evaluation-question-designlisted

# Evaluation Question Design for Development Programmes Evaluation questions are the contract between the people who commission an evaluation and the people who conduct it. Everything downstream (design, sampling, instruments, budget, report structure, the management response) is determined by them. Weak questions cannot be rescued by strong methods. In ToRs from UN agencies, multilaterals and NGOs, the usual problem is not missing questions but too many of them. Thirty to sixty "to what extent" questions get copied from a DAC template, none of them tied to a decision, a theory of change or a budget. The evaluation team then answers all of them thinly and none of them credibly. This skill exists to produce **a small number of real evaluative questions, each tied to a user and a decision, each answerable with the time, money, data and access available, and each with an explicit standard for what "good" looks like.** Three principles run through everything below: 1. **Utility first.** A question with no identifiable user or decision gets cut (Patton, *Utilization-Focused Evaluation*; UNEG Norm 2 on utility). 2. **Evaluative, not merely descriptive.** An evaluation question asks about *merit, worth or significance*, so it needs a standard to judge against (Scriven, 1991; Davidson, 2005). 3. **Theory-anchored and criteria-organised.** Questions come from the programme's theory of change and are then organised under the governing criteria framework (OECD-DAC 2019 by default).