Pp. 145-167 in A. Tashakkori and C. Teddlie (Eds.

), SAGE Handbook of Mixed

Methods in Social and Behavioral Research, 2nd ed. Sage Publications, 2010.

Realism as a Stance for Mixed Methods Research

Joseph A. Maxwell
George Mason University

Kavita Mittapalli, Ph.D.

Independent Researcher and Program Evaluator, MN Associates, Inc.
(Fairfax, VA)
and
Adjunct Faculty, Argosy University (Washington, DC campus) and American
University (Washington, DC)

Philosophical realism, a currently prominent approach in the

philosophy of science, is gaining increased attention as an alternative to both
positivism/empiricism and constructivism as a stance for research and
evaluation in the social sciences (Campbell, 1988; House, 1991; Mark, Henry,
& Julnes, 2000; Maxwell, 1990, 1992, 2004a, 2008; Pawson, 2006; Pawson &
Tilley, 1997; Sayer, 1992, 2000). Contemporary versions of realism have
presented sophisticated approaches to some of the contentious philosophical
issues involved in the “paradigm wars” over qualitative and quantitative
research.

Although there are now a considerable number of substantive mixed

method studies that have employed a realist perspective, realism has
received relatively little notice in discussions of mixed methodology
(exceptions include Greene, 2007; Greene & Hall, this volume; Lipscomb,
2008; and McEvoy & Richards, 2006). We argue that, as a philosophical
perspective that validates and supports key aspects of both qualitative and
quantitative approaches while identifying some specific limitations of each,
realism can constitute a productive stance for mixed method research, and
can facilitate a more effective collaboration between qualitative and
quantitative researchers.

There are many diverse versions of realism across the philosophical

landscape, but a common feature of the realist positions that we discuss
here is an integration of a realist ontology (there is a real world that exists
independently of our perceptions, theories, and constructions) with a
constructivist epistemology (our understanding of this world is inevitably a
construction from our own perspectives and standpoint, and there is no
possibility of attaining a “God’s eye point of view” that is independent of any
Maxwell & Mittapalli, Realism and mixed methods 2

particular viewpoint). In addition, these versions of realism acknowledge the

reality of mental phenomena and the value of an “interpretive” perspective
for studying these (Putnam, 1990, 1999; Sayer, 1992, 2000).

Different terms have been used for such versions of realism, including
"critical" realism (Bhaskar, 1989; Archer, Bhaskar, Collier, Lawson, & Norrie
1998), "experiential" realism (Lakoff, 1987), "subtle" realism (Hammersley,
1992), “emergent” realism (Henry, Julnes, & Mark, 1998; Mark, Henry, &
Julnes, 2000), "natural" realism (Putnam, 1999), “innocent” realism (Haack,
1998, 2003), and “agential” realism (Barad, 2007). We will use the term
“critical realism” in a broad sense to include all of these versions of realism.
(We provide a more detailed description of realism later in this chapter.)

There is a widespread view within mixed method research that the

appropriate philosophical “partner” for qualitative research is constructivism,
and that for quantitative research is postpositivist empiricism (Johnson &
Gray, this volume). This view would seem to make mixed method research a
philosophical oxymoron, or at least a problematic union. Postpositivism and
constructivism disagree on major issues concerning the nature of the objects
of research and our knowledge of these (Guba & Lincoln, 1989), and these
disagreements played a major role in what have been called the “paradigm
wars” between qualitative and quantitative approaches.

In response, methodological pragmatists (e.g., Patton, 2001;

Reichardt & Cook, 1979; Tashakkori & Teddlie, 1998) have claimed that
these philosophical disagreements are not fundamental, and that research
methods are not intrinsically linked to specific philosophical positions. They
have argued that methods can be combined on the basis of their practical
utility, and that paradigmatic conflicts can be ignored. This view has gained
substantial acceptance within the mixed method research community, and
pragmatism has been promoted as the appropriate philosophical stance for
mixed method research (Biesta, this volume; Johnson & Gray, this volume;
Maxcy, 2003; Morgan, this volume; Tashakkori & Teddlie, 2003).

We agree with pragmatists that research practices are not determined

by, or dependent on, philosophical paradigms. A research strategy or
method is not necessarily linked to a single philosophical stance, and any
approach may be informed by one or more of a number of paradigms
(Greene, 2002; Pitman & Maxwell, 1992). However, we believe that the
pragmatist position underestimates the actual influence of philosophical
assumptions on research methods, an influence that is particularly
significant for combining qualitative and quantitative approaches.
Ontological, epistemological, and axiological assumptions are real properties
of researchers and evaluators, part of what Henry, Julnes, and Mark (1998;
Mark, Henry, & Julnes, 2000) call “values.” These assumptions inevitably
Maxwell & Mittapalli, Realism and mixed methods 3

influence researchers’ purposes and actions to some degree, and are often
implicit and not easily abandoned or changed.

For example, mainstream quantitative research has traditionally

presupposed a Humean, regularity view of causation (Mohr, 1996; cf.
Johnson & Gray, this volume), although this is rarely explicit. This
philosophical assumption leads to, and supports, a variable-oriented
approach to research, an emphasis on replicability and general laws, and a
validity strategy based on experimental or statistical controls. These
characteristics, and the philosophical position that informs them, inherently
relegate qualitative research to a secondary role in investigating causality.
This restricts the range of questions for which qualitative methods are seen
as appropriate, and makes mixed method research both more difficult and
less productive (Maxwell, 2004a). On the other hand, qualitative researchers
who accept a “strong” constructivist philosophy reject quantitative
researchers’ characteristic assumption that objective, verifiable knowledge
about the world is possible (Schwandt, 1997, p. 20), and the view of the
world as analyzable in terms of causes (Guba & Lincoln, 1989). This
prevents these qualitative researchers from accepting (let alone using) some
central features of quantitative design, data collection, and analysis.

Urging researchers to simply set aside these assumptions is not just

unrealistic, but counterproductive. Paradigmatic assumptions function not
simply as constraints on methods, but as lenses for viewing the world,
revealing phenomena and generating insights that would be difficult to
obtain with other lenses. This idea is at the heart of Greene’s (2007, pp. 79-
80; Greene & Hall, this volume) “dialectic” stance for doing mixed method
research, in which the goal is to create a dialogue between diverse
perspectives on the phenomena being studied, so as to deepen, rather than
simply broaden or triangulate, the understanding gained. Greene considers it
the most valuable stance for mixed method research, because the
juxtaposition of different lenses or “mental models” that it requires is the
most likely to produce generative insights and depth of understanding, and
also because it promotes a meaningful engagement with difference and a
dialogue across paradigm boundaries.

In this chapter, we argue that realism—in particular, what we call

“critical realism”—can contribute to such a dialogue, and can help resolve
some of the problems created by other perspectives. Realism provides a
philosophical stance that is compatible with the essential methodological
characteristics of both qualitative and quantitative research, and can
facilitate communication and cooperation between the two (Mark, Henry, &
Julnes, 2000; Greene, 2002). However, we also argue that realism has some
specific implications that challenge certain practices in both qualitative and
Maxwell & Mittapalli, Realism and mixed methods 4

quantitative research, and that point to new ways of addressing some

important issues in mixed method research.

We are not arguing for realism as an “alternate paradigm” (Greene,

2007, pp. 82-86) that is the preferred stance for mixed method research. In
fact, we are skeptical of the entire concept of unified “paradigms” in
research, a concept that has dominated the discussion of the relationship
between philosophical assumptions and research methods. So before we
discuss what we see as the potential contributions of a realist perspective,
we want to address the larger issue of paradigms in mixed method research.

Paradigms in mixed method research

The main argument for combining qualitative and quantitative

paradigmatic positions, as well as methods, in mixed method research has
traditionally been their complementarity—that they have different strengths
and limitations, and that using them together allows the researcher to draw
conclusions that would not be possible using either method alone. However,
this argument has usually assumed that the quantitative and qualitative
traditions embody different “paradigms”—ontological, epistemological, and
value assumptions, as well as methodological differences—that are, even if
compatible, distinctly different from one another, and that these differences
are straightforward and easily categorized. Most textbooks or other general
presentations of mixed method research list the relative strengths and
limitations of qualitative and quantitative research, and use these to develop
strategies for combining the two. These lists of strengths are typically
dichotomous, and the characteristics of each approach are presented as
uniform, polar, and complementary.

This dichotomous and polar view of the two approaches has been
challenged by Hammersley (1992) and Howe (2003), and more recently by
Bergman (2008b), Biesta (this volume), Hammersley (2008), and Fielding
(2008), who argue for a more complex and contextualized understanding of
the potential contributions of each approach. Many of the contributors to the
volume edited by Bergman (2008a) have serious reservations about the way
mixed method research has been conceptualized, and see the
qualitative/quantitative distinction as much more problematic than has
traditionally been assumed. Bergman claims that the assumption of generic
“strengths” of each approach, based on paradigm differences, is fallacious,
and that the conventional divide between qualitative and quantitative
methods is to a considerable degree related to “delineating and preserving
identities and ideologies rather than to describe possibilities and limits of a
Maxwell & Mittapalli, Realism and mixed methods 5

rather heterogeneous group of data collection and analysis techniques” (p.

