What is a theory of change example?
A theory of change example shows how a particular intervention is expected to produce change, why those connections are plausible, and what conditions they depend on. A useful example connects activities to outcomes and makes the reasoning visible. It can include a diagram, a narrative and a measurement plan; there is no single mandatory layout.
The four examples below are illustrative models, not documented customer results. They show how the same planning questions change across workforce training, education, health services and environmental work. Start with the closest setting, then rewrite the assumptions for your own participants and partners. For definitions, read the complete theory of change guide.
How to read the examples
Read each pathway from the activity to the intended outcome. At every connection, ask why the next step would follow. Then look for a condition outside the activity itself: whether employers recognize a qualification, students attend regularly, or a service remains accessible.
- Separate what the team delivers from what changes for people.
- Treat each assumption as something to investigate, not as a disclaimer.
- Name the evidence that would support or challenge the explanation.
- Include negative and unexpected changes in the review.
BetterEvaluation recommends making the action and change theories explicit and checking them with stakeholders and existing evidence. Read its guidance on describing a theory of change.
Example 1: workforce training and sustained employment
A training provider wants to improve employment prospects for adults who have relevant experience but lack a recognized qualification. The theory must explain more than course completion: it needs a plausible route from learning to a suitable job and continued employment.
| Component | Illustrative workforce model |
|---|---|
| Problem | Participants cannot access suitable roles despite prior experience. |
| Inputs | Instructors, employer partners, practice equipment and participant support. |
| Activities | Teach job-specific skills, provide assessed practice, arrange interviews and offer follow-up support. |
| Outputs | Training sessions delivered, assessments completed and employer interviews arranged. |
| Outcomes | Participants demonstrate relevant skills, enter suitable work and remain employed. |
| Longer-term aim | More stable livelihoods and better employment choices. |
Assumptions to test: employers value the assessment, jobs are available, and transport or care responsibilities do not prevent participation. Training cannot solve a lack of local vacancies by itself.
Evidence plan: connect entry goals, skills assessments, completion records and follow-up interviews. Define “suitable work” with participants. Report employment at a named follow-up point alongside the number reached and the number whose status is unknown.
Review question: if completion is high but placement is low, is the missing connection employer demand, demonstrated skill, recruitment access or something else?
Example 2: reading support in education
A school offers small-group reading support to students who need additional instruction. The intended change is stronger reading ability, rather than simply attending more sessions.
| Pathway | What to check |
|---|---|
| Trained tutors deliver structured practice | Were sessions delivered consistently, with materials appropriate to the students? |
| Students participate and practice between sessions | Who misses sessions, and what prevents regular participation? |
| Reading accuracy and comprehension improve | Use an appropriate assessment at comparable points; check whether the task became easier. |
| Students engage more confidently in classroom learning | Gather classroom observations and student accounts, not just another attendance count. |
The main assumptions are that instruction addresses the students’ needs and that participation is sufficient for learning. A teacher’s observation can help explain an assessment change, but it does not replace a comparable assessment. Avoid claiming that every later school outcome was caused by the reading support.
Example 3: keeping people connected to a health service
A service team wants fewer people to lose contact after an initial appointment. This example concerns access and continuity of service, not a claim about a particular treatment’s clinical effectiveness.
| Component | Illustrative service model |
|---|---|
| Activities | Agree follow-up arrangements, send reminders and help people navigate access barriers. |
| Outputs | Appointments arranged, reminders sent and support contacts completed. |
| Near-term outcome | More people attend the agreed follow-up or arrange an appropriate alternative. |
| Assumptions | Contact details remain usable; the appointment is affordable and accessible; the service has capacity. |
| Evidence | Appointment records, reasons for missed contacts and permission-based follow-up feedback. |
Record why someone did not return without equating non-attendance with lack of motivation. Transport, timing, changed needs and service availability are different explanations. Review patterns at an appropriate group level and protect sensitive information.
Example 4: adoption of an environmental practice
A community initiative supports households to separate food waste. Providing containers is an output. Continued use of the collection service is a behavior change. Environmental benefits depend on what happens after collection as well.
| Connection | Assumption | Evidence to review |
|---|---|---|
| Containers and guidance → correct sorting | Households have space and understand what belongs in each container. | Participation feedback, sorting observations and contamination records. |
| Correct sorting → sustained collection | Collection is reliable and households can use it throughout the year. | Missed collections and repeated participation measures. |
| Collected material → intended environmental benefit | The processing route delivers the claimed benefit. | Verified processing records and an appropriate environmental assessment. |
A short participation survey cannot establish the final environmental benefit. Keep the outcome claim proportional to the evidence and describe the parts of the pathway that still need testing.
How to adapt an example without copying its assumptions
Begin with a short conversation with the people affected, delivery staff and key partners. Ask what is different in your location, what has already been tried and which barriers the proposed intervention cannot remove.
- Write the intended outcome in terms of a change for a defined group.
- Work backward to the conditions needed for that change.
- Choose activities your team can actually deliver or influence.
- Record the evidence behind important connections and where it is weak.
- Assign an owner and review point to the assumptions most likely to change.
Keep one versioned explanation across the diagram, narrative and measurement plan. If a partner leaves or the target population changes, update the reasoning as well as the boxes. A changed assumption is a reason to learn and adapt; it is not automatically evidence that the entire intervention failed.
Example, template, logic model and logframe: what changes?
| Format | Main purpose | What to carry across |
|---|---|---|
| Worked example | Shows how a model applies in a particular setting. | The reasoning and conditions, not only the layout. |
| Template | Provides a blank structure for your own model. | Replace sample wording with your actual context. |
| Logic model | Summarizes a sequence of resources, activities and results. | Retain a companion explanation of assumptions where needed. |
| Logframe | Organizes results, indicators, verification and assumptions for management. | Use consistent outcome definitions and evidence sources. |
These formats can complement one another. A logic model can include assumptions; it is not automatically an inferior theory of change. Follow the conventions your team or funder uses and make the underlying explanation available.
Keep the model connected to what happens next
Give each outcome and assumption a stable label. Link new surveys, interviews, delivery records and partner documents to the relevant group and period. This makes the next review about the evidence rather than reconstructing which file belonged to which cycle.
AI can help organize new narrative evidence against a review question. A reviewer still needs to check the sources, look for contradictory accounts and decide whether the explanation should change. A well-organized record supports that work; it does not establish causality on its own.
Continue learning in the Academy
Use these existing Academy guides for the practical next step. They are suggested companion readings; follow each guide’s course navigation for the full sequence.
Use the practical guides to connect the model to a collection plan, review new evidence and decide what to improve.
Frequently asked questions
Can I use these examples for my own program?
Yes, as starting points. Rewrite the population, context, outcomes and assumptions, and check the evidence for each important connection. The examples are illustrative rather than validated models for your setting.
Does every theory of change need six boxes?
No. Six components are a useful teaching structure, but complex interventions may need multiple pathways, feedback between stages and a narrative explanation.
What makes an example testable?
It identifies the intended changes, explains important causal connections, states assumptions and names evidence that could support or challenge those connections.
Is a theory of change the same as a logic model?
They overlap. A logic model usually provides a compact results sequence; a theory of change makes the explanation and context explicit. Either can include assumptions and evidence.
How often should we revise the model?
Use a planned review cycle and revisit it when important evidence or operating conditions change. Keep a record of what changed and why.
