cal phenomenon that exists independently of the experiences the process of going through a divorce.
researcher’s awareness of it. As such, this type of Finding that a participant experiences herself as
research is characterized by a discovery orientation “rejected by the whole world,” for example, consti-
(see Madill, Jordan, & Shirley, 2000). The role of tutes phenomenological knowledge irrespective of
the researcher in this situation is akin to that of a whether the participant really is being rejected by
detective who uses his or her skills, knowledge, and everyone she encounters.
experience to uncover hitherto hidden facts and The task of the researcher in this type of research
who, through his or her labor, makes what appeared is to get as close as possible to the research partici-
puzzling or mysterious intelligible. The kinds of pant’s experience, to step into that person’s shoes,
methods used by qualitative researchers who aim to and to look at the world through his or her eyes,
produce this type of (realist) knowledge include that is to say, to enter his or her world. Here, the

11
 Carla Willig

role of the researcher is similar to that of the person- means that questions about the nature of social and
centered counselor who listens to the client’s account psychological events and experiences are suspended
of his or her experience empathically, without judg- and instead the researcher is concerned with the
ing and without questioning the external validity of social construction of knowledge. Because language
what the client is saying. This means that the plays such an important part in the construction of
researcher assumes that there is more than one knowledge, qualitative researchers who adopt a
world to be studied. This is because researchers who social constructionist orientation to knowledge gen-
seek this type of knowledge are interested in the eration tend to study discourses and the ways in
experiential world of the participant (rather than the which they are deployed within particular contexts.
material, social, or psychological structures that may For example, a researcher might analyze the lan-
give rise to particular experiences—for example, the guage used in policy documents about antisocial
biochemical changes associated with psychosis or behavior to understand how the phenomenon of
the social processes that can give rise to stereotyp- concern—“antisocial behavior”—is constructed
Copyright American Psychological Association. Not for further distribution.

ing); what appear to be the “same” (material, social, within these documents and how the discourses
psychological) conditions (e.g., a divorce, a diagno- used in the documents position those who are con-
sis, an accident) can be experienced in many differ- structed as the targets of proposed interventions.
ent ways, and this means that there are potentially Such an approach to research is based on the
as many (experiential) worlds as there are individu- assumption that all human experience is mediated
als. A researcher who attempts to generate this type by language, which means that all social and psy-
of knowledge asks, “What is the world like for this chological phenomena are constructed in one way
participant?” (rather than “What is the world like or another. It also means that all knowledge about
and what is it about the world that makes a particu- the world and experience of the world is very much
lar experience possible?”). The kinds of methods socially mediated and that individual experiences
used by qualitative researchers who aim to produce are always the product of internalized social con-
this type of (phenomenological) knowledge, unsur- structions. In other words, when participants are
prisingly, tend to be phenomenological methods telling the researcher about their experiences, they
(such as interpretative phenomenological analysis are not seen to be giving voice to an inner reality (as
or descriptive phenomenology, but note that phe- in phenomenological research) or to be providing
nomenological methods engage with the process of information about social or psychological processes
interpretation in a variety of ways that are discussed (as in realist research); instead, the researcher is
in the section Varieties of Phenomenological Knowl- interested in how socially available ways of talking
edge later in this chapter). about the phenomenon of interest (i.e., discourses)
are deployed by the participant and how these may
Social Constructionist Knowledge shape the participant’s experience. Here, the role
Finally, a qualitative researcher can adopt a much of the researcher is to draw attention to the con-
more skeptical position in relation to knowledge structed nature of social reality and to trace the
and argue that what is of interest is not so much specific ways in which particular phenomena are
what is really going on (realist approach to knowl- constructed through discourse and to reflect on the
edge) or how something is actually experienced by consequences of this for those who are affected (that
participants (phenomenological approach) but is to say, who are “positioned”) by these social con-
rather how people talk about the world and, there- structions. As such, the role of the researcher is akin
fore, how they construct versions of reality through to that of an architect who looks at the phenomenon
the use of language. Here, the type of knowledge of interest with a view to how it has been con-
aspired to is not knowledge about the world or structed and from what resources and materials.
knowledge about how things are (experienced) but The most commonly used method to produce this
rather knowledge about the process by which such type of (social constructionist) knowledge is dis-
knowledge is constructed in the first place. This course analysis (of which there are several versions,