29).

In addition to the critiques by Bergman, Hammersley, and others of

the view that paradigms constitute a set of logically consistent assumptions
that have necessary connections to methods, the view that paradigms are
generally shared by members of a community of researchers is problematic.
While prominent advocates of this view (e.g., Denzin & Lincoln, 2005) now
concede that qualitative researchers don’t all share the same epistemological
assumptions, they still assume, or at least write as if, qualitative researchers
can be divided into distinct “camps” or “moments,” including postpositivist,
constructivist, and postmodern, that do share a particular paradigm.

This view is supported by the assumption that all communities are

united by shared beliefs, values, and practices, a theory exemplified in the
anthropological concept of “culture.” However, this assumption has
frequently been challenged in anthropology by authors who argued that it
denies or ignores the existence of substantial intra-cultural diversity in
communities, and misrepresents the actual processes that generate and
maintain social solidarity (Wallace, 1970; Hannerz, 1992; Maxwell, 1999). It
has also been challenged by postmodern scholars, who generally consider
diversity—within individual identities, as well as within social communities—
to be fundamental rather than superficial (e.g, Bernstein, 1992; Rosenau,
1992). Studies have demonstrated substantial, and often unrecognized,
diversity in the supposedly “paradigmatic” assumptions held by linguists
(McCawley, 1982) and qualitative evaluators (Pitman & Maxwell, 1992). This
issue will be addressed in more general terms below, in considering the
importance of diversity as a real phenomenon.

A perspective that makes little mention of postmodernism (and is

critical of specific aspects of postmodern theory when it does so), but that is
strikingly compatible with postmodernism’s overall insistence on the
pervasive significance of diversity, as well as with a “dialectic” stance for
mixed method research, has been presented by Abbott (2001, 2004). Abbott
argued that ontological and epistemological positions, rather than being
unified, foundational sets of premises that strongly shape the practices of
particular communities of scholars, function instead as heuristics, conceptual
tools that are used to solve specific problems in theory and research. He
stated that if we take any of a large number of debates between polar
positions, such as positivism vs. interpretivism, analysis vs. narrative,
realism vs. constructivism, and so on, we find that these issues can play out
at many different levels, even within communities of scholars that have
adopted one or the other of these positions as characterizing their field at a
broader level. Thus, within the community of sociologists of science, which is
Maxwell & Mittapalli, Realism and mixed methods 6

generally seen as constructivist in orientation, there are internal debates

that can be seen as involving realist vs. constructivist assumptions, and the
debates often employ both realist and constructivist theoretical “moves” by
particular scholars within that community.

One of the many examples that Abbott analyzed was Chambliss’s

study of competitive swimming (Chambliss, 1989; see Example 1). Abbott
argued that the debate over Chambliss’s work shows the power of making a
“realist” or “constructivist” move, even within a largely constructivist field,
creating new leads for research. He stated that “the idea of heuristics is to
open up new topics, to find new things. To do that, sometimes we need to
invoke constructivism, as have the students of occupational prestige.
Sometimes we need a little realism” (Abbott, 2004, p. 191). This position is
quite compatible with Hacking’s (1999) detailed and incisive analysis of
constructivism, uncovering the ways in which particular phenomena (mental
illness, child abuse, nuclear weapons, rocks) can be usefully seen as both
“real” and “social constructs.”

Example 1 [box]

On the basis of five years of ethnographic research, including coaching

swimming teams at different levels and observing and interviewing
swimmers, Chambliss argued that there is no such thing as “talent” as an
explanation of high performance; it is a myth that romanticizes and mystifies
what he called “the mundanity of excellence.” He supported this claim with
detailed evidence from his observations and interviews, showing that high
performance is simply the result of dozens of specific skills, learned or
stumbled upon, that are repeatedly practiced and synthesized into a
coherent whole. Abbott saw this as a constructivist move in the debate over
sports performance; it asserted that “talent” is a social construction that
does not refer to any real causal factor, but is simply a vacuous
“explanation” for high performance.

This move was consistent with the field of sociology of sport, which
was generally seen as constructionist in orientation. However, underlying
Chambliss’s argument for a constructivist interpretation of “talent” was a
realist move, identifying actual skills and practices, and “excellence” as the
outcome of these, as real phenomena rather than simply constructions. As a
result, his work was attacked by others in this field for not treating
“winning,” and the skills that led to this, as themselves social constructions.
Chambliss’s reply was that while selecting winners on the basis of elapsed
times, rather than the beauty or precision of their strokes, was certainly a
social construction, once that construction was made, the factors that lead to
Maxwell & Mittapalli, Realism and mixed methods 7

success in terms of that standard, and the outcomes of races, have a real
existence independent of how they are construed by participants and judges.

From this perspective, epistemological positions look less like the

traditional view of “paradigms”, and more like tools in a toolkit. “Logical
consistency” is the wrong standard to apply to a toolkit. You don’t care if the
tools are all “consistent” with some axiomatic principle; you care if, among
them, they enable you to do the job, to create something that can meet
your needs or accomplish your goals. In the same way, consistency is the
wrong standard to apply to an individual’s or a community’s ontological and
epistemological views. These views, seen as heuristics, are resources for
getting your work done. This approach is similar to Greene’s “dialectic”
stance, but puts more emphasis on the dialectic use of discrete conceptual
tools, rather than “paradigms” in a more global sense.

The rest of this paper explores some of the specific uses of realist
conceptual tools in social research. First, however, we need to describe
realism in more detail as a general approach in both the natural and social
sciences.

What is realism?

In the philosophy of science, including the philosophy of the social

sciences, realism has been an important, if not the dominant, approach for
over 30 years (Baert, 1998, pp. 189-190); realism has been a prominent
position in other areas of philosophy as well (Kulp, 1997). The proliferation
of realist positions has led one realist philosopher to claim that "scientific
realism is a majority position whose advocates are so divided as to appear a
minority" (Leplin, 1984, p. 1). The idea that there is a real world with which
we interact, and to which our concepts and theories refer, has proved to be
a resilient and powerful one that has attracted increased philosophical
attention following the demise of positivism.

Philosophic realism in general is defined by Phillips (1987:205) as "the

view that entities exist independently of being perceived, or independently of
our theories about them." More specifically, Lakoff lists the following
characteristics of what he terms "experiential realism":

(a) a commitment to the existence of a real world, (b) a recognition

that reality places constraints on concepts, (c) a conception of truth
Maxwell & Mittapalli, Realism and mixed methods 8

that goes beyond mere internal coherence, and (d) a commitment to

the existence of stable knowledge of the world. (1987:xv)

In the social sciences, the most important manifestation of realism is

the “critical realist” tradition most closely associated with the work of
Bhaskar (1978, 1989) and others in this tradition (Archer, Bhaskar, Collier,
Lawson, & Norrie, 1998). However, we also draw substantially from other
versions of realism that we see as compatible with the key ideas of this
tradition, in particular those of the philosophers Haack (1998, 2003),
Manicas (2006), and Putnam (1990, 1999), the physicist and historian of
science Barad (2007), the linguist Lakoff (1987; Lakoff & Johnson, 1999),
and the evaluation researchers Pawson and Tilley (1997).

The distinctive feature of these forms of realism is that they deny that
we have any objective or certain knowledge of the world, and accept the
possibility of alternative valid accounts of any phenomenon. All theories
about the world are grounded in a particular perspective and world view, and
all knowledge is partial, incomplete, and fallible. Lakoff states this distinction
between "objectivist" and "realist" views as follows:

Scientific objectivism claims that there is only one fully correct way in
which reality can be divided up into objects, properties, and
relations. . . . Scientific realism, on the other hand, assumes that "the
world is the way it is," while acknowledging that there can be more
than one scientifically correct way of understanding reality in terms of
conceptual schemes with different objects and categories of objects.
(1987, p. 265)

In taking this position, critical realism retains an ontological realism

while accepting a form of epistemological relativism or constructivism. This
position has achieved widespread, if often implicit, acceptance as an
alternative both to naïve realism and to radical constructivist views that
deny the existence of any reality apart from our constructions. Shadish,
Cook, and Campbell (2002) argued that “all scientists are epistemological
constructivists and relativists” in the sense that they believe that both the
ontological world and the worlds of ideology, values, and so forth play a role
in the construction of scientific knowledge (p. 29). Conversely, Schwandt
(1997) stated that

many (if not most, I suspect) qualitative researchers have a common-

sense realist ontology, that is, they take seriously the existence of
things, events, structures, people, meanings, and so forth in the
environment as independent in some way from their experience with
them. (p. 134)
Maxwell & Mittapalli, Realism and mixed methods 9

Schwandt also noted that most social constructivists in the sociology of

science “do not conclude that there is no material reality ‘out there’”
(Schwandt, 1997, p. 20; see also Shadish, Cook, & Campbell, 2002, pp. 28-
31). Ezzy (2002, pp. 15-18) argued similarly that while some
postmodernists deny that reality exists, others simply want to problematize
our assumptions about reality in light of the complexity of our process of
understanding it. He cites Kvale’s claim that while moderate postmodernism
rejects the idea of universal truth, it “accepts the possibility of specific, local,
personal, and community forms of truth with a focus on daily life and local
narrative” (1995, p. 21).

