Author: Roy Aulie Jacobsen, PhD research fellow at Kristiania University of Applied Sciences.
Roy Aulie Jacobsen explores why descriptive analysis should never be seen as second best. He argues that rigorous description provides the evidence base needed to make credible claims, avoid overstating impact and support better evaluation.
Description is not second best
In evaluations, descriptive analysis is often treated as the method you settle for when a causal design is out of reach. Descriptive work can sound like an inventory of activities, participants, outputs, and perceptions: useful, but not quite serious enough.
For many evaluators, that hierarchy does not fit the work. Descriptive analysis is often the evidence base that the commission, timetable, and available data permit. The task is not to make description do causal work. It cannot. The task is to make descriptive evaluations rigorous, useful, and clear about what the evidence can, and cannot, support.
The problem often begins with a gap between the question commissioners want answered and the data available to answer it. Evaluations are often commissioned around a large question. Did this programme work? Did outcomes change because of the programme, or because of wider conditions? Those are impact questions. They require some basis for comparison, whether through a baseline, comparison group, counterfactual reasoning, or theory-led design.
In practice, evaluators are often appointed after the most important design decisions have been made. There may be no baseline, no comparison group, and no reliable outcome data. Monitoring forms may capture attendance, satisfaction, and activities delivered, but say less about mechanism or longer-term change. The commissioner wants to know whether the programme had an impact, but the available evidence may only support an account of how the programme was delivered and what patterns were observed.
What good descriptive evaluation can tell us
This mismatch puts evaluators under pressure and can pull the report in two unhelpful directions. Observed change may be presented as impact, while descriptive findings may be treated as if they have little value. Both weaken evaluation practice.
Good descriptive evaluation answers basic but consequential questions. Who was reached, and who was not? What was delivered, and with what variation? Did participants receive enough of the intervention for change to be plausible? Where did delivery depart from the original plan?
These questions may look basic, but they are analytically important. They sit at the heart of process evaluation, the strand of work that the UK Government’s Magenta Book treats as central to policy learning. Before asking whether a programme caused an outcome, it is worth establishing what the programme actually was.
A common evaluation challenge
This situation will be familiar to many evaluators. A local authority commissions a short employability programme and appoints the evaluator near the end. The available evidence includes attendance registers, demographic data, feedback forms, and staff notes. There is no baseline, no comparison group, and no reliable follow-up on employment outcomes.
A less careful evaluation might still imply that the programme improved employability. A stronger descriptive evaluation would stop short of that claim. It could show whether the intended participants were reached, how much of the programme they received, where delivery varied, and which groups disengaged. That does not prove impact, but it gives commissioners evidence for redesign and targeting.
Why descriptive analysis matters
This is where description earns its place. As BetterEvaluation notes, analysing data to summarise it and look for patterns is part of every evaluation. Counts, proportions, distributions, medians, ranges, and subgroup patterns help evaluators establish what happened before moving to stronger claims about why it happened or what it caused.
The practical safeguard is a discipline of claim-making. A previous UK Evaluation Society blog on evaluation briefs argued that evaluators should ask what kind of answer an evaluation can realistically support before work begins. The same discipline is needed at reporting stage, where three moves often collapse into one.
First, the observation. What does the data show? “Sixty-eight per cent of participants completed at least three sessions.”
Second, the interpretation. Why might this matter? “This suggests moderate retention, but uneven exposure to the full intervention.”
Third, the boundary. How far can the claim go? “The data cannot show whether completion improved employment outcomes without a comparison or stronger outcome measures.”
This structure links back to the three-sentence stress test. It does not replace better design. Evaluators should still argue early for clearer outcomes, stronger baselines, feasible comparison strategies, and monitoring that captures mechanism as well as activity.
But where those conditions are absent, rigorous description remains essential. It prevents inflated claims, protects commissioners from false certainty, and creates a stronger foundation for future learning. The point is not to lower evaluation ambition. It is to make the first layer of evidence strong enough to carry the next one.
Descriptive analysis is not second best. Used well, it helps evaluators say, with confidence, what the evidence shows, what it suggests, and how far the claim can reasonably go.
Building stronger evaluations
This structure links back to the three-sentence stress test. It does not replace better design. Evaluators should still argue early for clearer outcomes, stronger baselines, feasible comparison strategies, and monitoring that captures mechanism as well as activity.
But where those conditions are absent, rigorous description remains essential. It prevents inflated claims, protects commissioners from false certainty, and creates a stronger foundation for future learning. The point is not to lower evaluation ambition. It is to make the first layer of evidence strong enough to carry the next one.
Descriptive analysis is not second best. Used well, it helps evaluators say, with confidence, what the evidence shows, what it suggests, and how far the claim can reasonably go.
Our thanks to Roy Aulie Jacobsen for this guest blog and for contributing their perspective to the UK Evaluation Society’s work.
How to cite this article
Aulie Jacobsen, J (2026) Descriptive analysis in evaluation: getting the evidence base right before claiming impact. UK Evaluation Society blog. Available at: https://evaluation.org.uk/blog-descriptive-analysis-in-evaluation/ (Accessed: [insert date]).
Roy Aulie Jacobsen
PhD Fellow in Political and Crisis Communication | YECREA Representative – ECREA Mediatization & Society Section | Data-Driven Insights | Strategic Messaging & Framing
Roy Aulie Jacobsen is a PhD research fellow working at the intersection of political communication, political science, and sociology, with a primary focus on strategic messaging from political leaders and their crisis response. His work centres on understanding how political actors communicate under pressure—whether during elections, public health emergencies, or high-stakes policy reforms.