12
 Perspectives on the Epistemological Bases for Qualitative Research

including discursive psychology, Foucauldian dis- of exploitation, of oppression) reflects a social real-
course analysis, and critical discourse analysis); ity that needs to be exposed, acknowledged, and
however, other methods such as narrative analysis understood. Again, to call such research “naive” is
and memory work also can be used. to disparage and devalue research that clearly does
Within these three basic approaches to conceptu- have its uses and significance. Perhaps a less value-
alizing the types of knowledge sought by qualitative laden term such as direct realism would be
researchers, each theme has variations (usually in preferable.
the form of more or less radical versions). In the fol- Critical realist approaches to knowledge genera-
lowing section, we identify a variety of positions tion differ from the more direct (or naive) version in
within each approach to knowledge generation. that they are formed on the basis of the assumption
that although the data can tell us about what is
Varieties of Realist Knowledge going on in the real (i.e., material, social, psycholog-
Realist aspirations to knowledge generation range ical) world, it does not do so in a self-evident,
Copyright American Psychological Association. Not for further distribution.

from what is sometimes referred to as naive to more unmediated fashion. In other words, a critical realist
critical varieties. Naive realist approaches are charac- approach does not assume that the data directly
terized by the assumption that a relatively uncom- reflect reality (like a mirror image); rather, the data
plicated and direct relationship exists between what need to be interpreted to provide access to the
presents itself (the data, the evidence) and what is underlying structures that generate the manifesta-
going on (the reality we want to understand). In tions that constitute the data. For example, if we
other words, we assume that the data more or less carry out a participant observation of the social ritu-
directly represent (mirror, reflect) reality. For exam- als and practices that characterize life within a par-
ple, if we wanted to find out how people make deci- ticular community, the data we collect (in the form
sions about whether to have an HIV antibody test of recordings of observations, conversations, inter-
and we interviewed individuals who recently made views, documents, and photographs which capture
such a decision, a naïve realist approach would dic- life in the community, perhaps) would provide us
tate that we take participants’ accounts at face value with information about what members of the com-
and that we accept that their accounts constitute munity do, how they relate to one another, and how
accurate descriptions of how they made their deci- they structure and manage their social life. However,
sion. The task of the researcher, therefore, would be the data would not tell us, directly and explicitly,
(a) to ensure that participants feel safe and comfort- what it might be (e.g., historically or politically) that
able enough to provide the researcher with accurate drives, shapes, and maintains these structures and
and detailed accounts and (b) to analyze the practices. To understand this, we need to move
accounts in such as way as to produce a clear and beyond the data and draw on knowledge, theories,
systematic model of the decision-making process (or and evidence from outside the particular study and
the variety of pathways for decision making if that is use these to account for what we have observed. For
what the accounts indicate). instance, a community’s history, its relations with
To call such research “naive realist” is to belittle neighboring communities or particular geographic
it. The label naive does imply a criticism, and it is conditions may help the researcher explain why
unlikely that a researcher would ever willingly people do what they do.
describe their own research as naïve realist—even if Crucially, from a critical realist standpoint, it is
he or she subscribed to the assumptions about not necessary (in fact, we would not usually expect)
knowledge generation that are associated with this that research participants be aware of the underlying
label. Also, some very valuable research aims to mechanisms or conditions that inform their overt
“give voice” to otherwise-marginalized individuals behaviors and experiences. Research informed by
and communities and is underpinned by the psychoanalytic theory is a good example of critical
assumption that what participants are telling the realist research in that it is assumed that the under-
researcher about their experiences (e.g., of suffering, lying (in this case, psychological) structures that

13
 Carla Willig

generate the manifest, observable phenomena as winning a prize, being invited on an unexpected
(behaviors, symptoms, dreams, slips of the tongue, holiday, or receiving a letter from a long-lost friend).
etc.) are not necessarily accessible to those who The analysis of the interviews would aim to generate
experience them (i.e., the research participants, the an understanding of what characterizes the experience
patients). This assumption, however, does not mean of being surprised; in other words, the researcher
that such structures are not “real.” Critical realist would want to know what it is that people experi-
research can vary in the extent to which it proclaims ence when they are surprised—for instance, the per-
the existence of underlying structures and mecha- son may experience a sense of a loss of control, of
nisms with anything approaching certainty. Some ambivalence, or of uncertainty about how to
researchers have presented their analyses with cau- respond, and perhaps also feelings of joy and excite-
tion and the proviso that the interpretations offered ment. We do not know what characterizes the expe-
are just that—interpretations that represent possibil- rience until we have conducted a phenomenological
ities rather than certainties (e.g., Frosh & Saville- analysis of the data and, as a descriptive phenome-
Copyright American Psychological Association. Not for further distribution.

