4.2.3.1 - Ethics in psychological research

Why ethics matters

Psychological research studies people. That sounds obvious, but it creates a special problem: the people being studied have rights, feelings, privacy, relationships and sometimes vulnerabilities. A method that produces valid data may still be unacceptable if it treats participants badly. Disclosing a precise hypothesis may change participants’ behaviour, but that methodological difficulty does not itself justify withholding information needed for a meaningful choice.

Ethics is therefore not just a list of "banned" procedures. It is a decision-making process. Researchers have to ask whether a study is worth doing, whether the risks are acceptable, whether participants can make a genuine choice, and how any unavoidable ethical problems will be managed.

For AQA, the key idea is that ethical issues can appear at three points:

The BPS code matters because it gives UK psychologists a shared professional framework. It is not a simple checklist that removes judgement. Instead, it helps researchers and ethics committees decide whether a study respects participants while still allowing worthwhile research. AQA also expects practical psychology work to follow BPS ethical guidance, so the code is part of research methods, not just professional administration.

The BPS Code of Ethics and Conduct is built around four main principles:

Ethics committees use the BPS framework, institutional rules and legal duties to review proposals before data collection. Their role is especially important when research involves deception, vulnerable participants, sensitive topics, possible distress or identifiable personal data.

Cost-benefit analysis is a useful AO3 idea because ethics often involves trade-offs. A mildly embarrassing task may be acceptable if the research question is valuable and the safeguards are strong. A highly distressing procedure may be rejected even if the findings could be interesting. The judgement depends on the severity of risk, the likelihood of harm, the value of the research and the quality of the safeguards.

The broader Code of Ethics and Conduct should be distinguished from the BPS Code of Human Research Ethics. The research code develops principles of respect for autonomy, privacy and dignity; scientific integrity; social responsibility; and maximising benefit while minimising harm. Poorly designed research can itself be unethical if it burdens participants without a reasonable prospect of useful knowledge. BPS research ethics guidance.

Ethical issues in design

Ethical research starts before a single participant is tested. A researcher should design the study so that participants know enough to choose freely, risks are minimised, private information is protected and any deception is justified.

The BPS often uses the phrase valid consent.

For example, a participant should not feel pressured by a teacher, employer, therapist or researcher who has power over them. Consent procedures also need to fit the person and context, such as using age-appropriate information for younger participants.

There is a tension here. If participants are told the full aim, they may guess what the researcher expects and change their behaviour. If they are not told enough, their consent is weakened. This is why ethics is built into design: researchers must decide how much information is necessary for a genuine choice without destroying the validity of the study.

Deception can involve telling participants a false aim or withholding important information. Not stating every prediction is different from concealing a distressing procedure: the ethical significance depends on what a person needs to decide freely. The BPS research code distinguishes limited withholding of hypothesis details from deliberately giving a false account of the purpose. It is ethically sensitive because participants cannot give fully informed consent if they are misled. However, some studies would be invalid without limited deception. A researcher investigating helping behaviour, for instance, may not be able to tell participants the exact aim beforehand because this would change whether they help.

Psychological harm includes stress, embarrassment, humiliation, guilt, lowered self-esteem, anxiety and distress caused by sensitive material. In design, the researcher should remove unnecessary risk, avoid topics they are not competent to handle, and prepare support if discomfort is possible.

The right to withdraw must be designed into the procedure. Participants need to know they can stop, and the study should make stopping realistic. If a participant is paid, for example, they should not lose all payment simply because they withdraw. If data are anonymous and cannot be removed later, that limit should be explained before consent.

Privacy is especially important in observations, online research, case studies and sensitive questionnaires. Just because behaviour can be seen does not always mean it is ethical to record and publish it. A public behaviour in a shopping centre is different from a private conversation in a clinic waiting room.

Confidentiality should be planned before data collection. The researcher should decide what data are needed, how they will be stored, who can access them, how long they will be kept and how findings will be reported without identifying individuals.

Some participants need extra care.

Vulnerability depends on the research context, not just a group label. A patient may freely choose a low-risk task, while a confident student may feel unable to refuse their own teacher. Do not assume that disability or a diagnosis means someone cannot consent. Use accessible information, assess relevant capacity where needed, reduce unequal power and seek appropriate support or permission. Avoid a procedure altogether if the risks cannot be justified.

Worked consent decision

A teacher invites their own class to complete a stress questionnaire. A signature is not sufficient if students think refusal will affect their grades. Use an independent invitation, make clear that the teacher will not see individual answers and provide an ordinary alternative activity. These arrangements address pressure and privacy directly; describing participation as "voluntary" while watching who refuses would undermine that reassurance.

Ethical issues in conduct

Even a well-designed study can become unethical during data collection if the researcher handles people badly. Ethical conduct means applying the plan sensitively in the moment, not just collecting signed consent forms.

The first conduct issue is pressure. Participants may technically have consented, but still feel they cannot leave. This can happen if the researcher is an authority figure, if the study is in school or work time, if the participant is embarrassed to stop, or if the instructions make withdrawal sound inconvenient. Researchers should use neutral language and remind participants that stopping is acceptable.

The second conduct issue is monitoring distress. A participant may become anxious, frustrated or embarrassed during a task. Ethical conduct requires the researcher to notice this and respond. Depending on the situation, they might pause the study, check whether the participant wants to continue, stop the procedure, or move straight to debriefing and support.

The third issue is confidentiality during the procedure. A study can protect names in the final report but still breach privacy during data collection. For example, if questionnaire answers are collected in a classroom where other students can see them, confidentiality is weak. If an interview about mental health is held in a public corridor, privacy is poor even if the final transcript uses initials.

The fourth issue is deception during conduct. If deception is used, it should be limited to what is necessary. The researcher should not add unnecessary lies, increase distress to make the effect stronger, or continue deception after the point where it is no longer needed.

The fifth issue is researcher competence. A researcher must be able to handle the method and the topic. An undergraduate student might competently run a low-risk memory task, but may not be competent to conduct unsupervised interviews about trauma, self-harm or severe mental health symptoms.