Research questions, methods, ethics and governance

• Recognise when a project is research under the UK health and social care framework.
• Develop answerable research questions and align them to appropriate quantitative, qualitative or mixed methods.
• Explain the roles of sponsor, chief investigator, sites, ethics review and proportionate governance.
• Identify core ethical principles including safety, scientific validity, fairness, consent and public involvement.
• Know when to seek specialist R&D, REC or regulatory advice rather than relying on local assumptions.

The current UK Policy Framework defines research around the attempt to derive generalisable or transferable new knowledge. Classification is not determined by whether the project is small, unfunded, undertaken by a trainee or intended for publication. Nor is it determined by whether the team personally believes the intervention is low risk.

The HRA “Is my study research?” tool provides a structured route to classification. Use it prospectively and retain the result where local processes require this. If the project sits near a boundary, discuss it with local R&D or research governance teams.

A wrong label can expose participants and staff to avoidable risk. Equally, treating every improvement activity as research can create unnecessary bureaucracy. Proportionate classification matters.

A good question specifies the population, phenomenon or intervention, comparator where relevant, and outcome or experience of interest. PICO is useful for many intervention questions; PICo, SPIDER and other frameworks can support qualitative or mixed-methods questions. The framework is a thinking aid, not a ritual.

The question should determine the design. Do not start with “I want to do a questionnaire” and then search for a question that fits. An intervention-effect question may require a controlled design; an experience question may need qualitative inquiry; a diagnostic question needs a design that can estimate relevant accuracy measures.

Feasibility, value and existing evidence should be considered before recruitment. Research that cannot answer a meaningful question wastes participant contribution.

Randomised trials can provide strong evidence about intervention effects when randomisation is feasible and ethical. Cohort, case-control, cross-sectional, diagnostic and other observational designs answer different questions and have different biases.

Design choice determines what can be inferred. A cross-sectional association cannot establish temporal sequence; a before/after study is vulnerable to secular change; a case-control study is vulnerable to selection and recall biases.

Sample size should be justified in relation to the primary question and analysis, not chosen by convenience alone. Seek statistical support early where the design or analysis is beyond the team's competence.

Qualitative research investigates meaning, experience, behaviour, culture and process. Common approaches include interviews, focus groups, observation and analysis of documents, using methodologies such as thematic analysis, grounded theory, phenomenology or ethnography depending on the question.

Rigour does not mean pretending qualitative data are quantitative. It requires a coherent sampling strategy, reflexivity, transparent analytic steps, attention to deviant cases and sufficient evidence to support interpretations.

Researchers should consider how their professional role and relationship with participants influence what is said, what is observed and how findings are interpreted.

Mixed-methods research integrates quantitative and qualitative evidence to answer a question better than either could alone. The key word is integrates. Running a survey and a few interviews without a plan for how the findings inform each other is merely parallel data collection.

Designs may be convergent, explanatory sequential, exploratory sequential or more complex. The timing and priority of each strand should be explicit, as should the integration point.

Mixed methods can be particularly useful in healthcare where an intervention's effect and its mechanism, acceptability or implementation are all important.

The UK Policy Framework applies across England, Scotland, Wales and Northern Ireland, with country-specific operational and legal differences. Research responsibilities may include a sponsor, chief investigator, research team, funder, sites and regulators. The sponsor has overall responsibility for proportionate arrangements to set up, run and report the project.

Ethics review is one safeguard among several. Some research requires NHS Research Ethics Committee review; some also requires MHRA or other regulatory approvals. Organisational capacity and capability, data protection, contracts, indemnity and local permissions may also apply.

Do not infer the approval route from a colleague's older project. Use current HRA/devolved administration guidance and specialist advice.

Valid consent requires appropriate information, voluntariness and capacity, with additional legal and ethical considerations for children, adults lacking capacity, emergency research and other contexts. Consent is a process, not merely a signed form.

Participant information should explain purpose, procedures, burdens, risks, benefits, confidentiality, withdrawal and what will happen to data or samples. Avoid overstating personal benefit.

Recruitment by clinicians requires attention to power imbalance. Patients should not feel that care depends on participation, and the research team should design proportionate safeguards.

The HRA framework explicitly treats scientific soundness as part of ethical conduct. A poorly designed study can expose participants to burden without a realistic prospect of generating useful knowledge.

Protocol development should therefore include literature review, method justification, feasible recruitment, data quality, analysis planning and arrangements for adverse events where relevant. Public and patient involvement can improve relevance, acceptability and clarity before recruitment starts.

“Low risk” does not excuse poor design. The ethical response to a weak question is to improve or abandon the study before participants contribute.

A team wants to test whether home monitoring after discharge reduces readmission. If the service has already been commissioned and the question is whether it is performing as intended locally, service evaluation may fit. If patients are prospectively allocated to different monitoring strategies to determine whether one causes better outcomes and the intention is to generate transferable knowledge, the project is likely research.

The same topic can therefore support different project types. Classification depends on the purpose and design. The team should use the HRA decision tool and engage the relevant governance route before recruitment or allocation begins.

Write the research question, design, population, inclusion/exclusion criteria, recruitment route, primary outcome or qualitative focus, sample rationale, data source, analysis plan, risks, consent process, public involvement plan and proposed governance route.

Now identify what you do not know. If the list includes whether REC, MHRA, sponsor, data-controller or cross-border approvals are needed, that uncertainty should trigger expert advice rather than an assumption. Use the Stage 4 Research Protocol Skeleton to structure the discussion.

• Assuming small or unfunded studies are not research.
• Choosing the method before defining the question.
• Treating ethics approval as the whole of research governance.
• Using publication intent as the sole classification test.
• Collecting data before confirming sponsor and approval requirements.

1. Health Research Authority (HRA): UK Policy Framework for Health and Social Care Research, version 3.4, updated 28 April 2026. https://www.hra.nhs.uk/planning-and-improving-research/policies-standards-legislation/uk-policy-framework-health-social-care-research/uk-policy-framework-health-and-social-care-research/
2. HRA Decision Tools: Is my study research? and Defining Research terminology. https://www.hra-decisiontools.org.uk/research/

Educational reference. Verify current local governance requirements before starting a project.