GEDSI indicators: a practical starter set
Oct 12 2026 · Templates
GEDSI indicators donors expect, with a means of verification for each, plus how to disaggregate core indicators so gaps stay visible.
GEDSI indicators measure whether a programme reaches and benefits women, people with disabilities and other marginalised groups, and whether it changes anything for them. A practical starter set covers reach, voice, access, resourcing and power: the share of participants who are women or have a disability; the share of women reporting influence over decisions; the share of activities held in accessible venues; the share of budget spent on inclusion activities; and the share of leadership roles held by marginalised groups. Each needs a means of verification, such as attendance registers disaggregated by sex and disability, or a short perception survey. The bigger discipline is disaggregating every core indicator rather than adding GEDSI as a separate annex.
What are GEDSI indicators?
GEDSI stands for gender equality, disability and social inclusion. GEDSI indicators are the measures that show whether a programme reaches women, people with disabilities and other marginalised groups, and whether it changes anything for them.
They come in two families. Disaggregated core indicators split your existing headline measures by sex, disability and other relevant characteristics. GEDSI-specific indicators measure things the core set misses, such as access, influence and budget allocation. Most programmes need both, but the disaggregation matters more.
What does a starter set of GEDSI indicators look like?
Six indicators cover reach, voice, access, resourcing and power. Adapt the wording to your context and keep the number small enough to collect well.
| Indicator | What it tells you | Means of verification |
|---|---|---|
| Share of participants who are women | Whether reach is balanced | Attendance registers disaggregated by sex |
| Share of participants with a disability | Whether disabled people are reached | Registers using disability questions consistent with the Washington Group short set |
| Share of women reporting influence over decisions | Whether participation translates into voice | Short perception survey, repeated at the same points |
| Share of activities held in accessible venues | Whether access is designed in | Venue accessibility checklist |
| Share of budget spent on inclusion activities | Whether inclusion is resourced | Budget and expenditure tracking |
| Share of leadership roles held by marginalised groups | Whether power is shared | Governance records and committee minutes |
Should you disaggregate core indicators or add GEDSI indicators?
| Approach | What it looks like | Strength | Risk |
|---|---|---|---|
| Disaggregated core indicators | Every headline indicator split by sex, disability and age | The gap is visible in the main results table | Needs clean collection at source |
| GEDSI-specific indicators | A separate set on access, influence and budget | Captures issues core indicators miss | Can sit in an annex and be ignored |
| Both, kept small | Disaggregate everything, add three to five GEDSI-specific indicators | Gaps stay visible and specific issues get measured | More data to manage and clean |
The practical answer is both, with discipline about volume. Disaggregate every core indicator, then add a short GEDSI-specific set that answers questions disaggregation alone cannot.
How do you disaggregate by sex and disability?
Disaggregation is a data collection decision, not a reporting decision. If the categories are not captured at registration, no amount of analysis will recover them later.
- Collect sex and disability data at enrolment, not retrospectively.
- Use a consistent disability question set rather than a single yes or no.
- Offer a prefer not to say option and record refusals separately.
- Train enumerators and data entry staff on why the categories matter.
- Report not known as its own category instead of dropping it.
- Check disaggregated counts before the report is due, not during it.
How do you write a GEDSI indicator that is measurable?
A weak indicator names a group but not a change. Use this checklist before the indicator goes into the logframe.
- Name the group precisely, for example women aged 18 to 24, or people with a disability.
- State the direction of change, such as increased, reduced or maintained.
- Define the denominator so the percentage means the same thing each round.
- Set a baseline and a target, even if the baseline is zero.
- Name the means of verification in the same row.
- Say who collects the data and how often.
- Decide the disaggregation before the first data collection round.
What means of verification work in practice?
Use sources you already have where possible, and add only what you will genuinely analyse.
- Attendance and enrolment registers, disaggregated by sex, age and disability.
- A short perception survey on influence and decision-making, repeated at the same points.
- A venue accessibility checklist completed before each event.
- Budget lines and expenditure records for inclusion activities.
- Committee minutes and governance records for leadership roles.
- Exit interviews or feedback channels for qualitative context.
Three indicators collected well beat ten collected badly.
What are the common mistakes?
- Bolting GEDSI on at reporting time, when the data was never collected.
- Treating the number of women reached as evidence of inclusion.
- No baseline, so no way to show change.
- Data collection that is itself inaccessible, such as online-only surveys.
- Disability recorded as a single yes or no with no consistent question.
- Indicators with no named owner and no collection schedule.
- Separate GEDSI annexes that never reach the main results table.
How does GEDSI fit into the logframe?
Put disaggregation in the indicator wording and in the means of verification column, so it is part of the main table rather than an annex. Add a small number of GEDSI-specific indicators at output or outcome level where the core set cannot capture access, influence or resourcing.
If you are rebuilding the logframe, start from a template that already has disaggregation columns.
What should you do next?
Pick three to five indicators, write the means of verification next to each, and check that your registration form captures sex and disability before the next round of data collection. Review the set at the next reporting point and drop anything you are not using.
Written for M&E professionals. Not a guarantee of accuracy — always verify every claim against its source, and let a named human sign off. Get the free handbook.