Young, 2008). Others have taken a much more nologist, we should not allow our experiences,
knowing stance and present their analyses as insights expectations, and assumptions regarding the experi-
into how things (actually, really) are (e.g., how peo- ence of surprise to inform our analysis of the data.
ple function psychologically or how communities are The end product of a descriptive phenomenological
formed; see Hollway & Jefferson, 2000). study would be an account of the structure of the
phenomenon of being surprised that is formed
Varieties of Phenomenological Knowledge entirely on the basis of participants’ accounts of
All phenomenological knowledge aspires to increase their experiences.
the researcher’s understanding of research partici- By contrast, interpretative phenomenology does
pants’ experience. As such, phenomenological not take accounts of experience “at face value” in
knowledge is insider knowledge—that is, a knowl- the same way; instead, interpretative phenomenolo-
edge that attempts to shed light on phenomena gists do move beyond the data in that they step out-
through an understanding of how these phenomena side of the account and reflect on its status as an
present themselves in or through experience; that is account and its wider (social, cultural, psychologi-
to say, how they appear to somebody within a par- cal) meanings. As Larkin, Watts, and Clifton (2006)
ticular context. Differences exist, however, in the put it in their discussion of interpretative phenome-
extent to which phenomenological knowledge bases nological analysis, such interpretative analysis “posi-
itself on the researcher’s interpretation of research tions the initial ‘description’ in relation to a wider
participants’ experience. This means that phenome- social, cultural, and perhaps even theoretical, con-
nological approaches to knowledge generation range text. This second-order account aims to provide a
from descriptive to interpretative varieties. Descrip- critical and conceptual commentary upon the partic-
tive phenomenology is very much concerned with ipants’ personal ‘sense-making’ activities” (p. 104).
capturing experience “precisely as it presents itself, For example, an interpretative phenomenologist
neither adding nor subtracting from it” (Giorgi, 1992, might want to explore the experience of women
p. 121). Descriptive phenomenology does not aim to who have tried and failed to conceive with the help
account for or explain the experience or to attribute of in vitro fertilization. The researcher would start
meanings to it that are imported from outside of the the research process in much the same way as a
account of the actual experience. In other words, it descriptive phenomenologist and conduct semi-
does not go beyond the data. For example, a descrip- structured interviews with women who recently
tive phenomenologist might be interested in the have had this experience. The next step (still in line
phenomenon of being surprised. To better under- with descriptive phenomenology) would be to
stand this phenomenon, the researcher might conduct engage with the interview transcripts with the aim
a series of semistructured interviews with individu- of entering the participant’s world, understanding
als who recently have experienced a surprise (such what it has been like for the participants to go

14
 Perspectives on the Epistemological Bases for Qualitative Research

through the experience, and producing a description constructionism is relativist in the sense that it con-
of the experience that captures its quality and tex- ceptualizes language as a form of social action that
ture, and that portrays its structure and essence. constructs versions of reality; here, it is discourse
The interpretative phenomenologist acknowl- that constructs reality rather than reality that deter-
edges that understanding the participant’s experi- mines how we describe or talk about it. More or less
ence presupposes a process of making sense of the radical strands of social constructionism exist, how-
participant’s account in the first place; in other ever, and not all social constructionist researchers
words, the researcher needs to give meaning to the would describe themselves as relativists. This means
account to understand it. Therefore, through a that social constructionist approaches to knowledge
hermeneutic circle of giving and recovering mean- production can range from radical to more moderate
ing, the researcher is intimately implicated in mak- versions. Research that is concerned with the ways
ing sense of the participant’s account of a failure to in which speakers within a particular social context
conceive a child. In a further interpretative move, strategically deploy discursive resources to achieve a
Copyright American Psychological Association. Not for further distribution.