Example 2 [box]

A particularly detailed and sophisticated statement of the sort of

realism we adopt here was presented by the physicist and historian of
science Evelyn Fox Keller (1992), with the assumption that this viewpoint is
so widely shared that it needs no explicit defense. She stated, “I begin with
a few philosophical platitudes about the nature of scientific knowledge upon
which I think we can agree, but which, in any case, will serve to define my
own point of departure” (p. 73):

Scientific theories neither mirror nor correspond to reality.

Like all theories, they are models, in Geertz’s (1973) terms, both
models of and models for, but especially, they are models for;
scientific theories represent in order to intervene, if only in search of
confirmation. And the world in which they aim to intervene is, first and
foremost, the world of material (that is, physical) reality. For this
reason, I prefer to call them tools. From the first experiment to the
latest technology, they facilitate our actions in and on that world,
enabling us not to mirror, but to bump against, to perturb, to
transform that material reality. In this sense scientific theories are
tools for changing the world.

Such theories, or stories, are invented, crafted, or constructed by

human subjects, interacting both with other human subjects and with
nonhuman subjects/objects.

But even granted that they are constructed, and even abandoning the
hope for a one-to-one correspondence with the real, the effectiveness
of these tools in changing the world has something to do with the
relation between theory and reality. To the extent that scientific
theories do in fact “work”—that is, lead to action on things and people
Maxwell & Mittapalli, Realism and mixed methods

10
that, in extreme cases (for example, nuclear weaponry), appear to be
independent of any belief system—they must be said to possess a kind
of “adequacy” in relation to a world that is not itself constituted
symbolically—a world we might designate as “residual reality.”

I take this world of “residual reality” to be vastly larger than any

possible representation we might construct. Accordingly, different
perspectives, different languages will lead to theories that not only
attach to the real in different ways (that is, carve the world at different
joints), but they will attach to different parts of the real—and perhaps
even differently to the same parts (pp. 73-74).

Such versions of realism share many characteristics with philosophical

pragmatism. It is worth noting, therefore, that some of the major figures in
pragmatism were also ontological realists (Maxcy, 2003, p. 56; Biesta, this
volume). Buchler (1940) said of Peirce, the founder of American
pragmatism, that

Underlying every phase of Peirce’s thought is his realism. The

supposition that there are real things—the real is ‘that whose
characters are independent of what anybody may think them to be’—
he regards as the ‘fundamental hypothesis’ of science, for it alone
explains the manner in which minds are compelled to agreement. (p.
xiv).

Contemporary philosophers who integrate pragmatism and realism include

Haack (2006) and Putnam (1990; Conant & Zeglen, 2002); Putnam once
commented that he should have called his version of realism “pragmatic
realism.”

Despite the widespread commonsense acceptance of combining

ontological realism and epistemological constructivism, the application of this
perspective to qualitative research, as advocated by Hammersley (1992) and
Maxwell (1992), was challenged by Smith and Deemer (2000), who asserted
that the ontological concept of a reality independent of our theories can
serve no useful function, since there is no way to employ this that will avoid
the constraints of a relativist epistemology. They concluded that “Maxwell is
unable to show us how to get reality to do some serious work” (p. 883). In
what follows, therefore, we attempt to show how a realist ontology can do
useful work in the methodology and practice of mixed method research, if it
is taken seriously and its implications systematically developed. We do so by
describing some specific implications of critical realism for quantitative,
Maxwell & Mittapalli, Realism and mixed methods

11
qualitative, and mixed method research, showing how a realist perspective
can provide new and useful ways of approaching problems and important
insights into social phenomena.

Given the prominence of realist views in philosophy, it is puzzling that

realism has not had a greater influence on research methodology. Despite
the contributions to a realist approach to social research by Campbell
(1988), Huberman & Miles (1985), Sayer (1992, 2000), Hammersley (1992),
House (1991), and others, philosophic realism seems still to be largely
unnoticed by most researchers (one exception is the field of program
evaluation, where realist approaches developed by Pawson and Tilley (1997)
and Henry, Julnes, and Mark (1998; Mark, Henry, & Julnes, 2000) have had
a significant impact). Even when realism is noticed, it tends to be seen by
quantitative researchers as a commonsense truism with no important
implications, and dismissed by qualitative researchers as simply positivism in
another guise (Mark, Henry, & Julnes, 2000, p. 166).

However, realism is strikingly different from positivism in many of its

premises and implications (Baert, 1998, pp. 192-193). One of the most
significant of these is the realist understanding of causality. Realists have
been among the strongest critics of the “regularity” view of causation that
was typical of positivism and is still dominant in quantitative research
(Maxwell, 2004a). In addition, most critical realists accept the reality of
mental states and attributes, and the importance of these for causal
explanation in the social sciences, positions rejected by both traditional
positivism and constructivism. Both of these aspects of realism are discussed
in more detail below, and constitute two areas in which critical realism can
make an important contribution to mixed method research.

Although some realists have been critical of quantitative and

experimental research (e.g., Sayer, 1992; Pawson & Tilley, 1997), we
believe that realism is a productive stance for both quantitative and
qualitative research (cf. Mark, Henry, & Julnes, 2000). Donald Campbell, a
major figure in the development of experimental methods in social research
and an important influence on quantitative methodology in general, was an
explicit critical realist in the broad sense we use here (Campbell, 1988; cf.
Maxwell & Lincoln, 1990), and his realist perspective was acknowledged by
Weisner (2005, p. 6) as an influence on the mixed method studies he was
involved in.

Realism is also compatible with some of the assumptions and

implications of constructivism and postmodernism, including the idea that
difference is fundamental rather than superficial, a skepticism toward
“general laws”, an antifoundationalist stance, and a relativist epistemology
Maxwell & Mittapalli, Realism and mixed methods

12
(Maxwell, 1995, 1999). It differs from these approaches primarily in its
realist ontology—a commitment to the existence of a real, though not an
“objectively” knowable, world—and its emphasis on causal explanation
(although a fundamentally different concept of causal explanation than that
of the positivists) as intrinsic to social science.

Such an ecumenical approach is so characteristic of realism that Baert

(1998, p. 194) accuses realists of ruling out almost nothing but extreme
positivism. It is true that realism is pragmatic in that it does not discard a
priori those approaches that have shown some ability to increase our
understanding of the world. However, the value of realism does not derive
simply from its compatibility with different approaches to research, or from
its pragmatic orientation to methods. Realism has important implications for
the conduct of research. In the remainder of this paper, therefore, we want
to take seriously a realist ontology, and to outline some of its most
important implications for mixed-method research.

Potential contributions of realism to mixed method research

There are many aspects of mixed method research for which realism
provides a valuable perspective. For example, it is useful to view research
designs as real entities--not simply as models for research, but also as the
actual conceptualizations and practices employed in a specific study. The
latter approach helps a reader of a research publication to understand the
real design of a study, its “logic-in-use,” which may differ substantially from
its “reconstructed logic” (Kaplan, 1964, p. 8) presented in publications
(Maxwell & Loomis, 2003; Maxwell, 2005). This conception of design as a
model of, as well as for, research is exemplified in a classic qualitative study
of medical students (Becker, Geer, Hughes, & Strauss, 1961; see example
3).

Example 3 [box]

Becker et al. (1961) begin their chapter on the "Design of the Study"
by stating that

In one sense, our study had no design. That is, we had no well-
worked-out set of hypotheses to be tested, no data-gathering
instruments purposely designed to secure information relevant to
these hypotheses, no set of analytic procedures specified in advance.
Maxwell & Mittapalli, Realism and mixed methods

13
Insofar as the term "design" implies these features of elaborate prior
planning, our study had none.

If we take the idea of design in a larger and looser sense, using it to

identify those elements of order, system, and consistency our
procedures did exhibit, our study had a design. We can say what this
was by describing our original view of the problem, our theoretical and
methodological commitments, and the way these affected our research
and were affected by it as we proceeded. (1961, p. 17)

A second example of the application of a realist perspective to

research design is to view the relationships that a researcher establishes
with participants and other stakeholders in a study as a real component of
the “design-in-use” of a study, one that is rarely addressed in discussions of
research design and that often is critical to the actual functioning of a study
(Maxwell, 2002, 2005).

