Insights
Research-to-practice notes. Each one states the problem, the evidence, its limits, and how it could be tested locally.
Article
A feasibility test is not a small experiment
A cohort of twelve to twenty people can determine whether an intervention can be delivered, understood, and reliably measured. It cannot establish whether the intervention works. Conflating the two is precisely how public programmes end up scaling on statistical noise.
Research-to-practice note
Article
What a cost-benefit analysis can tell a funder, and what it cannot
Economic cost-benefit analysis, financial analysis, and cost-effectiveness analysis answer fundamentally different questions. Switching values provide far better decision support than a single point-estimate NPV, and three common methodological pitfalls explain why appraisals so often mislead.
Research-to-practice note