the researcher may contextualize the participants’ particular interactional objective may be conducted
experience by reflecting on the social and economic from a radical relativist position. Such a position
structures within which women in our culture demands that the researcher abandons any ambition
experience reproduction, or the social and cultural to gain access to the participants’ inner experience
expectations and norms that prevail at the time of or indeed to understand how they make sense of
data collection. The aim of such reflection would be their experience. Instead, the researcher assumes
to make (further) sense of participants’ experiences that participants will construct different versions of
and to understand better how such experiences reality (i.e., of their experiences, their histories, their
are made possible by the context within which memories, their thoughts and feelings) depending
they occur. on the social context within which they find them-
Descriptive and interpretative versions of phe- selves and the stake that they have in this context.
nomenological research therefore differ in their In other words, from a radical social construc-
approach to reflexivity. Although descriptive phe- tionist perspective, there is nothing outside of the
nomenologists believe that it is possible to produce text. Reality is what participants are constructing
descriptions that capture and comprehend the phe- within a particular interaction through discourse.
nomenon as it presents itself, interpretative phenom- This reality does not survive the context within
enologists argue that it is not, in fact, possible to which it has been constructed, as a different reality
produce a pure description of experience in any case will be constructed to suit the next context. This
and that description always involves a certain means that the radical version of social construc-
amount of interpretation. At the most basic level, it is tionism foregrounds the variability and flexibility of
argued, one’s choice of words shapes the meaning of accounts. It aims to understand how and why dis-
what they are trying to convey and this means that, cursive objects and positions are constructed in par-
inevitably, the researcher adds meaning to the data. ticular ways within particular contexts and it
explores the consequences of such constructions for
Varieties of Social Constructionist those who are using them and those who are posi-
Knowledge tioned by them (i.e., the speakers in a conversation).
By way of contrast with realist approaches, the social For example, a researcher might be interested in
constructionist perspective is often described as how people who have decided to commence psycho-
relativist. It is relativist in the sense that it questions therapy introduce themselves to their new psycho-
the “out-there-ness” of the world and it rejects the therapist and how they explain why they are there.
idea that objects, events, and even experiences pre- To obtain suitable data, the researcher would need
cede and inform our descriptions of them. Indeed, to obtain recordings of first sessions of a number
it rejects the notion of description altogether of therapist–client dyads. These recordings would
and replaces it with that of construction. Social be transcribed and then analyzed. The aim of the

15
 Carla Willig

analysis would be to identify the ways in which the Having identified dominant discourses surrounding
participants in the sessions deploy discursive psychotherapy in the 21st century, the researcher
resources and with what consequences. For might then explore the ways in which such dis-
instance, the researcher might observe that some cli- courses position people (e.g., as damaged by their
ents begin by pointing out that they had waited until past, as in need of expert help, as responsible for
they had reached the “end of their tether” before working through their issues) and with what conse-
making the appointment. The researcher might quences (e.g., as a society, we may expect individu-
observe that by doing this, clients position them- als to invest in their mental health and well-being).
selves within a moral discourse and construct them- By grounding discourses in social, cultural, eco-
selves as deserving of help because they have tried nomic, and material structures, more moderate
very hard to sort out their own problems before ask- social constructionist researchers are making refer-
ing for help. Clients may also disclaim an (undesir- ence to something outside of the text. They invoke a
able) identity, perhaps that of a needy person, by reality that preexists and indeed shapes the ways in
Copyright American Psychological Association. Not for further distribution.

emphasizing that they have never sought help before which individuals construct meaning within partic-
and that their present visit to the therapist was an ular contexts. This means that the moderate social
exceptional event. In this way, clients might posi- constructionist position has an affinity with the criti-
tion themselves as responsible adults whose help- cal realist position (see the section Varieties of Real-
seeking is not a sign of weakness or of ist Knowledge). Although radical social
psychopathology. constructionists emphasize people’s ability to play
The important thing to remember is that a radi- with discursive resources and to use them creatively
cal social constructionist researcher would not be to construct the social realities that suit their needs
interested in the validity of these accounts—indeed, at a particular moment in time, moderate social con-
he or she would not believe in the relevance or even structionists are more concerned with the ways in
the possibility of establishing these accounts’ valid- which available discourses can constrain and limit
ity. In other words, it is irrelevant whether clients what can be said or done within particular contexts.
really are seeking help for the first time or whether Figure 1.1 provides a summary of what charac-
they really are (or feel) weak, strong, or needy. The terizes the three different types of knowledge that
point of social constructionist research is to examine qualitative researchers can aim to produce. In this
localized, context-specific discursive productions chapter, I have kept the use of specialist (philosophy
(e.g., of the self as “adult,” as “strong,” “normal,” or of science) terminology to a minimum and instead
“deserving”) and their action orientation and conse- have focused on a description of the assumptions
quences within the specific context. In other words, (about the nature of knowledge, about the world,
the radical social constructionist researcher would about the role of the researcher) that underpin and
be interested only in the particular reality con- characterize the three approaches and that define
structed for the purposes of a specific conversation. their differences. I have argued that what matters is
By contrast, more moderate (that is to say, less that we ask the right questions about a study (i.e.,
relativist) approaches to social constructionist What kind of knowledge is being produced? What
research would want to go beyond the study of are the assumptions that have been made about the
localized deployments of discursive resources and world that is being studied? What is the role of the
make connections between the discourses that are researcher in the research process?) and that these
used within a particular local context and the wider answers will help us to identify (and make explicit)
sociocultural context. For example, the researcher its epistemological foundations. I would argue that
might be interested in exploring contemporary ther- how we then label a particular epistemological posi-
apy culture more generally, looking at self-help tion is of secondary importance as long as we are
texts; television shows that reference psychotherapy; clear about its parameters. Those who are familiar
and “problem pages” in newspapers and magazines, with the qualitative research methodology literature
where experts answer letters from troubled readers. will be aware that, as Ponterotto (2005) has pointed