In what follows, we focus on four issues for which we feel realism can
make a particularly important contribution to mixed method research: causal
explanation, mind and reality, validity, and diversity.

1. A process approach to causality

For most of the twentieth century, the dominant conception of

causality in the philosophy of science was based on David Hume’s analysis,
generally known as the “regularity” theory of causation (Salmon, 1989;
House, 1991). Hume argued that we can’t directly perceive causal
relationships, only the observed regularities in associations of events, and
rejected any reference to hypothesized or inferred entities and mechanisms.
This view treats the actual process of causality as unobservable, a “black
box,” and focuses simply on discovering whether there is a systematic
relationship between inputs and outputs. This conception of causality is “the
basis of ordinary quantitative research and of the stricture that we need
comparison in order to establish causality” (Mohr, 1996, p. 99).

In quantitative research, the regularity theory of causation is intrinsic

to an approach to explanation that Mohr (1982) called “variance theory.”
Variance theory deals with variables and the correlations among them; it is
based on an analysis of the contribution of differences in measured values of
particular variables to differences in values of other variables. The
comparison of conditions or groups in which the presumed causal factor
takes different values, while other factors are held constant or statistically
controlled, is central to this approach to causation. Thus, variance theory
Maxwell & Mittapalli, Realism and mixed methods

14
tends to be associated with research that employs experimental or
correlational designs, quantitative measurement, and statistical analysis. As
Mohr noted, "the variance-theory model of explanation in social science has
a close affinity to statistics. The archetypal rendering of this idea of causality
is the linear or nonlinear regression model" (1982, p. 42).

In philosophy, the most widely accepted alternative to the regularity

approach to causality is a realist approach that sees causality as
fundamentally referring to the actual causal mechanisms and processes that
are involved in particular events and situations. For the philosophy of science
in general, this approach was most systematically developed by Salmon
(1984; 1989; 1998), who referred to it as the “causal/mechanical” view.
This approach

makes explanatory knowledge into knowledge of the . . . mechanisms

by which nature works. . . . It exhibits the ways in which the things we
want to explain come about. (Salmon, 1989, p. 182-183)

For the social sciences, this approach to explanation closely resembles

what Mohr (1982) called “process theory.” Process theory deals with events
and the processes that connect them; it is based on an analysis of the causal
processes by which some events influence others. It is fundamentally
different from variance theory as a way of thinking about scientific
explanation. Sayer (1992) argued that

much that has been written on methods of explanation assumes that

causation is a matter of regularities in relationships between events,
and that without models of regularities we are left with allegedly
inferior, ‘ad hoc’ narratives. But social science has been singularly
unsuccessful in discovering law-like regularities. One of the main
achievements of recent realist philosophy has been to show that this is
an inevitable consequence of an erroneous view of causation. Realism
replaces the regularity model with one in which objects and social
relations have causal powers which may or may not produce
regularities, and which can be explained independently of them. In
view of this, less weight is put on quantitative methods for discovering
and assessing regularities and more on methods of establishing the
qualitative nature of social objects and relations on which causal
mechanisms depend. (pp. 2-3)

This approach is quite different from variance theory. Pawson and Tilley
(1997), in their realist approach to program evaluation, stated that
Maxwell & Mittapalli, Realism and mixed methods

15
When realists say that the constant conjunction view of one event
producing another is inadequate, they are not attempting to bring
further “intervening” variables into the picture . . . The idea is that the
mechanism is responsible for the relationship itself. A mechanism
is . . . not a variable but an account of the makeup, behaviour and
interrelationship of those processes which are responsible for the
regularity. (pp. 67-68)

Similar distinctions to that between variance and process theory have

been presented by many other writers, including the distinctions between
variable-oriented and case-oriented approaches (Ragin, 1987), propositional
knowledge and case knowledge (Shulman, 1990), and factor theories and
explanatory theories (Yin, 1993, pp. 15-21). Sayer (1992, pp. 241-251)
similarly distinguished between extensive and intensive research designs;
extensive designs address regularities, common patterns, and distributions
of features of populations, while intensive designs focus on how processes
work in particular cases.

These arguments suggests that realist, process-oriented qualitative

investigations deserve a more prominent place in social research, including
experimental research, complementing regularity-based quantitative
research. Shadish, Cook , and Campbell (2002), in what is arguably the
most detailed and sophisticated presentation of the case for experimental
research, stated that

the unique strength of experimentation is in describing the

consequences attributable to deliberately varying a treatment. We call
this causal description. In contrast, experiments do less well in
clarifying the mechanisms through which and the conditions under
which that causal relationship holds—what we call causal explanation.
(p. 9)

Referring to a “delicate balance” between causal descriptions and

causal explanations, they assert that “most experiments can be designed to
provide better explanations than is the case today” (p. 12), and describe
several studies in which qualitative methods were used to substantially
strengthen the understanding of causal mechanisms in experimental
investigations (pp. 390-392).

Realist social researchers also place considerable emphasis on the

context-dependence of causal explanation (e.g., Sayer, 1992, pp. 60-61;
Huberman & Miles, 1985, p. 354). Pawson and Tilley (1997) sum up this
position in their formula “mechanism + context = outcome” (p. xv). They
maintain that “the relationship between causal mechanisms and their effects
Maxwell & Mittapalli, Realism and mixed methods

16
is not fixed, but contingent” (p. 69); it depends on the context within which
the mechanism operates. This is not simply a claim that causal relationships
vary across contexts; it is a more fundamental claim, that the context within
which a causal process occurs is, to a greater or lesser extent, intrinsically
involved in that process, and often cannot be “controlled for” in a variance-
theory sense without misrepresenting the causal mechanism (Sayer, 2000,
pp. 114-118). For the social sciences, the social and cultural contexts of the
phenomenon studied are crucial for understanding the operation of causal
mechanisms.

We argue that a realist alternative to the dominant regularity model of

causality can provide a way out of the somewhat polarized confrontation
between qualitative and quantitative researchers on this issue of causal
investigation (Maxwell, 2004a, 2004b, 2008). It recognizes the explanatory
importance of the context of the phenomena studied, and does so in a way
that does not simply reduce this context to a set of “extraneous variables.”
It relies fundamentally on an understanding of the processes by which an
event or situation occurs, rather than simply a comparison of situations
involving the presence and absence of the presumed cause. Finally, it
legitimates a concern with understanding particular situations and events,
rather than addressing only general patterns. A process theory of causation
does not require abandoning quantitative, variance-based methods for
investigating causality; it simply requires recognition that process-based
approaches are as legitimate as, and often complementary to, variance-
based ones.

2. Mind as part of reality

The neglect of mental phenomena, or the attempt to deal with these

solely within a behavioral, variable-oriented framework, is one of the main
problems qualitative researchers identify with quantitative research, and one
of the main arguments that qualitative researchers make for adopting a
constructivist or interpretivist stance for research, since these approaches
inherently recognize the important of the mental realm. However, the types
of realism that we discuss here treat mental entities as equally real to
physical ones, and as relevant to causal explanations of individual and social
phenomena. Sayer stated that “social phenomena are concept-
dependent . . . What the practices, institutions, rules, roles, or relationships
are depends on what they mean in society to its members” (1992, p. 30).
Emotions, beliefs, values, and so on are part of reality; they are not simply
abstractions from behavior or constructions of the observer. Realism in this
sense therefore does not entail materialism, nor is it simply a cover for a
reductionist agenda that would attempt to eliminate mental concepts from
scientific discourse (Putnam, 1999, p 74 ff.).
Maxwell & Mittapalli, Realism and mixed methods

17
However, realists are not dualists, postulating two different realms of
reality, the physical and the mental. In our view, the clearest and most
credible analysis of this issue has been that of Putnam (1990, 1999), who
argued for the legitimacy of both “mental” and “physical” ways of making
sense of the world. He advocated a distinction between mental and physical
perspectives or languages, both referring to reality, but from different
conceptual standpoints. He argued that “The metaphysical realignment we
propose involves an acquiescence in a plurality of conceptual resources, of
different and mutually irreducible vocabularies . . .coupled with a return not
to dualism but to the ‘naturalism of the common man.’” (1999, p. 38)

Thus, while realism rejects the idea of "multiple realities" in the sense
of independent and incommensurable worlds in which different individuals or
societies live, it is quite compatible with the idea that there are different
valid perspectives on the world. However, it holds that these perspectives,
as held by the people we study, are part of the world that we want to
understand, and that our understanding of these perspectives can be more
or less correct (Phillips, 1990).

