Guide
Baseline and endline surveys
What they are, how they measure what a programme changed, and how to design and run them well.
What they are
A baseline survey measures the situation before a programme starts; an endline survey measures it again after the programme ends, so the two can be compared to show what changed. The baseline sets the reference point for every indicator - for example, the share of households with a reliable water source - and the endline re-measures the same indicators, with the same method, at the end.
Together they turn a claim like "the programme improved access to water" into a measured number: the difference between where things started and where they finished. They are the backbone of most MEAL systems and of the independent evaluations donors expect.
How they measure change
The measured change is the gap between the baseline and the endline - but only part of that gap is the programme's doing. Prices, seasons, other actors and the wider situation all move indicators too. A good evaluation accounts for what would likely have happened anyway (the counterfactual), so the programme is credited with its real contribution rather than everything that changed around it.
That is why method discipline matters: the endline has to measure the same indicators, in the same population, the same way as the baseline. Change the questions or the sampling between the two rounds and the comparison quietly stops meaning anything.
Designing them well
Five decisions made at the baseline determine whether the endline can answer the question at all.
-
Indicators first
Lock the indicators - tied to the results framework - before designing the questionnaire, not the other way round.
-
A sample you can repeat
Choose a sampling approach the endline can reproduce, sized for confidence you can actually act on.
-
Right timing
Run the baseline before activities reach people, and the endline once effects have had time to show - not the day the funding closes.
-
Comparability
Keep questions, definitions and mode identical across rounds, so any difference reflects real change, not a change in measurement.
-
Access & ethics
Plan for limited access and safe, dignified data collection - especially where the endline population may have moved.
Getting it right
Most weak endlines fail for a reason set at the baseline: shifting indicators, an un-repeatable sample, or a questionnaire that quietly changed. Designing both rounds together - ideally with the same team - avoids that. We design and quality-assure baseline and endline studies for programmes across Ukraine; see our services, our guide to choosing an M&E consultant, or get in touch.