16
 Perspectives on the Epistemological Bases for Qualitative Research

out, numerous classification schemas in the litera- clarified whether we are applying them to describe
ture aim to classify approaches to qualitative the status of the data (e.g., as descriptions of reality,
research in meaningful and helpful ways and that as witness statements, as individual constructions,
use terminology lifted from the philosophy of sci- as social constructions, etc.) or to the status of our
ence. For example, we find references to “modern- analysis (as accurate knowledge of reality, as an
isms, postmodernism, social constructionism and interpretation, as a construction, as an artistic pro-
constructivism” (Hansen, 2004); “positivism, post- duction, etc.).
positivism, constructivism-interpretivism, and
critical-ideological” approaches (Ponterotto, 2005);
Evaluation
and “positivism, postpositivism, critical theory, con-
structivism, and participatory” approaches (Guba & How can we assess the quality and value of a partic-
Lincoln, 2005). ular piece of qualitative research? Given that quali-
Such classification schemas often are developed tative research is concerned with meaning, and
Copyright American Psychological Association. Not for further distribution.

within the context of formulating a critique of quan- given that it usually takes the form of descriptions
titative research in cases in which qualitative (often or interpretations of research participants’ context-
referred to as “new paradigm”) approaches are con- specific experiences, it follows that the criteria tradi-
trasted with quantitative (often characterized as pos- tionally used to evaluate quantitative research (i.e.,
itivist and postpositivist) approaches. Such critiques reliability, representativeness, generalizability,
are important in their own right, but it is not neces- objectivity, and validity) are not applicable to quali-
sarily helpful to present classifications of qualitative tative research. Does this mean that qualitative
epistemologies within such a context. A preoccupa- research cannot, or should not, be evaluated? Does
tion with contrasting quantitative with qualitative it mean that in qualitative research “anything goes”?
perspectives can lead to a homogenizing of qualita- Opinion is divided on this subject, with some quali-
tive research and a lack of attention to the differ- tative researchers (e.g., Forshaw, 2007) rejecting the
ences between qualitative approaches. As a result, whole notion of “method” in qualitative research
we often find representations of both quantitative (and with it any aspirations to “rigor”), proposing
and qualitative perspectives that lack sophistication that the aim of qualitative research ought to be to
and differentiation and that (despite the use of eru- produce ideas that resonate with readers and that
dite terminology) actually simplify and sometimes generate debate rather than to produce insights that
even caricature both perspectives. Often, a simple claim to have some validity or even truth value. It
dichotomy between a positivist (old paradigm) follows that it is not meaningful to assess the value
quantitative perspective and a constructivist (new of qualitative research in terms other than its cre-
paradigm) qualitative perspective is constructed ativity and originality.
(and this usually positions the former as flawed and Others (myself included; see Willig, 2007, 2008)
in need of replacement by the latter; see also Shad- disagree with this argument, proposing instead that
ish, 1995, for a discussion of common errors and qualitative research involves a process of systematic,
misrepresentations in epistemological debates in the cyclical, and critical reflection whose quality can be
social sciences). The problem with such dichoto- assessed. Like everything else in qualitative research,
mous classifications is that they do not acknowledge however, evaluation is not a simple or a straightfor-
the full range of qualitative epistemologies that, as ward matter. This is because the criteria we use for
indicated, can reach from naive (or better, direct) evaluating a qualitative study must be informed by
realism to radical social constructionism. In other the study’s epistemological position. In other words,
words, not all qualitative research is constructivist, to be able to evaluate a study’s contribution to
not all of it is relativist, and not all of it is interpre- knowledge in a meaningful way, we need to know
tivist. Furthermore, as discussed in the section Dif- what it was the researchers wanted to find out and
ferences Among Qualitative Approaches, references what kind of knowledge they aimed to generate.
to these terms do not mean anything until we have Several authors have compiled lists of generic criteria

