Ethical Considerations In Psychology Research

Ethics in psychology refers to the set of principles and guidelines that protect the rights, dignity, and wellbeing of people who take part in research or receive psychological services.

It ensures that psychologists act responsibly, avoid harm, respect confidentiality, and use honesty and fairness in their work.

Ethical standards are vital because they build trust, safeguard participants, and prevent the kinds of abuses seen in past controversial studies

What are Ethical Guidelines?

In Britain, ethical guidelines for research are set by the British Psychological Society (BPS), while in the United States they are published by the American Psychological Association (APA).

These codes aim to protect participants, uphold the reputation of psychology, and safeguard researchers.

Balancing Risks and Benefits

Ethical decisions are rarely clear-cut.

A study may cause distress or harm, yet also produce discoveries that benefit individuals or society.

As Rosenthal and Rosnow (1984) note, failing to conduct research can also carry costs.

The guiding principle is that participants’ interests must always come first when weighed against the aims of researchers.

This is beneficence in practice. The Belmont Report’s principle requires researchers to maximise possible benefits and minimise possible harms, not just avoid the worst outcome.

Role of Ethics Committees

To ensure this balance, all studies must be reviewed before implementation.

• US: Institutional Review Boards (IRBs).

• UK: Departmental, Institutional, or External Ethics Committees (e.g., NHS research).

Committees review proposals to assess if the potential benefits of the research are justifiable in light of the possible risk of physical or psychological harm.

These committees may request researchers make changes to the study’s design or procedure or, in extreme cases, deny approval of the study altogether.

Codes of Practice

The British Psychological Society (BPS) and American Psychological Association (APA) have issued a code of ethics in psychology that provides guidelines for conducting research:

These codes translate broad ethical principles into concrete requirements. They cover informed consent, limits on deception, protection from harm, confidentiality, the right to withdraw, and debriefing. An independent committee then checks that a specific study meets them before it can start.

---

Historical Origins of Research Ethics

Modern ethical guidelines grew out of real historical abuses, not abstract theory. Each founding document responds to a specific scandal.

The Nuremberg Code (1947)

The Nuremberg Code followed the “Doctors’ Trial” of Nazi physicians who performed lethal experiments on concentration-camp inmates. The judges who presided over the trial wrote the code themselves.

It was part of their verdict.

Its ten principles established that research participation must be freely chosen. Most famously, “the voluntary consent of the human subject is absolutely essential.”

The stakes were high for everyone involved.

A participant must be free to end the experiment at any point. Risk must never outweigh expected benefit, and unnecessary suffering must always be avoided.

The investigator was bound too.

They had to be ready to terminate the experiment the instant harm became likely. The Code was written for physician-researchers and carried no formal enforcement mechanism. It remains, even so, the direct ancestor of every later ethics code.

The Declaration of Helsinki (1964)

The World Medical Association wrote the Declaration of Helsinki to close gaps in the Nuremberg Code. It has been revised repeatedly, most recently in 2013 (World Medical Association, 2013).

It separated therapeutic from non-therapeutic research.

It also emphasised the primacy of a participant’s welfare over scientific and societal interests. It allowed legally authorised representatives to consent on behalf of participants who could not consent themselves.

Later revisions added one more requirement.

Clinical trials now had to be registered in advance, with results reported publicly regardless of outcome.

The Tuskegee Study and the Belmont Report

The Tuskegee Syphilis Study (1932–1972) forced US federal reform. The US Public Health Service enrolled 399 African-American men with latent syphilis, plus a comparison group of 201 uninfected men.

It followed them for four decades.

Researchers told the men they were being treated for “bad blood.” They withheld penicillin after it became the standard cure for syphilis in the mid-1940s (Jones, 1981). An Associated Press reporter exposed the study in 1972, and it was shut down.

Direct institutional consequences followed.

The National Research Act arrived in 1974, and the Belmont Report five years later.

The Belmont Report (1979) was the government’s response.

It distilled research ethics into three principles that still guide US regulation today. Belmont frames them as guiding principles, though, not decision rules that settle a case by themselves.

Respect for persons means autonomy, with extra protection for those with diminished autonomy. Beneficence means maximising benefit and minimising harm. Justice means a fair distribution of research’s burdens and benefits.

---

Informed Consent

Before a study begins, researchers must explain the purpose, procedures, risks, and potential benefits, and then obtain participants’ voluntary agreement to take part.

Consent must be:

• Informed – participants clearly understand what they are agreeing to.