A realist approach thus recognizes the reality and importance of

meaning, as well as of physical and behavioral phenomena, as having
explanatory significance, and the essentially interpretive nature of our
understanding of the former (Sayer, 2000, pp. 17-18). Combining this view
with a process-oriented approach to causality can resolve the long-standing
perceived contradiction between “reason” explanations and “cause”
explanations, and integrate both in explanatory theories. Weber's sharp
distinction between causal explanation and interpretive understanding
(1905) obscured the importance of reasons as causal influences on actions,
and thus their role as essential components of any full explanation of human
action. Realism can deal with the apparent dissimilarity of reason
explanations and cause explanations by showing that reasons can plausibly
be seen as real events in a causal nexus leading to the action.

Realism also supports the idea that individuals' social and physical
contexts have a causal influence on their beliefs and perspectives. While this
proposition is widely accepted in everyday life, constructivists have tended
to deny the “reality” of such influences, while positivism and some forms of
post-positivist empiricism tend to simply dismiss the reality or importance of
individuals’ perspectives, or to “operationalize” these to behavioral variables.
From a realist perspective, not only are both individuals’ perspectives and
their situations real phenomena, they are separate phenomena that causally
interact with one another.
Maxwell & Mittapalli, Realism and mixed methods

18
In doing this, a realist perspective can provide a framework for better
understanding the relationship between individuals’ perspectives and their
actual situations. This issue has been a prominent concern in the philosophy
of social science for many years (e.g., MacIntyre, 1967; Menzel, 1978), and
is central to “critical” approaches to qualitative research. Critical realism
treats both individuals’ perspectives and their situations as real phenomena
that causally interact with one another. In this, realism supports the
emphasis that critical theory places on the influence that social and
economic conditions have on beliefs and ideologies. Sayer (1992, pp. 222-
223) stated that the objects of “interpretive” understanding (meanings,
beliefs, motives, and so on) are influenced both by the material
circumstances in which they exist and by the cultural resources that provide
individuals with ways of making sense of their situations. However, critical
realism approaches the understanding of this interaction without assuming
any specific theory of the relationship between material and ideational
phenomena, such as Marxism.

A realist perspective also legitimates and clarifies the concept of

“ideological distortion”—that cultural forms may obscure or misrepresent
aspects of the economic or social system or the physical environment—while
affirming the causal interaction between the physical and social environment
and cultural forms. In particular, realism is compatible with what have been
called “ideological” or “non-reflectionist” approaches to culture, in which
cultural forms that contradict aspects of social structure may serve
ideological functions that act to sustain the social system or constitute
adaptive responses to the physical or social environment (e.g., Maxwell,
1978). An emphasis on causal processes, rather than regularities or laws, in
explaining sociocultural phenomena also allows explanations to be tailored to
single cases and unique circumstances, so that different individuals or social
groups may have different responses to similar situations, depending on
differences in specific personal or cultural characteristics that are causally
relevant to the outcome.

3. Validity and inference quality

Validity and quality are issues for which there have been substantial
disagreements between qualitative and quantitative researchers. The types
of “validity” (many qualitative researchers don’t even use this term)
employed in each tradition have little overlap (Teddlie & Tashakkori, 2003),
and the basic assumptions involved in the two approaches are radically
different. Teddlie and Tashakkori went so far as to recommend abandoning
the term “validity” entirely in mixed method research, arguing that the term
has taken on such diverse meanings that it is losing its ability to
communicate anything (pp. 12, 36-37).
Maxwell & Mittapalli, Realism and mixed methods

19
Despite these differences, there is an important similarity between the
typical quantitative and qualitative approaches to validity. Both focus largely
on the procedures used in collecting data and drawing inferences from these
data. This is particularly obvious in the movement for “evidence-based”
research, which relies almost entirely on the type of research design as the
bases for assessing the validity of the results, with randomized experiments
as the “gold standard” for design quality. However, it also characterizes
prominent approaches to validity (or its analogues) in qualitative research.

A realist concept of validity is quite different from these procedure-

based approaches. Validity, from a realist perspective, is not a matter of
procedures, but of the relationship between the claim and the phenomena
that the claim is about (Norris, 1983; House, 1991; Hammersley, 1992;
Maxwell, 1992). Shadish, Cook, and Campbell, in what is currently the
definitive work on experimental and quasi-experimental research, state that

Validity is a property of inferences. It is not a property of designs or

methods, for the same designs may contribute to more or less valid
inferences under different circumstances. . . . No method guarantees
the validity of an inference. (2002, p. 34; emphasis in original)

Also, as argued by Keller in the passage quoted earlier (Example 2), a

realist approach to validity does not entail that concepts, theories, or claims
“reflect” or “correspond to” reality, only that whether these claims “work”
depends on their relationship to a reality independent of our constructions
(cf. Barad, 2007). While critical realism denies that we can have any
“objective” perception of these phenomena to which we can compare our
claims, it does not abandon the possibility of testing these claims against
evidence about the nature of the phenomena.

We see this process of testing claims against the evidence that is

relevant to the claims as fundamental to a scientific approach in general.
However, the nature of the evidence that is relevant to a claim depends on
the nature of the claim. A claim about a person’s beliefs requires a different
sort of evidence from a claim about the outcome of a randomized trial of a
new drug. Specifically, claims about meanings and perspectives, which fall
under the general category of “interpretive” claims, require quite different
sorts of evidence from claims about behavior, let alone claims about the
relationships between variables. A realist approach to validity also entails
that a valid description, explanation, or interpretation must not only be
supported by evidence, but must address plausible alternative descriptions,
explanations, or interpretations of the phenomenon about which the claim is
made.
Maxwell & Mittapalli, Realism and mixed methods

20
For these reasons, the main approach to validity in experimental
research (e.g., Shadish, Cook, and Campbell, 2002) is grounded in the
concept of a validity threat—a possible way that a conclusion might be
wrong—and ways to address these threats. However, the emphasis has
largely been on the designs and methods used to deal with these threats.
This has been facilitated by the fact that this literature has, consistently with
a regularity view of causality as inherently general, dealt mainly with types
of validity threats, rather than emphasizing the actual ways a specific
conclusion might be wrong in a given study. The importance of the latter
point is implicit in the realist argument above, that validity is not simply
determined by procedures (although procedures are obviously relevant to
the validity of a conclusion), but must be assessed in the specific context of
a particular study. It is also an implication of a realist view of causality as
inherently local rather than general.

A realist perspective on validity can thus be of value to mixed method

researchers by focusing attention on the specific plausible threats to the
conclusions drawn in a given study, which depend on the context and
purposes of that study as well on the methods used.

4. Diversity as a real phenomenon

Finally, realism implies that diversity is itself a real phenomenon. This

fact is most obvious in evolutionary biology, where variation among
organisms is the precondition of evolutionary change. Lewontin (1973)
argued that the Darwinian revolution replaced a Platonic, idealist view of
variation (that variations were simply imperfect approximations to the
“ideal” or “type” of a species) with a realist view that saw actual variation as
the fundamental fact of biology and the cornerstone of evolutionary theory.
Mayr (1982), arguing against the prevailing view that the most important
characteristic of a species is the normal “type” of the organism, stated that
"the most interesting parameter in the statistics of natural populations is the
actual variation, its amount, and its nature" (p. 47).

Similar arguments about diversity have been made for the social
sciences, as described above in discussing paradigms. However, both
qualitative and quantitative research have tendencies, theoretical as well as
methodological, to ignore or suppress diversity in their goal of seeking
general accounts, though in different ways (Maxwell, 1995). Quantitative
research often aggregates data across individuals and settings, and ignores
individual and group diversity that cannot be subsumed into a general
explanation (Shulman, 1986). Because of its emphasis on general
descriptions and causal theories, it tends to impose or generate wide-
ranging but simplistic models that do not take account of individual
Maxwell & Mittapalli, Realism and mixed methods

21
variation, unique contextual influences, diverse meanings, and idiosyncratic
phenomena.

However, qualitative researchers also tend to neglect diversity.

Theoretically, this is often the result of social theories that emphasize
uniformity; such theories include a definition of culture as beliefs or practices
that are shared by members of a society (Maxwell, 1999), and approaches
to community and social order that assume a dichotomy between consensus
and conflict (Maxwell, 1996). Methodologically, the sample size and sampling
strategies used in qualitative studies are often inadequate to fully identify
and characterize the actual diversity that exists in the setting or population
studied, and can lead to simplistic generalizations or the assumption of
greater uniformity or agreement than actually exists.

Mixed method research provides one way to help overcome the

theoretical and methodological characteristics that lead to the neglect of
diversity. Qualitative methods and approaches, which focus on particular
phenomena and processes and their unique contexts, can help to overcome
the biases inherent in universalizing, variable-oriented quantitative methods.
Conversely, quantitative methods can provide systematic evidence for
diversity, and can help to correct a tendency to ignore complexity and to
focus on typical characteristics and shared concepts and themes. However,
doing so effectively requires recognizing the reality of diversity.