17
 Carla Willig

for evaluating qualitative research (e.g., Elliott, for others it is a self-critical lens. Some
Fischer, & Rennie, 1999; Henwood & Pidgeon, researchers utilize reflexivity to intro-
1992; Yardley, 2000) and although some overlap spect, as a source of personal insight,
exists between these, as I have argued elsewhere, while others employ it to interrogate the
“it is clear that authors approach the question of rhetoric underlying shared social dis-
evaluation from the particular standpoint afforded courses. Some treat it as a methodologi-
by their own preferred methodological approach” cal tool to ensure “truth,” while others
(Willig, 2008, p. 152). exploit it as weapon to undermine truth
I concur with Madill et al. (2000) and Reicher claims. (p. x)
(2000), who have argued that no such thing as a
unified qualitative research paradigm exists and, This means that reflexivity can be used in differ-
therefore, that the criteria we use to evaluate quali- ent ways and for different purposes. For example,
tative studies need to be tailored to fit the particular for a direct realist researcher, reflexivity can be a way
Copyright American Psychological Association. Not for further distribution.

methodology they are meant to appraise. For exam- of acknowledging and bracketing off personal
ple, Madill et al. proposed that objectivity (i.e., the expectations and assumptions so that they do not
absence of bias on the part of the researcher) and make their way into the analysis and distort (or even
reliability (i.e., the extent to which findings have silence) the participant’s voice that is trying to make
been triangulated) are criteria that can be applied itself heard. By contrast, an interpretative phenome-
meaningfully to evaluate realist research, whereas nological researcher may draw on his or her own
from a radical constructionist point of view, any cri- thoughts and feelings about what the participant is
teria that are concerned with the accuracy or saying to uncover meanings within it that are not
authenticity of accounts would be meaningless. immediately obvious to the participant. Finally, a
Instead, to evaluate such studies, we would need to radical social constructionist researcher can use
assess their internal coherence (i.e., the extent to reflexivity to trace the ways in which his or her own
which the analytic narrative “hangs together” with- contributions to the conversation with the partici-
out internal contradictions), to establish deviant case pant have positioned the participant and how this
analysis (i.e., the extent to which the limits of the may have shaped the interview.
applicability of the analytic insights have been iden- Again, these differences have implications for the
tified), and reader evaluation (i.e., the extent to evaluation of a qualitative study in that the use of
which the study is perceived by its readers to reflexivity within the design of the study needs to be
increase their insights and understanding). Finally, assessed in its own terms. In other words, we need
an evaluation of what Madill et al. described as con- to ask whether reflexivity has been used in a way
textual constructionist research (and that is compati- that is compatible with the epistemological orienta-
ble with the phenomenological perspective tion of the study and whether the use of reflexivity
identified in this chapter) requires scrutiny of the within the study’s design has met its own objectives.
study’s use of reflexivity and the extent to which it From our discussion of evaluation so far, it should
explores (and ideally theorizes) the relationship have become clear that to make meaningful evalua-
between accounts (i.e., both the participants’ tion possible, a study’s author needs to clearly iden-
accounts, that is to say the data as well as the tify the study’s epistemological position. Therefore,
researcher’s analytic account) and the context(s) the most important criterion for evaluating qualita-
within which these have been produced. Finlay and tive research ought to be epistemological reflexivity
Gough (2003) proposed that different “versions of (i.e., the extent to which a study clearly and unam-
reflexivity” reflect different epistemological orienta- biguously identifies its epistemological stance) as
tions so that this is a precondition for any further evaluation.
Indeed, Madill et al. (2000) concluded that “qualita-
for some, reflexivity is celebrated as part tive researchers have a responsibility to make their
of our essential human capacity, while epistemological position clear, conduct their