• Voluntary – given freely, without pressure, coercion, or excessive inducement.

• Rational – the person has the capacity to weigh the decision.

Adults (18+) can provide their own consent, while parents or guardians must provide consent for minors.

---

What Participants Must Be Told

To make a meaningful decision, participants should receive clear information about:

• The voluntary nature of participation, and that refusal has no consequences.

• The purpose of the research and what it involves.

• All foreseeable risks and discomforts (both physical and psychological).

• The procedures they will undergo.

• Potential benefits for society and/or individuals.

• The expected duration of participation.

• Who to contact with questions, concerns, or in case of harm.

• Their right to confidentiality and how their data will be protected.

• Their right to withdraw – the entitlement to stop taking part, or have their data removed, at any time without penalty.

Consent forms should be accompanied by a plain-language participant information sheet and researcher contact details.

---

Special Considerations

Researchers have a duty to respect participants’ dignity and autonomy, being sensitive to power imbalances between researcher and participant.

Consent processes must be adapted to the needs of the study population:

• Children and adolescents – require parental/guardian consent as well as the child’s own agreement where possible.

• Vulnerable groups (e.g., psychiatric patients, people with intellectual disabilities) – capacity must be assessed individually, with support provided to aid comprehension.

• Cultural context – language and procedures should be tailored to the participant’s background.

• Inducements – payments or rewards must be fair and not so large that they unduly influence participation.

---

Alternatives and Ongoing Consent

When direct consent is not possible, researchers may use presumptive consent, asking a comparable group how they would feel about participation.

However, this method is limited, since imagined reactions often differ from actual experience.

Importantly, consent is an ongoing process, not a one-off form.

Researchers should check participants’ willingness throughout, remain alert to discomfort, and allow them to withdraw at any time.

The BPS also advises researchers to keep a record of ethical decision-making, particularly when dilemmas arise.

---

Debrief

Purpose of Debriefing

Debriefing is a key ethical requirement in psychology.

After a study, researchers must explain the purpose of the research, give participants the chance to ask questions, and address any potential harm.

The aim is to:

• Educate participants about the study.

• Correct misconceptions or anxieties.

• Restore dignity and wellbeing, leaving participants in a similar state of mind as when they entered the study (Aronson, 1988).

As Harris (1988) put it, the goal is to ensure participants leave with “a sense of dignity, knowledge, and a perception of time not wasted.”

---

What Debriefing Should Include

An effective debrief goes beyond a polite thank you.

It should include:

• Study aims – explaining what the researcher was investigating and why.

• Participant’s role – clarifying how their participation contributed.

• Deception disclosure – if deception was used, participants must be told and given a clear justification for why it was necessary.

• Opportunity for questions – participants must be encouraged to ask questions, and researchers are obliged to answer them openly and fully.

• Comprehension check – experimenters should take reasonable steps to ensure participants actually understand the debriefing.

Debriefing should take place as soon as possible, ideally immediately after participation, and use plain, accessible language.

---

Ongoing Responsibility for Participant Welfare

Debriefing is not just the end of a study; it reflects the researcher’s ongoing duty of care.

Ethical researchers should:

• Monitor for distress – watch for signs of discomfort during or after participation.
• Provide support – offer resources or referrals if sensitive issues were triggered.
• Follow up if needed – check in when participation may have long-term effects (e.g., socially sensitive research).
• Reflect and record – document ethical concerns raised to improve future studies.

---

Exceptions and Limitations

Although debriefing is normally mandatory, there are rare exceptions:

• Methodological reasons – in some studies, immediate debriefing could compromise validity if participants might share details with others who have not yet taken part.

• Risk of harm from debriefing – for example, revealing deception to children in a way that could create distress or undermine trust.

In these cases, researchers must:

• Seek explicit approval from an Institutional Review Board (IRB) or Ethics Committee.

• Justify why a full or immediate debrief is not possible.

• Provide a delayed or partial debrief when appropriate, ensuring participants are not left permanently misinformed.

---

Protection of Participants

Researchers must protect participants from physical, psychological, social, and emotional harm throughout the research process.

This obligation covers both direct harm (e.g., pain, stress) and indirect harm (e.g., reputational damage, privacy violations).

Participants should leave the study in a state of no worse wellbeing than when they entered.

---

Types of Harm to Avoid

Researchers should take active steps to prevent:

• Physical harm – injury, illness, excessive fatigue, or physical risk.

• Psychological harm – anxiety, stress, guilt, loss of self-esteem, embarrassment, or trauma.