To sum up this section, we are not simply claiming that realism is a

productive stance for mixed method research because it is compatible with
both qualitative and quantitative research, and treats the two perspectives
as equally valid and useful. We have also argued that realism has important
implications for both approaches, ones that push both qualitative and
quantitative researchers to examine more closely some issues that they
typically dismiss or ignore. Realism can therefore not only help to integrate
the two approaches into a more coherent combination, and promote closer
and more equal cooperation between qualitative and quantitative
researchers, but can serve to increase the usefulness of both approaches.

Applications of Critical Realism in Mixed Methods Practice

Explicit use of realist perspectives in mixed method research is still

relatively uncommon, and sometimes involves little more than an
acknowledgment that realism has informed the research. For example,
Weisner, in his introduction to a collection of papers on mixed method
studies of children’s development and family life, paid homage to Campbell’s
Maxwell & Mittapalli, Realism and mixed methods

22
realist and multiplist approach, saying that this skeptical realism and holism
“provides the context and tradition for much of our work” (2005, pp. 5-6),
but doesn’t discuss how, specifically, realism did so.

As noted above, the one area in which realist perspectives have had a
major influence on mixed method studies is in program evaluation. The work
of Tilley (described in Pawson & Tilley, 1997) and Mark, Henry, and Julnes
(2000; Henry, Julnes, & Mark, 1998), much of which combined qualitative
and quantitative approaches, has provided a realist alternative to traditional
ways of conceptualizing program evaluations. In addition, Pawson’s (2006)
analysis of literature reviews for evidence-based policy constitutes a major
critique of standard ways of integrating qualitative and quantitative results in
a literature synthesis, and presents a realist alternative to these approaches.

There is also the potential within realist approaches for incorporating

features highlighted by an emancipatory paradigm, and promoting social
justice (House, 1991); this is a significant aspect of Bhaskar’s version of
critical realism, which has been more prominent in Europe than in the United
States. Some important advances based on critical realism have recently
been made in mixed methods research in accounting (Brown & Brignall,
2007; Covaleski & Dirsmith, 1983, 1990), operations management
(Fleetwood, 2000; Reed, 2005a; Mingers, 2000, 2006) economics
(Downward, Finch & Ramsay, 2002; Fleetwood, 1999; Lawson (1989, 1997,
1998, 2001), political science (Patomäki, 2002), and nursing (Lipscomb,
2008; Stickley, 2006).

Modell (2007) used critical realism to develop a unified approach for

validating mixed methods research in accounting management. He argues
that whereas critical realism shows many similarities to, and has indeed
borrowed key concepts from, the pragmatist tradition, it constitutes a more
relevant philosophical foundation to this end. Based on examples from the
field of management accounting and budgeting in the U.S. nursing area and
the jute industry in Bangladesh (see Covaleski & Dirsmith, 1983; Hogue &
Hopper, 1994, 1997), Modell’s work explicates how critical realism may
inform management accounting research by effectively integrating
qualitative and quantitative methods. The examples illustrate the role of
context-specific conditions that may only be captured through deeper
empirical probing and re-conceptualization. In a later study, Covaleski &
Dirsmith (1990) conceded that their quest for a deeper understanding of
budgeting derived from a growing realization of the problematic nature using
budgeting based on the traditional, positivist approach. The authors describe
“freeing” themselves from a priori theories to develop a more contextualized
understanding of the lived experiences of interviewees, and produced a more
multifaceted conceptualization of budgeting (Covaleski & Dirsmith, 1990).
Maxwell & Mittapalli, Realism and mixed methods

23
In economics, critical realism points to the main limitations of
neoclassical economics (based on econometrics principles that are
reductionist in nature and presuppose that concepts can be measured,
counted, manipulated, and cross-classified), and it provides a philosophical
and methodological foundation for a broad set of alternative approaches (see
Downward, Finch & Ramsay, 2002; Downward & Mearman, 2007;
Fleetwood, 1999; Lawson, 1989, 1997, 1998, 2001). In this regard, within
economics, critical realism supports Lawson’s view that the exclusive
dependence on mathematical/statistical modelling in economics is misguided
(Castellacci, 2006).

Olsen (2004) illustrated some of the limitations of the latter approach

in two studies of Indian grain markets and peasant farmers, focusing on
distress sales in the first, and the gendered nature of poverty in the second.
She used both theoretical and methodological triangulation in these studies.
Theoretically, she combined neoclassical, Marxist political economy, and
feminist political economy perspectives; methodologically, she used a survey
of a random sample of farmers, ethnographic observations, in-depth
interviews, family histories, and the analysis and documents and secondary
data. Her results exposed deficiencies in both the neoclassical and Marxist
approaches, and showed the necessity of adding qualitative to quantitative
methods to understand the phenomena studied. She concludes with three
rules of thumb for realist research: a complex and stratified ontology,
explicit value analysis, and getting behind the numbers and mathematical
models to causal mechanisms.

Within the field of psychiatry, where hierarchy and control prevail, a

critical realist perspective offers a model that does not submit to the
dominant discourse but rather recognizes that service users now possess
decision-making power, especially in terms of being able to provide services
that statutory services providers now require (Stickley, 2006). Based on a
power/knowledge discourse, Stickley suggested a critical realist framework
that offers a theoretical explanation for cause and change with an argument
for an alternative to accepted models of service user involvement. He
argued that since mental health nurses are often the workers who have the
most contact with service users, it is essential that they give consideration to
the philosophies and approaches that underpin these models that are
emancipatory for people who use mental health services.

McEvoy and Richards (2006) justified using a critical realist framework

for mixed methods in a case study in nursing of how and why gatekeeping
decisions emerge at the interface between primary care and community
mental health teams. The quantitative survey helped them to identify
patterns of practice, which were confirmed and elaborated by the findings
Maxwell & Mittapalli, Realism and mixed methods

24
from semi-structured interviews. Using mixed methods gave the inquiry a
“better sense of balance and perspective” (p. 66). Additionally, the findings
from both approaches stimulated retroductive reasoning, a process that
involves the construction of hypothetical models as a way of uncovering the
real structures, contexts and mechanisms that are presumed to produce
empirical phenomena (Bhaskar, 1978, 1986, 1989). In addition, reliance on
retroduction necessitates that the researcher is being explicit about what is
being done during the process, including data collection and analysis
(Bollingtoft, 2007), leading to the development of a theoretical model that
explained why gatekeeping decisions tended to emerge in the way they did.

In summary, we believe that realist perspectives and approaches can

make important contributions to mixed method research. These
contributions involve not simply an overall perspective within which
qualitative and quantitative methods and assumptions can be better
integrated, but also specific insights and strategies that can enable mixed
method researchers to better understand the phenomena they study.

References

Abbott, A. (2001). Chaos of disciplines. Chicago: University of Chicago Press.

Abbott, A. (2004). Methods of discovery: Heuristics for the social sciences.

New York: W. W. Norton.

Archer, M., Bhaskar, R., Collier, A., Lawson, T., & Norrie, A. (Eds.) (1998).
Critical realism: Essential readings. London: Routledge.

Baert, P. (1998). Social theory in the twentieth century. Washington Square,

NY: New York University Press.

Barad, K. (2007). Meeting the universe halfway: Quantum physics and the
entanglement of matter and meaning. Durham, NC: Duke University
Press.

Becker, H. S. (1970). Sociological work: Method and substance. New

Brunswick, NJ: Transaction Books.

Becker, H. S. (1986). Writing for social scientists. Chicago: University of

Chicago Press.
Maxwell & Mittapalli, Realism and mixed methods

25
Becker, H. S. (1990). Generalizing from case studies. In Eisner, E. W. &
Peshkin, A. (Eds.), Qualitative inquiry in education: The continuing debate
(pp. 233-242). New York: Teachers College Press.

Becker, H. S., Geer, B., Hughes, E. C., and Strauss, A. L. (1961). Boys in
white: Student culture in medical school. Chicago: University of Chicago
Press. Reprinted by Transaction Books, 1977

Bergman, M. M. (2008a). The straw men of the qualitative-quantitative

divide and their influence on mixed method research. In M. Bergman
(Ed.), Advances in mixed method research (pp. 11-21). London: Sage
Publications.

Bergman, M. M., Ed. (2008b). Advances in mixed method research. London:

Sage Publications.

Bernstein, R. (1992). The new constellation: The ethical-political horizons of

modernity-postmodernity. Cambridge MA: MIT Press.

Bhaskar, R. (1978). A realist theory of science, 2nd ed. Brighton, England:

Harvester.

Bhaskar, R. (1986). Scientific realism and human emancipation. London:

Verso.

Bhaskar, R. (1989). Reclaiming reality: A critical introduction to

contemporary philosophy. London: Verso.

Biesta, G. (2010). Pragmatism and the philosophical foundations of mixed

methods research. Forthcoming in A. Tashakkori & C. Teddlie, Handbook
of mixed methods in social and behavioral research, 2nd edition.
Thousand Oaks, CA: Sage.

