Funders increasingly ask a simple question: what difference did your programme make? Answering it convincingly requires measuring conditions before, during and after the programme. That is what baseline, midline and endline surveys are for.
This guide explains each survey, when to run it, how to design it so the results can be compared, and the mistakes that most often weaken programme evidence in Ghana.
The three surveys at a glance
| Survey | When | Main purpose |
|---|---|---|
| Baseline | Before activities start, or as early as possible | Record the starting point for each indicator and set realistic targets |
| Midline | Roughly halfway through | Check progress and adjust the programme while there is still time |
| Endline | At or near the end | Measure change against the baseline and report results |
Baseline: setting the starting point
A baseline survey measures the situation your programme intends to change. If the goal is to raise household income among smallholder farmers in a district, the baseline records current income, along with the factors you expect the programme to influence, such as yields, access to inputs or market prices.
A good baseline does three things:
- Gives a value for every indicator in your results framework.
- Tests your assumptions. Sometimes the baseline shows the problem is different from what the proposal assumed, and the design can be adjusted early.
- Allows realistic targets. Targets set without a baseline are guesses.
If a programme has already started without a baseline, collect one as soon as possible and document the timing. Some organisations reconstruct a baseline from records or recall questions, but this is weaker evidence and should be described honestly in reports.
Midline: learning while you can still act
The midline is often skipped to save money, but it is where learning pays off. It tells you whether activities are reaching the right people, whether early outcomes are moving, and whether anything needs to change. A midline that finds a problem in year two of a four-year programme is far more valuable than an endline that finds the same problem when the money has been spent.
Endline: measuring the change
The endline repeats the baseline measurement so you can compare. The comparison is only valid if the two surveys are consistent. That means:
- The same indicators, defined the same way.
- The same or equivalent questions, asked in the same way.
- A comparable sample from the same population.
- Fieldwork at a similar time of year, where seasons affect the indicator. Agricultural income, school attendance and some health outcomes vary strongly by season in Ghana.
Before and after is not the same as impact
A change between baseline and endline shows that conditions improved. It does not prove the programme caused the improvement. Other factors, such as a good harvest, new government policies or another NGO working in the same area, may have contributed.
Stronger designs compare programme participants with a similar group that did not take part. Where a full comparison group is not possible, you can still strengthen the evidence by:
- Asking participants what difference the programme made and why, through interviews or focus groups.
- Tracking a logical chain of results, from activities to outputs to outcomes.
- Documenting other events that could have affected the results.
Be clear in your report about what the design can and cannot show. Funders respect honesty about limitations more than inflated claims.
Designing the survey: a practical checklist
- Start from the results framework. Every indicator needs a clear definition, a data source and a method of calculation.
- Calculate the sample size based on the change you expect to detect. Detecting a small change needs a larger sample.
- Plan for comparison groups if you want to compare districts, genders or other groups, since each needs enough respondents.
- Translate and test the questionnaire in the local languages respondents will use, such as Twi, Ewe, Ga, Dagbani or others, and pilot it in a community similar to the programme area.
- Train enumerators on the questionnaire, consent and how to ask sensitive questions.
- Use digital data collection tools such as KoboToolbox or ODK to reduce errors and allow quality checks during fieldwork.
- Store the baseline instrument and data carefully so the endline team can replicate it exactly, even if staff change.
Ethics and data protection
Programme surveys often collect sensitive information about income, health, family and community relationships. Respondents should understand the purpose of the survey, agree to take part, and know they can decline any question. Personal data must be protected under Ghana’s Data Protection Act, 2012 (Act 843), and reports should not identify individuals without consent.
Common mistakes
- Changing indicators halfway. If the definition changes, the baseline and endline cannot be compared.
- Collecting too much. Long questionnaires tire respondents and reduce data quality. Measure what your framework needs.
- Losing the baseline. Data, questionnaires and sampling lists go missing when staff leave. Archive them.
- Different seasons. A baseline in the lean season and an endline after harvest will exaggerate change.
- Reporting averages only. Results for women, youth or remote communities can differ sharply from the overall average.
Frequently asked questions
Should the same organisation run the baseline and endline?
It helps continuity, but it is not essential if the baseline is well documented. Many funders prefer an independent firm for the endline evaluation.
Do we need to interview the same households?
Interviewing the same households (a panel) gives powerful evidence of change, but people move and drop out. A fresh, comparable sample is a common and valid alternative.
How large should the sample be?
It depends on the size of change you want to detect, the number of groups you will compare and how variable the indicator is. An experienced M&E consultant can calculate it for your programme.
Get your programme evidence right from the start
Pinnacle Gate’s monitoring and evaluation consultants design and run baseline, midline and endline studies for NGOs, foundations and government programmes across Ghana. Request a proposal to discuss your programme.
