Friday afternoon, four o'clock: the project has just been delivered, and the project manager still needs to get the evaluation report out the door before the weekend — what ends up on paper is usually a retrospective without a fixed structure, an account of what happened without a systematic answer to why it went that way, with what input, and whether it delivered what it was supposed to deliver. The CIPP model by Daniel Stufflebeam offers that structure after the fact: four questions that can be asked just as well in retrospect as in advance.
Four letters, four questions
CIPP stands for Context, Input, Process and Product: four evaluation moments that complement each other rather than overlap.
Context asks: what was the situation in which the project started, and what need underpinned it? This is comparable to the OECD relevance question, but applied to the starting point.
Input asks: which resources, strategies and plans were deployed to address that need, and were those choices sound given the alternatives? This evaluation moment normally precedes execution, but also belongs in an ex-post retrospective: was the chosen approach, in hindsight, the right one?
Process looks at the execution itself: was the plan carried out as intended, and what deviations occurred along the way?
Product, the component that carries the most weight for ex-post evaluation, is further broken down by Stufflebeam into four sub-questions: impact (what broader effects are there), effectiveness (were the goals achieved), sustainability (do the results persist) and transportability (is the approach transferable to other situations or projects). For that last component, the model prescribes a fixed set of assessment criteria: quality, cost-effectiveness, probity, feasibility, safety, equity and significance.
The requirement most evaluations skip: the meta-evaluation
The most distinctive element of the CIPP model is not the four-part structure itself, but a requirement that is hard to separate from it: Stufflebeam explicitly prescribes that the evaluation itself be tested against meta-evaluation standards: utility, feasibility, propriety, accuracy and evaluator accountability. In other words: it is not only the project that gets evaluated, the evaluation itself is also tested for quality. Stufflebeam explicitly recommends that the client commission an independent meta-evaluation, while the evaluator who does the actual work must remain independent in forming their own judgement.
That is a double layer of independence rarely applied in practice, yet it is the sharpest normative argument for a principle that recurs in virtually every authoritative framework, from Gate 5 to the RPE 2022: an evaluation written solely by the project team itself, without any form of challenge or external perspective, is by definition weaker than one that has been tested.
An explicit requirement for lessons, too
Alongside the meta-evaluation requirement, the CIPP model prescribes that final evaluation reports explicitly name lessons, including mistakes to avoid. This connects directly to what we discuss in another article about the difference between documenting and learning: the CIPP model does not just require that lessons be written down, but that they be concrete enough to actually name mistakes to avoid: not a vague observation, but a specific, traceable lesson.
Why this is more than an academic model
The CIPP model comes from the core methodological literature on evaluation research, not from project management practice itself. That is exactly its value: where OECD DAC, Gate 5 and PRINCE2 practice were built mainly from practice, CIPP provides the methodological foundation that explains why independence, pre-set criteria and action-oriented lessons are not coincidentally recurring themes. They are methodologically anchored in how evaluation research as a field is itself constructed.
How this lands in the EvaluatieScore rubric
The CIPP model is one of four normative anchors behind the EvaluatieScore rubric, and the strongest evidence for independence as an explicit quality criterion: category 4 of the rubric (methodological quality & independence) directly tests whether an evaluation method has been named, whether a demonstrable independent or external check has taken place, and whether the evaluation's sources and data are transparent.
Put category 4 to the test against your own report: upload it for €15 at dutchmind.com/producten/evaluatiescore and see whether the independent check Stufflebeam prescribes is also reflected in your evaluation.
Sources
- Stufflebeam, D.L. (2015), CIPP Model checklist (2e ed.) — rszarf.ips.uw.edu.pl.
- HM Government / Cabinet Office–IPA (nu NISTA) (2021), Gate 5: Operations Review and Benefits Realisation (Assurance Portfolio Standard, V1.0) — assets.publishing.service.gov.uk.
- Regeling periodiek evaluatieonderzoek 2022 (RPE) — wetten.overheid.nl.