Process Evaluation
- Introduced by
- Health-promotion & MRC evaluation tradition (Saunders et al.; Moore et al.)
- Year
- 2015
- Type
- Implementation-focused program evaluation
- Family
- Program evaluation methodology
What does it do?
Process evaluation examines how a program or policy was actually implemented, rather than only whether it achieved its outcomes.
It documents what was delivered, to whom, how much, how well and in what context, so that outcome findings can be interpreted correctly. By assessing implementation fidelity, dose, reach, and the mechanisms and contextual factors at work, process evaluation explains why an intervention succeeded or failed and distinguishes a flawed program theory from a sound theory that was poorly delivered. The UK Medical Research Council's 2015 guidance and earlier health-promotion frameworks consolidated it as a core component of evaluating complex interventions.
The intuition behind the method
Suppose a school anti-bullying program shows no effect. Before concluding the idea is worthless, you should ask whether it was ever really delivered. Maybe only half the schools ran the lessons, teachers skipped the hardest sessions, or the students who most needed it never attended. Process evaluation is the part of the study that watches the program being delivered and counts what actually happened on the ground. It is the difference between a doctor concluding a drug does not work and a doctor checking whether the patient ever took it. Without that check, a null result is uninterpretable: you cannot tell a bad treatment from a treatment that was never properly given.
History
Process evaluation has roots in the health-promotion and public-health evaluation traditions of the 1980s and 1990s, where researchers recognised that null trial results were uninterpretable without knowing whether interventions had been delivered. Implementation-science concepts such as fidelity, dose and reach were formalised in frameworks by Steckler, Linnan, Baranowski and others, and Saunders, Evans and Joshi's 2005 how-to guide provided a practical planning template. The UK Medical Research Council's 2015 guidance, led by Graham Moore and colleagues, synthesised these strands into authoritative guidance for the process evaluation of complex interventions, embedding it in the standard evaluation toolkit.
When to use it
Use process evaluation alongside any outcome or impact evaluation of a complex intervention, particularly when delivery is likely to vary across sites or providers, when an intervention is being piloted or scaled, or when understanding why an effect occurred is as important as estimating its size. It is essential for interpreting null or surprising outcome results and for deciding whether a program is ready for wider roll-out. It is less central for very simple, automatically delivered interventions where fidelity is not in question. It is most powerful when planned from the outset and embedded in the same study as the outcome evaluation.
Limitations and considerations
- Describes implementation rather than establishing the program's causal impact, so it complements rather than replaces an outcome evaluation.
- Comprehensive fidelity, dose and reach measurement is data-intensive and can burden providers and participants.
- Self-reported delivery and attendance data are prone to social-desirability and recall bias, requiring validation against observation.
- Linking process to outcomes is often correlational across a modest number of sites, limiting strong causal claims about which components drove results.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Makes outcome results interpretable by revealing whether the intervention was actually delivered and received as intended.
- Distinguishes implementation failure from theory failure, preventing sound programs from being abandoned because of poor delivery.
- Explains heterogeneity in effects across sites and subgroups by documenting variation in fidelity, dose, reach and context.
- Provides actionable feedback for improving and scaling a program, identifying which components and delivery conditions matter.
Applications
- Embedded within randomised trials of complex health, education and social interventions to explain trial results.
- Pilot and feasibility studies assessing whether an intervention can be delivered as intended before a full evaluation.
- Scale-up and dissemination research examining how fidelity and reach change as a program moves from pilot to system-wide delivery.
- Quality improvement and implementation-science studies in healthcare, schools and community settings.
- Public-policy programs where delivery is decentralised across many local agencies with variable capacity.
How it works
How it works
The evaluation begins by clarifying the intervention's intended form and its theory of change — the activities prescribed, the causal mechanisms they are meant to trigger, and the outcomes expected. This specification of what was supposed to happen provides the benchmark against which actual delivery is later compared.
Key process dimensions are defined, drawing on established frameworks: fidelity (was it delivered as designed?), dose delivered and dose received, reach (which of the intended population was exposed?), recruitment, context and, in the MRC framing, the mechanisms of impact. Each dimension is operationalised into specific, measurable questions.
Mixed-methods data are collected to capture implementation: attendance and delivery logs, structured observations of sessions, checklists and fidelity ratings, surveys of participants and providers, and interviews that probe how and why delivery varied across sites. Data are gathered throughout delivery, not only at the end, so problems can be detected as they arise.
The collected data are analysed to quantify and characterise how faithfully and fully the intervention was delivered and received, and how this varied across sites, providers and subgroups. This reveals implementation strengths and shortfalls — for example, high fidelity but low reach, or adequate dose delivered but poor dose received.
Finally, the process findings are related to outcome results. Strong outcomes with weak implementation, or null outcomes with high fidelity, sharpen the interpretation: they help distinguish theory failure from implementation failure, identify which mechanisms operated, and explain heterogeneity in effects across contexts — informing whether and how the program should be scaled or revised.
Common pitfalls
- Treating process evaluation as an afterthought added once outcomes disappoint, rather than planning measures before delivery begins.
- Measuring only dose delivered while ignoring dose received and reach, so a program can look well implemented yet never reach its target population.
- Failing to capture adaptations made in the field, which obscures the difference between intended and actual intervention and confuses fidelity assessment.
- Collecting rich process data but never relating them to outcome findings, which wastes the explanatory value of the evaluation.
Frequently asked questions
How does process evaluation differ from impact or outcome evaluation?
Outcome and impact evaluation ask whether the program produced its intended effects and how large they were; process evaluation asks how the program was delivered and received. The two are complementary and ideally run together: process data tell you whether a null impact reflects a genuinely ineffective program or one that was never properly implemented, and they explain why effects varied across sites. Without process evaluation, an outcome result is a number without a story.
What are fidelity, dose and reach?
These are core dimensions of implementation. Fidelity is the degree to which the intervention was delivered as designed. Dose has two senses: dose delivered (how much of the intervention providers actually offered) and dose received (how much participants actually engaged with). Reach is the proportion and representativeness of the intended population that was exposed. Together they describe whether enough of the right intervention reached enough of the right people in the right way for an effect to be plausible.
Should process evaluation allow for adaptation, or insist on strict fidelity?
Modern guidance, including the MRC framework, treats this as a balance rather than a contradiction. Fidelity to a program's essential functions and active ingredients is important, but local adaptation of surface form is often necessary and even beneficial for fit. The key is to document adaptations systematically, so the evaluation can distinguish principled adaptation that preserves the mechanism from drift that undermines it, and can interpret outcomes accordingly.
Related methods
Sources
Moore, G. F., Audrey, S., Barker, M., Bond, L., Bonell, C., Hardeman, W., et al. (2015). Process evaluation of complex interventions: Medical Research Council guidance. BMJ, 350, h1258.
Open DOI referenceCopied.Saunders, R. P., Evans, M. H., & Joshi, P. (2005). Developing a process-evaluation plan for assessing health promotion program implementation: A how-to guide. Health Promotion Practice, 6(2), 134–147.
Open DOI referenceCopied.