• Social harm – stigmatization, stereotyping, or damage to relationships.

• Economic or legal harm – loss of employment, financial costs, or exposure to criminal liability if sensitive data is mishandled.

These four categories matter because they map onto different routes of injury, from a physical injury to a damaged reputation.

---

Acceptable Risk Threshold

Researchers must carefully manage the level of risk participants are exposed to, ensuring that harm is minimized and never excessive.

Key expectations are:

• No greater than everyday life – risks should not exceed what participants face in normal daily activities.
• Higher risks require approval – any added risk must be minimized and cleared by an ethics committee or IRB.
• Protect participants first – risk–benefit analysis must prioritize welfare over research outcomes.
• Stop if unsafe – studies should not proceed if risks cannot be adequately managed.

---

Special Safeguards for Vulnerable Groups

Extra protections are required when working with populations such as:

• Children and adolescents – sessions should be short, age-appropriate, and avoid unnecessary fatigue or confusion.

• Elderly adults – consider health conditions, mobility, and cognitive load.

• People with intellectual or developmental disabilities – ensure consent and participation are tailored to their comprehension level.

• Psychiatric patients – safeguard against triggering symptoms or distress.

• Marginalized groups – ensure participation does not reinforce stigma or discrimination.

In all cases, capacity to consent and cope must be individually assessed, and procedures adapted accordingly.

---

Monitoring During the Study

Ethical responsibility continues throughout the research process, requiring researchers to remain alert to participants’ wellbeing.

This involves:

• Watching for distress – continuously observe for signs of discomfort, fatigue, or anxiety.

• Reinforcing autonomy – remind participants of their right to withdraw at any time, without penalty.

• Building in check-ins – include breaks or wellbeing assessments in sensitive or demanding studies.

These checks continue even after a session ends, since some effects only surface later.

---

Managing Unexpected Harm

Even with careful planning, harm can occur. Researchers must:

• Intervene immediately if distress becomes evident.
• Offer therapeutic debriefing or psychological support if needed.
• Provide referrals to professional services if harm persists after the study.

Famous examples like Zimbardo’s Stanford Prison Experiment show the dangers of failing to stop a study when participants are clearly suffering.

Ethics committees may also require the incident to be reported and reviewed afterwards, so future studies build in stronger safeguards.

---

Deception

Deception in research occurs when participants are not given the full truth about the study. It can take two main forms:

1. Active deception (commission): participants are deliberately misled, e.g.:

   • Using confederates (actors posing as real participants).

   • Giving false feedback (e.g., telling participants they failed a test when they did not).

   • Creating staged events or giving deceptive instructions.

2. Passive deception (omission): researchers withhold information, such as:

   • Not explaining the true aim of the study.

   • Describing the task ambiguously so participants draw incorrect conclusions.

Both forms reduce participants’ ability to give informed consent and limit autonomy, which is why deception is considered ethically sensitive.

---

Why Deception Is Used

Some psychological questions cannot be investigated honestly if participants are fully aware of the study’s true purpose.

When people know what is being tested, they may consciously or unconsciously change their behavior to fit expectations, a phenomenon known as demand characteristics.

To avoid this, researchers have sometimes relied on deception.

Baumrind (1985) argued that psychologists should prefer deception-free designs wherever possible. One alternative, proposed by Mixon (1972), is role-playing: telling participants the scenario in full and asking them to enact how they would respond.

Classic examples:

• Milgram (1963) – participants believed they were administering painful electric shocks to study obedience.
• Asch (1951) – confederates gave obviously wrong answers in a line-judging task to study conformity.
• Schachter & Singer (1962) – participants were misled about an injection’s true effects to test theories of emotion.

These studies show how deception has led to groundbreaking insights but also caused lasting debate about ethical costs.

Milgram’s Obedience Study: A Key Ethical Case

Milgram’s obedience studies (Milgram, 1963, 1974) produced the most consequential ethical dispute in twentieth-century psychology. The details below are compact; the full Milgram experiment is covered on its own page.

The stakes were high.

Aim: to test whether ordinary people would obey a legitimate-seeming authority’s instructions to shock an innocent stranger, against their own moral objections.

Method: The setup was deceptively simple. 40 male volunteers were told by a lab-coated experimenter to deliver shocks rising in 15V steps to a labelled 450V to an unseen “learner.” Four scripted prods pushed any hesitant participant to continue, including “you have no other choice, you must go on.”

Results: 65% of participants delivered the full 450V shock. Every participant continued to at least 300V, despite visible signs of acute stress.

