---
name: independent-methodology-design
description: Use when designing or assessing a research methodology that must answer a specified question under practical constraints.
---

# Design a methodology

Extract the research question, intended inference, available data or participants, resources, and constraints. Separate supplied facts from assumptions, and identify any information whose absence changes the design. Give a provisional design where reasonable and ask for the critical missing details.

Choose and justify a design appropriate to the question and feasible setting. Describe the units of analysis, eligibility or sampling, comparison or controls where applicable, operational definitions, procedures, instruments or sources, and planned analysis. For quantitative work specify variables, estimands, missing data handling, assumptions, uncertainty, and a defensible approach to sample size or precision; do not fabricate numerical inputs. For qualitative work describe recruitment, data collection, analytic approach, reflexivity, and the basis for stopping recruitment. For computational work specify inputs, implementation, verification, convergence or sensitivity checks, and reproducibility.

Identify plausible confounding, selection and measurement biases, alternatives, and safeguards. Match causal claims to the design and distinguish exploratory analyses from prespecified ones. Describe consent, privacy, data management, safety, and institutional review requirements as relevant, but do not claim approval or legal compliance without evidence. Preserve citations and use verified sources for consequential technical or ethical claims.

Return a usable protocol outline, rationale for major choices, required resources, limitations, and open decisions. Check current institutional or regulatory guidance when required and available. A proposed protocol is not a completed experiment, validated instrument, approved study, or analysis result.