Realist Evaluation in Practice: Health and Social Work
| Author | Eric Weir,Mansoor Af Kazi |
| DOI | 10.1177/1035719X0700700108 |
| Published date | 01 March 2007 |
| Date | 01 March 2007 |
| Subject Matter | Article |
BOOK REVIEWS
Evaluation Journal of Australasia, Vol. 7, No. 1, 2007
48
This book is intended as a
contribution to the development
of the realist approach to
evaluation. Kazi means to show
what realist evaluation—which
he feels has remained ‘largely
at the level of a philosophy of
science, and as a manifesto for
evaluation research’ (p. 1)—
would be like if it were actually
put to work.
Current approaches
Kazi prefaces his application
of the realist model with an
assessment of the approaches
that have dominated evaluation
heretofore. He describes three,
but two are most relevant to his
argument.
‘Empirical practice’ is roughly
characterized by its focus on
program outcomes, reliance on
randomized controlled trials to
establish program effectiveness,
and on quantitative analysis. Its
strength is its focus on outcomes.
Its weakness is its neglect
of program implementation
and contexts. It is capable of
establishing that a program
worked, but tells us nothing
about how, or the conditions
under which, it did so. As a
result, its fi ndings are not as
helpful as they might be.
‘Interpretivism’ is
characterized by a concern with
helping participants become
more critically aware, in-depth
study of particular situations,
and use of qualitative methods.
Its strength is rich description
of programs and contexts.
Its weakness is its neglect of
outcomes. For this reason, it,
too, is less helpful than it might
to be.
Title: Realist Evaluation in Practice: Health and Social Work
Author: Mansoor AF Kazi
Publisher: Sage, Thousand Oaks, California
Publication date: 2003
Extent/type: 180 pages, paperback
Price: $A71 from Footprint Books which offers a 15% discount to AES members, phone (02) 9997 3973
ISBN: 0-7619-6996-9
The realist model
Kazi began his career
comfortably within the
empirical practice fold, albeit
with a heightened sensitivity to
practitioners’ concerns regarding
the utility of evaluation,
conducting outcomes-focused
studies of programs in education,
health, and social work. He
had already begun trying to
address the need for more
information about programs
and contexts when he fi rst
encountered Ray Pawson and
Nick Tilley’s writings1 relating
realist philosophy of science to
evaluation practice. And it was
realism’s potential for addressing
the ‘how’ and ‘where’ of
causation that appealed to him.
Obviously there is more
to causation than that the
cause precedes the effect. As
a philosophical account of
causation, realism developed in
response to failed efforts to give
an account of the ‘more’ that
is consistent with empiricism.
For realism, the ‘more’ is a real
power in things to produce
their effects. As Kazi puts it,
‘causation is generative rather
than successive’ (p. 25). Things,
in other words, are agents.
Talk of ‘powers’ is apt to be a
bit mystifying, but for realism
powers are themselves explained
by the causal mechanisms by
which causes produce their
effects. This may seem to be
replacing one mystery with
another. A causal mechanism,
however, is simply the structure
of the object possessing the
power—the entities, relations,
and processes that make it
up, and by means of which it
produces the effect. Relating this
to social programs, Kazi explains
that powers do not reside in
individual events, objects, or
properties, but ‘are the result of
complex transactions of many
different kinds of structures on
many different levels’ (p. 23).
For the realist, the complex
‘structured’ and ‘stratifi ed’
nature of reality, the ‘embedded’
nature of agents and events,
makes it imperative that inquiry
address the real-world context of
interaction of causes, conditions,
and contexts. He explains this in
terms of ‘emergence’. Just as the
powers of an agent are emergent
from the powers operative in
the mechanism that explains
them, when we intervene in
the world we are very likely to
encounter unanticipated effects.
The effectiveness of social
programs, Kazi says, ‘can best
be understood in relation to the
structures and elements that
exist, and in relation to how the
interventions of services interact
with other elements in this
stratifi ed reality … What works
in one time-space may not work
in another’ (p. 24).
It is important to note that
this conception of causation
is non-reductive. Explanation
of powers manifested on
one level involves appeal to
powers operating on another
level. Description of a causal
mechanism establishes at least a
tentative conceptual connection
between the cause and the effect.
It is in virtue of the powers
appealed to and the connection
thus established that we can say
that the cause ‘produced’ the
effect; that when the cause is
present, the effect ‘must’ happen;
that if it does not we must either
explain why not or rethink the
proposed mechanism2.
Kazi teases out the
implications of the realist
account of causation for
evaluation in what he calls ‘the
To continue reading
Request your trial