The role of theory is important in process evaluation. Process evaluations use a variety of theory and conceptual frameworks to guide the evaluation of interventions, which are detailed below.
- Theory of Change
The UK MRC guidance (Skivington et al. 2021) highlights the importance of using a theory-based approach to evaluation in order to describe “how an intervention is expected to lead to its effects and under what conditions” (p4). Theory-based evaluations often make use of the Theory of Change (ToC) or Diagrammatic Logic Model. A theory of change or diagrammatic logic model explains how an intervention is intended to produce the desired effect.
Using a theory-driven approach such as a Theory of Change (ToC) in process evaluations can provide a framework for synthesising empirical and theoretical evidence to help us understand how and why interventions work or don’t work (underlying causal mechanisms) and the processes through which interventions are implemented. This information can inform intervention design adaptations for optimisation towards achieving intended outcomes (implementation science) and understanding of under what conditions interventions might work best.
De Silva et al. (2014) provide examples of how a Theory of Change can concurrently support the development of interventions and respective evaluations: a ToC (1) acts as a framework for increasing stakeholder engagement and helps in the design of a contextually-relevant interventions and evaluations; (2) enables systematic identification of knowledge gaps (sometimes termed as the evaluation ‘black box’) for generating research questions that strengthen interventions and inform evaluation design; (3) helps map indicators along a causal pathway which can help combine process and effectiveness evaluations into one. Diagrammatic logic models visually map the components, activities, and intended causal pathways of an intervention.
A theory of change also helps identify and test assumptions and hypotheses that have been made for an intervention or programme to be effective and anticipate possible unintended consequences which might emerge if such assumptions are not correct or met (McGill and Borghi, 2015). Use of a ToC approach in process evaluations can also support intervention scale-up and translation into different contexts, as well as research, for example the fields of implementation and programme science (McIntyre et al. 2020).
- Realist evaluations
Realist evaluation is a theory-driven approach that focuses on understanding how, why, and in what contexts interventions generate outcomes. It centres on developing Context-Mechanism-Outcome (CMO) configurations, which explain how mechanisms are triggered by contextual conditions. In process evaluation, realist methods allow for the exploration of complex causal pathways and variation in outcomes across different settings.
Realist evaluation acknowledges that interventions do not work universally. Instead, they work differently for different groups in different settings. For example, a community-based family planning intervention may trigger trust and engagement among service users in one locality (leading to improved health-seeking behaviour), but indifference or resistance in another due to prior negative experiences with health services. Realist evaluators seek to explain these variations by examining contextual influences. Methodologically, realist evaluation typically uses mixed methods, combining qualitative and quantitative data to understand intervention processes.