Bollingtoft, A. (2007). A critical realist approach to quality in observation

studies. In H. Neergaard and J.P. Ulhoi (Eds.) Handbook of qualitative
research methods in entrepreneurship. Northampton MA: Edward Elgar.

Brown, R. & Brignall, S. (2007). Reflections on the use of a dual-

methodology research design to evaluate accounting and management
practice in UK university central administrative services, Management
Accounting Research, 18, 32-48.

Bryman, A. (1988). Quantity and quality in social research. London: Unwin

Hyman.
Maxwell & Mittapalli, Realism and mixed methods

26
Buchler, J. (1940). “Introduction,” in Buchler (Ed.), The philosophy of Peirce:
Selected writings. Routledge and Kegan Paul.

Campbell, D. T. (1988). Methodology and epistemology for social science:

Selected papers (S. Overman, Ed.). Chicago: University of Chicago Press.

Cartwright, N. (1999). The dappled world: A study of the boundaries of

science. Cambridge: Cambridge University Press.

Castellacci, F. (2006). A critical realist interpretation of evolutionary growth

theorizing. Cambridge Journal of Economics, 30, 861–880.

Chambliss, D. (1989). The mundanity of excellence: An ethnographic report

on stratification and Olympic swimmers. Sociological Theory 7, 70-86.

Conant, J., & Zeglen, U. M. (2002). Hilary Putnam: Pragmatism and realism.
London: Routledge.

Covaleski, M.A., & Dirsmith, M.W. (1983) Budgets as a means of control and
loose coupling. Accounting, Organizations and Society, 8, 323-340.

Covaleski, M.A., & Dirsmith, M.W. (1990). Dialectical tension, double

reflexivity and the everyday accounting researcher: on using qualitative
methods. Accounting, Organizations and Society, 15, 543-547.

Denzin, N. K., & Lincoln, Y. S. (2005). Introduction: The discipline and

practice of qualitative research. In N. K. Denzin & Y. S. Lincoln (Eds.),
Handbook of qualitative research, 3rd ed. (pp. 1-42). Thousand Oaks,
CA: Sage.

Downward, P., Finch, J. H. and Ramsay, J. (2002). Critical realism, empirical

methods and inference: A critical discussion. Cambridge Journal of
Economics, 26, 481–500.

Downward, P., & Mearman, A. (2007). Retroduction as mixed-methods

triangulation in economic research: reorienting economics into social
science. Cambridge Journal of Economics, 31, 77-99.

Ezzy, D. (2002). Qualitative analysis: Practice and innovation. London:

Routledge.

Fielding, N. (2008). Analytic density, postmodernism, and applied multiple

methd research. In M. Bergman (Ed.), Advances in mixed method
research (pp. 37-52). London: Sage Publications
Maxwell & Mittapalli, Realism and mixed methods

27
Fleetwood, S. (1999). Critical realism in economics: Development and
debate, London: Routledge.

Geertz, C. (1973). The interpretation of cultures. New York: Basic Books.

Greene, J. (2002). Understanding social programs through evaluation. In N.

K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research, 2nd
edition (pp. 981-1000). Thousand Oaks, CA: Sage.

Greene, J. (2007). Mixed methods in social inquiry. San Francisco: Jossey-

Bass.

Greene, J., & Hall, J. (this volume). The dialectical stance and pragmatism:
Interplay between philosophy and practice in mixed method research
Forthcoming in A. Tashakkori & C. Teddlie, Handbook of mixed methods
in social and behavioral research, 2nd edition. Thousand Oaks, CA: Sage.

Guba, E.G., & Lincoln, Y.S. (1989). Fourth generation evaluation. Thousand
Oaks, CA: Sage.

Haack, S. (1998). Manifesto of a passionate moderate. Chicago: University

of Chicago Press.

Haack, S. (2003). Defending science—within reason. Amherst, NY:

Prometheus Press.

Haack, S. (2003). Pragmatism, old and new: Selected writings. Amherst,

NY: Prometheus Books.

Hacking, I. (1999). The social construction of what? Cambridge, MA:

Harvard University Press.

Hammersley, M. (1992). Ethnography and Realism. In M. Hammersley,

What's Wrong with Ethnography? Methodological Explorations (pp. 43-
56). London: Routledge.

Hammersley, M. (2008). Troubles with triangulation. In M. Bergman (Ed.),

Advances in mixed method research (pp. 22-36). London: Sage
Publications.

Hannerz, U. (1992). Cultural complexity: Studies in the social organization

of meaning. New York: Columbia University Press.

Henry, G., Julnes, J., & Mark, M. (1998). Realist evaluation: An emerging
theory in support of practice. New Directions for Evaluation 78. San
Francisco: Jossey-Bass.
Maxwell & Mittapalli, Realism and mixed methods

28
Hoque, Z., & Hopper, T. (1994). Rationality, accounting and politics: A case
study of management control in a Bangladeshi jute mill. Management
Accounting Research, 5, 5-30.

Hoque, Z., & Hopper, T. (1997). Political and industrial relations turbulence,
competition and budgeting in the nationalised jute mills of Bangladesh.
Accounting and Business Research, 27, 125-143

House, E. (1991). Realism in research. Educational Researcher, 20 (6), 2-9.

Huberman, A. M., & Miles, M. B. (1985). Assessing local causality in

qualitative research. In D. N. Berg and K. K. Smith, Eds., Exploring
clinical methods for social research (pp. 351-382). Beverly Hills:
Sage Publications.

Johnson, B., & Gray, R. (this volume). The history of integrated research in
the social and behavioral sciences. Forthcoming in A. Tashakkori & C.
Teddlie, Handbook of mixed methods in social and behavioral research,
2nd edition. Thousand Oaks, CA: Sage.

Kaplan, A. (1964). The conduct of inquiry. San Francisco: Chandler.

Keller, E. F. (1992). Secrets of life, secrets of death: Essays on language,

gender, and science. New York: Routledge.

Kulp, C. B., Ed. (1997). Realism/antirealism and epistemology. Lanham MD:

Rowman and Littlefield.

Kvale, S. (1995). Interviews: An introduction to qualitative research

interviewing. Thousand Oaks, CA: Sage.

Lakoff, G. (1987). Women, fire, and dangerous things: What categories

reveal about the mind. Chicago: University of Chicago Press.

Lakoff, G., & Johnson, M. (1999). Philosophy in the flesh: The embodied
mind and its challenge to western thought. New York: Basic Books.

Lawson, T (1989). Abstraction, tendencies and stylized facts: A realist

approach to economic analysis. Cambridge Journal of Economics, 13,
59–78.

Lawson, T. (1997). Economics and Reality. London, Routledge.

Lawson, T. (1998). Transcendental realism. in J. Davis, D.W. Hands, and

Ma¨ki, U. (Eds), The Handbook of Economic Methodology.
Cheltenham: Edward Elgar.
Maxwell & Mittapalli, Realism and mixed methods

29
Lawson, T. (2001). Reorienting Economics. London, Routledge.

Leplin, J. (1984). Introduction. In Leplin, J., Ed., Scientific Realism.

Berkeley: University of California Press.

Lewontin, R. C. (1973). Darwin and Mendel—The Materialist Revolution. In

J. Neyman, Ed., The heritage of Copernicus: Theories "more pleasing
to the mind". Cambridge: MIT Press.

Lipscomb, M. (2008). Mixed method in nursing studies: A critical realist

critique. Nursing Philosophy, 9, 32-45.

MacIntyre. A. (1967). The idea of a social science. Aristotelian society

Supplement 41. Reprinted in B. R. Wilson, Ed., Rationality (pp. 112-
130) (1970). New York: Harper & Row.

Manicas, P. T. (2006). A realist philosophy of social science: Explanation and

understanding. Cambridge: Cambridge University Press.

Mark, M. M., Henry, G. T., & Julnes, G. (2000). Evaluation: An integrated

framework for understanding, guiding, and improving policies and
programs. San Francisco: Jossey-Bass.

Maxcy, S. (2003). Pragmatic threads in mixed method research in the social

sciences: The search for multiple modes of inquiry and the end of the
philosophy of formalism. In A. Tashakkori & C. Teddlie (Eds.), Handbook of mixed
methods in social and behavioral research (pp. 51-89). Thousand Oaks, CA: Sage.

Maxwell, J. A. (1992). Understanding and validity in qualitative research.

Harvard Educational Review 62, pp. 279-300.

Maxwell, J. A. (1995). Diversity and methodology in a changing world.

Pedagogía 30, 32-40 (1995-96).

Maxwell, J. A. (1999). A realist/postmodern concept of culture. In E. L.

Cerroni-Long (Ed.), Anthropological theory in North America (pp. 143–
73). Westport, CT: Bergin & Garvey.

Maxwell, J. A. (2002). Realism and the role of the researcher in qualitative

psychology. In Mechthild Kiegelmann (Ed.), The role of the researcher in
qualitative psychology. Tuebingen, Germany: Verlag Ingeborg Huber.

