Unit
- Research question anchoring: every policy research project begins with a precisely worded question that identifies the population affected, the intervention or policy variable under study, and the outcome of interest, since vague questions produce unfocused data collection and inconclusive findings that decision makers cannot act on.
- Positive versus normative framing: positive questions ask what is happening or what would happen under a policy change, using observable evidence, while normative questions ask what should happen and require value judgments; conflating the two weakens the credibility of a research brief presented to a legislature or ministry.
- Policy relevance test: a well-framed question must connect to an actual decision point, such as a budget allocation, a regulatory threshold, or a program renewal deadline, because research divorced from a live decision rarely influences outcomes regardless of its analytical rigor.
- Scoping the unit of analysis: researchers must decide whether the analysis operates at the individual, household, firm, regional, or national level, since the choice determines which datasets are usable and what causal claims are defensible given the data structure.
- Feasibility screening: before committing resources, analysts should assess whether the required data exists, whether the timeline matches the policy cycle, and whether ethical or legal constraints limit access to sensitive populations or administrative records.
- Literature triangulation: reviewing prior studies, government reports, and multilateral agency publications helps identify what is already known, preventing duplicated effort and revealing where genuine analytical gaps remain unaddressed by existing research.
- Stakeholder consultation in question design: engaging with the officials or program managers who will use the findings early in the process ensures the question reflects real decision needs rather than purely academic curiosity.
- Hypothesis articulation: once the question is set, researchers state a testable hypothesis, such as an expected direction and rough magnitude of effect, which guides the choice of data and method and provides a benchmark against which findings are later compared.
- Iterative refinement: initial research questions are rarely final; as data availability and institutional constraints become clear, the question is narrowed or reframed, and documenting this evolution protects the credibility of the final research design.
Strategic Framework
Framework: PICO for Policy Research (Population, Intervention, Comparison, Outcome)
Purpose: structures a policy research question into testable components so that data requirements and analytical methods can be identified early.
- Population: the group affected by or eligible for the policy, defined with clear boundaries such as age, geography, or income bracket.
- Intervention: the specific policy, program, or regulatory change being studied, described precisely enough to distinguish it from related measures.
- Comparison: the counterfactual condition, such as a prior policy regime, a control group, or a comparable jurisdiction without the intervention.
- Outcome: the measurable indicator used to judge success or failure, chosen for both policy relevance and data availability.
Quick Revision
- A strong research question specifies population, intervention, and outcome precisely.
- Positive and normative framings serve different purposes and should not be mixed.
- Feasibility and stakeholder input must shape the question before data collection begins.
- The PICO framework provides a repeatable structure for framing policy questions.