Research ethics

The BPS Code of Human Research Ethics (2021) rests on four principles. Scientific integrity is one of them: the Code says poorly designed research wastes resources and devalues the contribution of the participants, so whether a study can answer its question is an ethical matter, not only a methodological one.

• Respect for the autonomy, privacy and dignity of individuals, groups and communities (section 2.1).
• Scientific integrity (section 2.2).
• Social responsibility (section 2.3).
• Maximising benefit and minimising harm (section 2.4). Normally, the risk of harm should be no greater than in ordinary life.

Consent must be given freely, with enough information to make an informed choice, by means proportionate to the risk. For non-sensitive data that are anonymous from the start, the Code accepts consent given by taking part or ticking a box; sensitive topics need fuller information. Consent is an ongoing process, renewed where taking part is long or repeated (sections 4 and 4.14).

Participants may ask for their data to be withdrawn. Where there is a necessary limit, such as the point at which data are aggregated, it must be stated when consent is sought (section 2.1).

For young people, the Code says 16 should be acceptable for sole consent in low-risk research, with parental consent sought as well if in doubt, and notes that safeguarding duties continue to 18 (section 4.5).

Incentives should match the burden of taking part and never the degree of risk. Reimbursing expenses such as travel is normally ethically required (section 4.12).

Confidentiality is not absolute. The Code says it can be overridden by more compelling duties, such as protecting someone from harm; where it cannot be guaranteed, participants must be told before they agree, and the procedure belongs in the protocol (section 5).

Deception is acceptable only where it is essential, there is no alternative, the research has strong scientific merit and harm is managed. If participants are likely to react with discomfort, anger or objections when it is revealed, it is inappropriate (section 7). Debriefing should undo any after-effects: the Code’s example is inducing a happy mood before someone leaves an experiment that induced a negative one (section 10).

Much of what clinical psychologists evaluate in services is not research. The Health Research Authority’s Defining Research table separates the kinds of project by purpose, and the difference decides which approvals are needed.

• Research: aims at generalisable or transferable new knowledge. Normally needs NHS research ethics committee (REC) review, but not always.
• Service evaluation: “What standard does this service achieve?” Does not need REC review.
• Clinical audit: “Does this service reach a predetermined standard?” Does not need REC review.
• Each has its own governance to arrange before it starts.

In England and Wales, section 30 of the Mental Capacity Act 2005 makes intrusive research with an adult who lacks capacity to consent unlawful unless an appropriate body has approved the project and it follows sections 32 and 33. Section 32 requires the researcher to consult a carer, or a nominated person with no connection to the project; if they advise that the person would not want to take part, the person does not take part. The BPS Code adds that capacity is assumed unless it is shown to be lacking (section 4.9).

Consent to take part is not the legal basis for processing the data. The Health Research Authority’s guidance says universities and NHS bodies should rely on “task in the public interest”, not consent. Consent to take part is still needed, as an ethical requirement and to meet the common law duty of confidentiality.