Maxwell, J. A. (2004a). Causal explanation, qualitative research, and

scientific inquiry in education. Educational Researcher 33 (2), 3-11
(March 2004).
Maxwell & Mittapalli, Realism and mixed methods

30
Maxwell, J. A. (2004b). Using qualitative methods for causal explanation.
Field Methods 16(3), 243-264 (August 2004).

Maxwell, J. A. (2005). Qualitative research design: An interactive approach,

2nd edition. Thousand Oaks, CA: Sage.

Maxwell, J. A. (2008). The value of a realist understanding of causality for

qualitative research. In N. Denzin (Ed.), Qualitative research and the
politics of evidence (pp. 163-181). Walnut Creek, CA: Left Coast Press.

Maxwell, J. A., & Lincoln, Y. (1990). Methodology and epistemology: A

dialogue. Harvard Educational Review 60 (4), 497-512.

Maxwell, J. A., & Loomis, D. (2003). Mixed methods design: An alternative

approach. In A. Tashakkori & C. Teddlie (Eds.), Handbook of mixed
methods in social and behavioral research (pp. 241-271). Thousand Oaks,
CA: Sage.

Mayr, E. (1982). The growth of biological thought: Diversity, evolution, and

inheritance. Cambridge: Harvard University Press

McCawley, J. D. (1982). Thirty million theories of grammar. Chicago:

University of Chicago Press.

McEvoy, P., & Richards, D. (2006). Rationale for using a combination of

quantitative and qualitative methods. Journal of Research in Nursing, 11,
66-78.

Menzel, H. (1978). Meaning--who needs it? In M. Brenner, P. Marsh, and M.

Brenner (Eds.), The social contexts of method (pp. 140-171). New York:
St. Martin's Press.

Mingers, J. (2000). The contribution of critical realism as an underpinning

philosophy for OR/MS and systems, Journal of the Operational Research
Society, 51, 1256-1270.

Mingers, J. (2003). The Place of Statistical Modelling in Management

Science: Critical Realism and Multimethodology. Canterbury Business
School, Canterbury. Working Paper Series No. 45.

Modell, S. (2007). Integrating qualitative and quantitative methods in

management accounting research: A critical realist approach.
Unpublished manuscript. Manchester Business School (Manchester) and
School of Accounting, University of Technology, Sydney.
Maxwell & Mittapalli, Realism and mixed methods

31
Mohr, L. B. (1982). Explaining organizational behavior. San Francisco:
Jossey-Bass.

Mohr, L. B. (1996). The causes of human behavior: Implications for theory

and method in the social sciences. Ann Arbor: University of Michigan
Press.

Morgan, D. (this volume). Pragmatism as a paradigm for mixed methods

research. Forthcoming in A. Tashakkori & C. Teddlie, Handbook of mixed
methods in social and behavioral research, 2nd edition. Thousand Oaks,
CA: Sage.

Norris, S. P. (1983). The inconsistencies at the foundation of construct

validation theory. In Ernest R House (Ed.), Philosophy of Evaluation (pp.
53-74). New Directions for Program Evaluation, no. 19. San Francisco:
Jossey-Bass.

Olsen, W. (2004). Methodological triangulation and realist research: An

Indian exemplar. In B. Carter & C. New, Making realism work: Realist
social theory and empirical research (pp. 135-150). London: Routledge.

Patomäki, H. (2002). After international relations: Critical realism and the

(re)construction of world politics. London: Routledge.

Patton, M.Q. (2001). Qualitative research and evaluation methods, third

edition. Thousand Oaks, CA: Sage Publications.

Pawson, R. (2006). Evidence-based policy: A realist perspective. London:

Sage.

Pawson, R., & Tilley, N. (1997). Realistic evaluation. London: Sage.

Pitman, M. A., and Maxwell, J. A. (1992). Qualitative approaches to

evaluation. In M. D. LeCompte, W. L. Millroy, & J. Preissle (Eds.), The
handbook of qualitative research in education (pp. 729-770). San Diego:
Academic Press.

Phillips, D. C. (1987). Philosophy, science, and social inquiry: Contemporary

methodological controversies in social science and related applied
fields of research. Oxford: Pergamon Press.

Platt, J. R. (1966). "Strong Inference". Science 146, 347-353 (Oct 16,

1966). Reprinted in H. S. Broudy, R. Ennis, & L. I. Krimerman (Eds.),
Philosophy of Educational Research. New York: John Wiley & Sons,
1973.
Maxwell & Mittapalli, Realism and mixed methods

32
Putnam, H. (1990). Realism with a human face, edited by James Conant.
Cambridge: Harvard University Press.

Putnam, H. (1999). The threefold cord: Mind, body, and world. New York:
Columbia University Press.

Ragin, C. C. (1987). The comparative method: Moving beyond qualitative

and quantitative strategies. Berkeley: University of California Press.

Reichardt, C. S., & Cook, T. D. (1979). Beyond qualitative versus

quantitative methods. In T. D. Cook & C. S. Reichardt, Eds., Qualitative
and quantitative methods in program evaluation (pp. 7-32). Thousand
Oaks CA: Sage Publications.

Rosenau, P. M. (1992). Post-modernism and the social sciences. Princeton:

Princeton University Press.

Salmon, W. C. (1984). Scientific explanation and the causal structure of the

world. Princeton: Princeton University Press.

Salmon, W. C. (1989). Four decades of scientific explanation. In P. Kitcher &

W. C. Salmon (Eds.), Scientific explanation (pp. 3-219). Minneapolis:
University of Minnesota Press.

Salmon, W. C. (1998). Causality and explanation. New York: Oxford

University Press.

Sayer, A. (1992). Method in social science: A realist approach, 2nd ed.

London: Routledge.

Sayer, A. (2000). Realism and social science. London: Sage.

Schwandt, T. A. (1997). Qualitative inquiry: A dictionary of terms. Thousand

Oaks, CA: Sage.

Shadish, W. R., T. D. Cook, and D. T. Campbell (2002). Experimental and

quasi-experimental designs for generalized causal inference. Boston:
Houghton Mifflin.

Shulman, L. (1986). Paradigms and programs. In M. C. Wittrock (Ed.),

Handbook of research on teaching, 3rd ed. Reprinted separately by the
American Educational Research Association, 1990.

Schwandt, T. A. (1997). Qualitative inquiry: A dictionary of terms. Thousand

Oaks, CA: Sage.
Maxwell & Mittapalli, Realism and mixed methods

33
Smith, J. K., & Deemer, D. K. (2002). The problem of criteria in the age of
relativism. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of
qualitative research, 2nd edition (pp. 877-896). Thousand Oaks, CA:
Sage.

Stickley, T. (2006). Should service user involvement be consigned to

history? A critical realist perspective. Journal of Psychiatric and Mental
Health Nursing, 13, 570–577.

Tashakkori, A., and Teddlie, C. (1998). Mixed methodology: Combining

qualitative and quantitative approaches. Newbury Park CA: Sage
Publications.

Tashakkori, A., and Teddlie, C. (2003). The past and future of mixed method
research: From data triangulation to mixed model designs. In A.
Tashakkori & C. Teddlie (Eds.), Handbook of mixed methods in social
and behavioral research (pp. 671-701). Thousand Oaks, CA: Sage.

Wallace, A. F. C. (1970). Culture and Personality, 2nd ed. New York:

Random House.

Weber, M. (1905). Critical studies in the logic of the social sciences. Archiv
fuer Sozialwissenschaft und Sozialpolitik. Reprinted in Weber, The
Methodology of the Social Sciences. New York, The Free Press, 1949.

Weisner, T. S. (2005). Discovering successful pathways in children’s

development: Mixed methods in the study of childhood and family life.
Chicago: University of Chicago Press.

Yin, R. (1993). Applications of case study research. Thousand Oaks, CA:

Sage.

Objectives:

Upon finishing this chapter you should be able to:

1. Understand the main characteristics of critical realism, as the term is

used in this chapter.

2. Identify some of the important differences between realism,

positivism, constructivism, and pragmatism that are relevant to mixed
method research.
Maxwell & Mittapalli, Realism and mixed methods

34
3. Understand why the view of paradigms as logically unified sets of
premises that are shared by members of a research community is
problematic.

4. Identify some of the aspects of mixed method research for which

realism can provide a useful perspective.

Questions for reflection and discussion:

1. How does the realist perspective presented in this chapter fit with your
own assumptions about qualitative, quantitative, and mixed method
research? Has the chapter changed your thinking about any of these? Do
you disagree with any of the chapter’s arguments? Why?

2. How could you apply the specific realist approaches described here to
an actual study that you might conduct? What difference would these make
in how you design and carry out the study?

3. How does the perspective on mixed method research presented in this

chapter differ from that in other chapters of the Handbook? How are these
different views helpful to you in understanding mixed method research
publications, or thinking about how to do mixed method research?