science; this prevailed during the 1960s, 1970s and 1980s, prior to Guba and
Lincoln’s contribution. The ideas of Evert Vedung (1997) and Peter Rossi and
colleagues (1999) are used here to exemplify the goal-bound tradition of
programme evaluation. Vedung (1997: 223–5), for example, considered a
programme to be an entity understood by means of intervention theory consist-
ing of causes (hypotheses underlying intervention logic), intervention per se
(programme), planned conversion (expected results), intended output (in
concrete, measurable terms), intended immediate outcomes, intended inter-
mediate results, and intended ultimate (societal) outcomes. The difference
between this approach and that of Guba and Lincoln is much more than
semantic. According to this paradigm it is important to map inter-linkages
between outcomes (or effects), results, outputs, activities, inputs and objectives.
This approach has recently been labelled result-chain thinking (Uusikylä and
Valovirta, 2002). In the goal-bound approach, the evaluator aims to map causal
interpretations between objectives, inputs, outcomes and effects.
Finally, the third paradigm, which derives from realist evaluation, tries to
combine the elements of goal-free and goal-bound approaches. Pawson and
Tilley (1997: 63–4) argue that social programmes are ‘. . . undeniably, unequivo-
cally, unexceptionally social systems. They . . . comprise the interplays of indi-
vidual and institution, of agency and structure, and of micro and macro social
processes.’ There seems therefore to be much to learn from investigating the
‘social nature’ of programmes. Pawson and Tilley (1997) suggest that an evalu-
ator should focus on certain themes in carrying out evaluation assignments,
including embeddedness, mechanisms, contexts, regularities and change.
The key idea in Pawson and Tilley’s reasoning is the connections between
contexts (C), mechanisms (M) and outcomes (O). This is what they call the CMO
principle. As we understand this logic, an evaluator should focus on the inter-
vention but this alone is not adequate. S/he should also construct the hypotheses,
implement data-collection strategies and analyse data in terms of CMO configur-
ation. This means in practice that an evaluator tries to understand what might
work for whom in certain circumstances, uses multi-method data-collection oper-
ations and data analysis, and finally consolidates his/her findings with the context
in which the programme operates (thus answering the question: what worked for
whom in what circumstances?).
Pawson and Tilley’s conception comes close to the classic ideas of causality
presented by David Hume almost 400 years ago. Hume further examined the
concept of causality and stressed the importance of contiguity – the relationship
of being connected in space and time (see Hume, 1963). Table 1 lists the main
methodological differences between the alternative approaches to programme
evaluation.
Micro–macro conversion will now be considered and the ideas of James
Coleman will form a conceptual bridge to develop a novel conception of causal-
ity in programme evaluation. In addition, conceptual ideas, originally developed
by Kaufmann (1987), related to the intermediate sphere between the micro and
macro, will be linked to ideas presented by Coleman to formulate a better frame-
work for understanding the logic and usefulness of programme evaluation.
Virtanen and Uusikylä: Exploring the Missing Links between Cause and Effect
81