18
 Perspectives on the Epistemological Bases for Qualitative Research

research in a manner consistent with that position, constituting alternative visions of what valid or use-
and present their findings in a way that allows them ful knowledge should look like, are simply provid-
to be evaluated appropriately” (p. 17). ing three different angles from which to view human
experience. They shed light on three different
aspects of human experience. From this point of
Conclusion
view, qualitative research is about attempting to dis-
The aim of this chapter is to review and clarify the cover new aspects of a totality that never can be
various ways in which qualitative researchers accessed directly or captured in its entirety.
approach the production of knowledge. It has been Cohn (2005) referred to this as the “amplifica-
suggested that qualitative researchers can aim to tion” of meaning. To illustrate this way of thinking,
produce three types of knowledge and these were and to illustrate what amplification of meaning may
given the labels realist, phenomenological, and social involve, let us imagine a researcher who wants to
constructionist. Each of these types of knowledge understand what happens when someone is diag-
Copyright American Psychological Association. Not for further distribution.

was shown to be formed on the basis of different nosed with a terminal illness. First, the researcher
answers to questions about the nature and status of might want to listen to first-person accounts of this
knowledge claims, the assumptions the researcher experience. To this end, she conducts semistruc-
makes about the social and psychological worlds she tured interviews with a number of participants who
or he is studying, and the role of the researcher in have had this experience. Her aim is to shed light on
the research process. It was proposed that different the experience of receiving a terminal diagnosis. At
methods of data collection and analysis are required this point, the researcher adopts a realist approach,
to generate the different types of knowledge, and taking the accounts at face value. She produces a
that the evaluative criteria we use to assess the value thematic analysis that aims to capture and systemati-
and quality of a qualitative study may differ depend- cally represent how the participants experienced the
ing on the type of knowledge the study aspires to process of being given their diagnosis. She identifies
produce. To develop these epistemological argu- a number of interesting patterns in relation to the
ments and to clearly distinguish among the three ways in which participants were treated by medical
positions, we have foregrounded their differences. staff and perhaps also in the ways in which the par-
In this concluding section, I return to the bigger pic- ticipants’ loved ones responded to the situation. The
ture and reflect on the ways in which the three research could end here, having produced some use-
approaches complement one another. Each research ful and important insights.
project is motivated and driven by a research ques- Let us assume that the researcher has the time
tion that specifies which aspect or dimension of and motivation to continue with the research. Let us
social or psychological reality the study aims to shed also assume that the researcher had noticed that,
light on. No study ever seeks to simply study (the despite their many shared experiences with medical
meaning of) life as such or to understand the world staff and loved ones, the participants gave quite dif-
in general. Even realist research only ever seeks to ferent meanings to their illnesses. She also noticed
establish the truth about something in particular that this seemed to inform the participants’ sense of
rather than simply the truth. In addition, every themselves as a terminally ill patient and how they
study will have to work within a set of practical con- felt about their illness. To better understand these
straints (such as available time and finances, for differences, the researcher arranges further inter-
example) which set limits to what it can aspire to views with the participants, this time using a phe-
find out. nomenological approach to explore their subjective
All this means that even the most carefully experience in greater depth. This phase of the
designed study can never achieve more than to shed research generates a further set of themes, this time
light on one small part of a much bigger whole. It capturing the existential dimensions of the experi-
could be argued, therefore, that the three types of ence of being diagnosed with a terminal illness and
knowledge identified in this chapter, rather than the range of existential meanings that can be given

19
 Carla Willig

to such an experience. Again, the research could end British Journal of Clinical Psychology, 38, 215–229.
at this point. Let us assume, however, that the doi:10.1348/014466599162782
researcher is still willing and able to continue with Finlay, L., & Gough, B. (Eds.). (2003). Reflexivity: A prac-
tical guide for researchers in health and social sciences.
her project. She reflects on the fact that all the par-
Oxford, England: Blackwell.
ticipants included references to the question of
Forshaw, M. J. (2007). Free qualitative research from the
responsibility (for the illness) and that many of them shackles of method. The Psychologist, 20, 478–479.
grappled with issues around blame (for the illness)
Frosh, S., & Saville-Young, L. (2008). Psychoanalytic
in their accounts. She decides that she wants to find approaches to qualitative psychology. In C. Willig
out more about this and adopts a social construc- & W. Stainton Rogers (Eds.), The Sage handbook
tionist approach, focusing on the use of discourses of qualitative research in psychology (pp. 109–126).
London, England: Sage.
of individual responsibility within the context of ter-
minal illness. She returns to the data (both sets of Giorgi, A. (1992). Description versus interpretation:
Competing alternative strategies for qualitative
interviews) and analyzes them again, this time using
Copyright American Psychological Association. Not for further distribution.

