Theory of Change Example: Girls' Education, From Diagram to Database
Most theory of change examples you’ll find online stop at a diagram. This one keeps going: from the outcomes pathway to the exact indicators, the records a system must hold, and the forms field staff fill in. It is the worked example inside our free theory of change template, written out in full so you can adapt it for any education programme. The programme, “Stay & Learn”, is fictional, but every design choice comes from systems we’ve built for education nonprofits.
The programme in one paragraph
Stay & Learn works with girls aged 11 to 16 in Grades 6 to 10 across 40 rural government schools in one district. Girls here tend to leave around the move from Grade 8 to Grade 9, often well below grade-level in reading and maths. The programme runs after-school learning camps grouped by level, holds caregiver meetings each term, and sends a mentor to any girl who misses five school days in a row. Some causes are outside its control (distance to secondary schools, poverty, early-marriage norms), and a good theory of change says so.
Step 1: The goal and the outcomes pathway, built backwards
Long-term goal (G1): Girls complete Grade 10 with the reading and maths skills to choose their next step in education or work. This is shared with schools and families, so the programme claims contribution, not sole credit.
Working backwards, asking “what must be true first?” each time:
| ID | Outcome | Term | Leads to |
|---|---|---|---|
| LO1 | Girls complete Grade 10 at or near grade-level | Long | G1 |
| IO1 | Girls move from Grade 8 to Grade 9 | Intermediate | LO1 |
| SO1 | Girls attend school and camps regularly | Short | SO2, IO1 |
| SO2 | Girls' reading and maths levels improve | Short | IO1, LO1 |
| SO3 | Caregivers support girls continuing to secondary school | Short | IO1 |
Only then do activities come in: learning camps (AC1), caregiver meetings (AC2), home visits after five consecutive absences (AC3), and three assessment rounds a year (AC4). Their outputs are sessions held as planned (OP1), on-time home visits (OP2) and termly caregiver meetings (OP3).
Step 2: The assumptions worth testing
Three assumptions carry most of the weight:
- AS1 (high risk): Home visits change attendance, because absence is mostly about things a family can act on. If girls are absent because of distance or work, visits won’t help.
- AS2 (medium): Girls whose learning improves are more likely to move to Grade 9.
- AS3 (high): Caregiver support is enough when a secondary school is within reach. Where it isn’t, support alone may not change transition.
Each one shapes the data model. AS1 is why the home-visit form records a reason code and the system tracks attendance for four weeks after each visit. AS3 is why every school record carries a distance band. Without those fields, the programme could never learn whether its theory holds.
Step 3: Indicators with exact definitions
Here are six of the ten indicators in the template. Each has a numerator and denominator so two people reach the same number from the same records.
| ID | Indicator | Definition | Frequency | Source |
|---|---|---|---|---|
| I2 | Girls attending ≥75% of camp sessions | Girls with ≥75% attendance in sessions held ÷ girls active on the 1st of the month | Monthly | Session attendance form |
| I3 | Girls who moved up ≥1 level | Girls whose level rose since the last round ÷ girls assessed in both rounds | Each round | Assessment entry |
| I4 | Caregiver support for Grade 9 | Caregivers answering "Yes, definitely" ÷ caregivers surveyed | Twice a year | Caregiver survey |
| I5 | Grade 8 → 9 transition | Girls in Grade 8 at year end enrolled in Grade 9 by the end of admissions ÷ girls in Grade 8 at year end | Annually | Annual status check |
| I8 | Home visits on time | Visits within 7 days of the 5th consecutive absence ÷ girls who reached 5 consecutive absences | Weekly | Home-visit form |
| I10 | Attendance after a visit | Girls with ≥75% attendance in the 4 weeks after a visit ÷ girls visited | Monthly | Home-visit form + attendance |
Notice that I10 measures an assumption, not an outcome. That is deliberate. If you’d like the full method for writing these, read our guide to writing SMART indicators from outcomes.
Step 4: From indicators to entities and forms
You don’t build a “transition” table. You build records for the things that happen, and calculate transition from them.
| Entity | One record per | Why it exists |
|---|---|---|
| Girl | girl, with an ID that survives school changes | Every indicator |
| Caregiver | caregiver | I4 |
| School | school, with distance band | I5, AS3 |
| Session | camp session scheduled or held | I2, I7 |
| Attendance | girl × session | I2 |
| Assessment result | girl × round × subject, level calculated from bands | I3 |
| Home visit | visit, with reason code | I8, I10, AS1 |
| Annual status | girl × academic year | I5, I6 |
Seven forms feed these records, all filled by mentors, mostly offline. Levels, level changes, consecutive absences and the transition flag are calculated; nobody types them. The form most programmes forget is the annual status check. It is the only way to know a girl reached Grade 9, and it must happen after admissions close, verified against the school register.
Step 5: Who uses what
- Mentor, weekly: which girls have missed five days, or are slipping (I1, I2, I8).
- Coordinator, monthly: which centres are missing sessions or not moving levels (I2, I3, I7).
- Programme lead, each term: which villages need caregiver work before admissions (I4, I5).
- Director and funders, yearly: is transition rising, and for whom (I5, I6)?
Because girls’ records include disability status (Washington Group questions), each view can be split to check no one is left behind. Decide first who can see that data and why; our DPDP compliance checklist covers consent and access.
Adapting this example
Swap the outcomes, keep the structure. A skilling programme might replace “Grade 8 → 9 transition” with “in a job and still employed at 3 months”, which needs the same kind of follow-up record. The editable version is inside the free template on the theory of change template and data model guide.
Your next step: see whether your current tools could produce these indicators every month. Take the free 10-minute technology assessment.