Four ideas worth keeping separate.
Inputs are the resources contributed: money, time, equipment or expertise. Outputs are the activities or services delivered. Outcomes are changes in people’s circumstances, capabilities or experience. Impact concerns the wider or longer-term effects to which the work contributes.
The links between them are assumptions to investigate. An activity can be well delivered without producing the intended change. An outcome may be real while several other factors also contributed to it. Neither possibility makes evaluation pointless; both make careful questions important.
A FICTIONAL WORKED EXAMPLE / NOT A REAL PROGRAMME
A community learning room.
Imagine a local group wants to make it easier for adults to use everyday digital services. The idea begins with conversations about access, confidence, language and the tasks people actually want to complete. Participants help shape the offer and can choose whether to share feedback.
INPUTSA borrowed room, volunteer time and a small equipment budget. The example deliberately sets no real funding or performance target.
OUTPUTSDrop-in sessions take place, devices are available and people receive help with tasks they choose. Attendance records describe delivery; they do not by themselves demonstrate benefit.
OUTCOMESSome participants may become more confident or complete a chosen task independently. Ask them, observe with consent where appropriate and check whether the change lasts beyond the session.
WIDER EFFECTSImproved access to opportunities is a possible longer-term contribution. It would require follow-up and attention to other influences, not simply multiplication of attendance figures.
Before claiming that support caused a change, ask what might have happened without it. Participants may also have learned from family, another service or their own practice. There may be people the room did not reach because of transport, disability, cost or trust. Record these limitations and use them to improve the offer.
Choose evidence for the decision.
If the immediate question is whether session times are suitable, a short conversation may be more useful than a complex survey. If the question is whether a costly programme should expand, a stronger design and relevant specialist evaluation may be needed. Decide what evidence is credible enough for the decision and feasible to collect.
Use quantitative and qualitative information together where helpful. Counts can reveal patterns; lived experience can reveal why those patterns matter or what the categories missed. Neither should be used to erase inconvenient findings from the other.
Make room for negative results.
Ask about unintended effects, who did not benefit and whether the activity displaced something valuable. Report an outcome that did not improve. Distinguish a weak idea from poor delivery, an unsuitable measure or a change in context before deciding what to do next.
Protect privacy and avoid collecting personal data “just in case”. Explain why information is requested, how it will be used and what choice people have. Do not publish a story or small-group dataset that could identify someone without the appropriate basis and permissions.
Further reading: evaluation lenses.
The OECD DAC evaluation criteria offer a structured way to examine relevance, coherence, effectiveness, efficiency, impact and sustainability. Select questions appropriate to the context; a framework should support judgement rather than replace it.
Explore the OECD evaluation criteria ↗