research. Journal of Phenomenological Psychology, 23,

a discourse analytic approach. To contextualize her 119–135. doi:10.1163/156916292X00090
participants’ use of discourse in their constructions Glaser, B. (1992). Basics of grounded theory analysis. Mill
of meaning around their terminal diagnosis, the Valley, CA: Sociology Press.
researcher analyzes newspaper articles and television Guba, E. G., & Lincoln, Y. S. (2005). Paradigmatic con-
documentaries about terminal illness and compares troversies, contradictions, and emerging influences.
In N. K. Denzin & Y. S. Lincoln (Eds.), The Sage
the discursive constructions used in those docu- handbook of qualitative research (3rd ed., pp. 191–215).
ments with those deployed by the participants. Thousand Oaks, CA: Sage.
Much more could be done to shed further light Hansen, J. T. (2004). Thoughts on knowing: Epistemic
on the experience of being diagnosed with a termi- implications of counselling practice. Journal of
nal illness, but let us take pity on our hypothetical Counseling and Development, 82, 131–138.
researcher and stop here. It remains for us to con- Henwood, K. L., & Pidgeon, N. F. (1992). Qualitative
clude that, rather than being mutually exclusive, research and psychological theorising. British Journal
of Psychology, 83, 97–111. doi:10.1111/j.2044-
realist, phenomenological, and social constructionist 8295.1992.tb02426.x
forms of knowing can be thought of as providing
Hollway, W., & Jefferson, T. (2000). Doing qualitative
access to different aspects of our social and psycho- research differently: Free association, narrative and the
logical world(s) and that our choice of which one(s) interview method. London, England: Sage.
to mobilize within the context of a particular Kendall, M., & Murray, S. A. (2005). Tales of the unex-
research project is a question of knowing what we pected: Patients’ poetic accounts of the journey to a
want to know on this particular occasion. diagnosis of lung cancer: A prospective serial qualita-
tive interview study. Qualitative Inquiry, 11, 733–751.
doi:10.1177/1077800405276819
References Langdridge, D. (2007). Phenomenological psychology:
Theory, research and method. London, England:
Cohn, H. W. (2005). Interpretation: Explanation or Pearson Prentice Hall.
understanding. In E. van Deurzen & C. Arnold-
Baker (Eds.), Existential perspectives on human issues. Larkin, M., Watts, S., & Clifton, E. (2006). Giving voice
A handbook for therapeutic practice (pp. 221–226). and making sense in interpretative phenomenologi-
Basingstoke, England: Palgrave Macmillan. cal analysis. Qualitative Research in Psychology, 3,
102–120. doi:10.1191/1478088706qp062oa
Denzin, N. K., & Lincoln, Y. S. (Eds.). (2005). The Sage
handbook of qualitative research (3rd ed.). Thousand Madill, A., Jordan, A., & Shirley, C. (2000). Objectivity
Oaks, CA: Sage. and reliability in qualitative analysis: Realist,
contextualist and radical constructionist episte-
Drisko, J. W. (1997). Strengthening qualitative studies mologies. British Journal of Psychology, 91, 1–20.
and reports: Standards to promote academic integ- doi:10.1348/000712600161646
rity. Journal of Social Work Education, 33, 185–197.
Ponterotto, J. G. (2005). Qualitative research in counselling
Elliott, R., Fischer, C. T., & Rennie, D. L. (1999). psychology: A primer on research paradigms and phi-
Evolving guidelines for publication of qualitative losophy of science. Journal of Counseling Psychology,
research studies in psychology and related fields. 52, 126–136. doi:10.1037/0022-0167.52.2.126

20
 Perspectives on the Epistemological Bases for Qualitative Research

Reicher, S. (2000). Against methodolatry: Some Willig, C. (2007). Qualitative research: The need for sys-
comments on Elliott, Fischer and Rennie. tem. [Letter to the editor]. The Psychologist, 20, 597.
British Journal of Clinical Psychology, 39, 1–6. Willig, C. (2008). Introducing qualitative research in psy-
doi:10.1348/014466500163031 chology (2nd ed.). Maidenhead, England: McGraw
Shadish, W. R. (1995). Philosophy of science and the Hill Open University Press.
quantitative-qualitative debates: Thirteen common Yardley, L. (2000). Dilemmas in qualitative health
errors. Evaluation and Program Planning, 18, 63–75. research. Psychology and Health, 15, 215–228.
doi:10.1016/0149-7189(94)00050-8 doi:10.1080/08870440008400302
Copyright American Psychological Association. Not for further distribution.

21