Conclusion: Milgram (1974) concluded that ordinary people will obey a legitimate authority even to the point of apparently endangering another person’s life.

The reaction was immediate. Baumrind (1964) argued that the study inflicted severe emotional distress on participants who had trusted the experimenter. No scientific yield, she wrote, could justify that cost. She also argued that debriefing could not undo the harm of learning what one was willing to do.

Milgram (1964) replied that only 1.3% of participants regretted taking part, and that a psychiatric follow-up found no evidence of lasting harm. He argued that unpleasantness felt during a study is not the same as harm from it.

The exchange was never resolved. It shaped the modern requirements for prior ethics review, minimised deception, and proper debriefing that the rest of this article describes.

---

Ethical Safeguards

Ethics codes (APA, BPS) state that deception may be used only when strict conditions are met:

• Last resort – deception is allowed only if no non-deceptive method can answer the research question.

• Scientific justification – the knowledge gained must have strong potential benefits to science, society, or participants.

• Independent approval – deception must be judged acceptable by an ethics committee or IRB.

• Minimal impact – deception should be as limited as possible and must not cause lasting physical, psychological, or social harm.

• Anticipate reactions – researchers should consider cultural context and whether participants are likely to feel angry, betrayed, or harmed once deception is revealed.

---

Debriefing After Deception

Because deception interferes with informed consent, thorough debriefing is essential. This involves:

• Revealing the truth – explaining the actual purpose of the study as soon as possible.

• Justifying the deception – clarifying why misleading participants was scientifically necessary.

• Reducing harm – addressing potential negative reactions (e.g., guilt, embarrassment, anger).

• Offering support – providing reassurance or referrals if distress lingers.

If participants are likely to be distressed or refuse consent once they learn the truth, then the study is considered ethically unacceptable.

---

Consequences of Poorly Managed Deception

When deception is not handled carefully, it can have damaging consequences for both participants and the discipline of psychology.

• Loss of trust – participants may feel betrayed, leading to reluctance to volunteer in future studies.

• Reputational harm – psychology as a discipline may be seen as manipulative or exploitative.

• Long-term participant distress – guilt, lowered self-esteem, or suspicion of researchers.

• Historical backlash – controversial studies like Milgram’s and Zimbardo’s have shaped modern ethics codes precisely because of concerns about deception and harm.

---

Confidentiality

Protecting participant confidentiality is a core ethical obligation in psychology.

It shows respect for individuals, promotes honest and open participation, and reduces risks such as embarrassment, stigma, or legal repercussions.

Participants must be assured that any information they provide will be handled with care and not made personally identifiable unless they have explicitly given consent.

For example, names should never appear in published reports, and data should be presented in ways that prevent identification.

Methods of Protecting Confidentiality

The steps taken to safeguard privacy should match the methodology and risks of the study.

In face-to-face research, coding systems replace names with numbers. Restricted access to data files and secure physical storage add further protection.

In online research, additional challenges arise, such as the risk of third-party access.

Here, encryption, password protection, and secure data servers are essential. Participants should be informed of these protections too.

Researchers can further protect privacy by collecting only the necessary information. For instance, they might record age rather than a full birthdate.

Limits of Confidentiality

Researchers must also be transparent about the limits of confidentiality.

Participants have a right to know whether there are circumstances where their information may be disclosed.

While confidentiality is the default, there can be exceptions.

For example, when researchers encounter information that suggests risk of serious harm to the participant or others.

Modern data-sharing norms complicate this further. Open, reproducible science depends on sharing data, but data that can be shared freely is rarely still confidential (Meyer, 2018). Researchers now manage the tension with de-identification and controlled-access repositories for higher-risk data.

Disclosure Dilemmas

One recurring ethical dilemma is whether to disclose information about a participant’s criminal behavior.

In most jurisdictions, researchers have no automatic legal duty to report crimes uncovered during a study.

Decisions to disclose involve a genuine conflict. Researchers must weigh their primary duty to the participant against their duty to protect the wider community.

This judgment is always contextual and should align with both the research aims and professional ethical codes.

---

Withdrawal from an Investigation

Participants must be free to leave a study at any time if they feel uncomfortable, without penalty or pressure. This right extends not only to their participation but also to the data they have provided.

Communicating the Right

Researchers have a responsibility to clearly explain the right to withdraw at all stages of participation.

• Inform at the start – participants should be told clearly that they can withdraw at any point, for any reason.

• Remind during the study – this right should be reinforced throughout, especially in longer or more demanding experiments.

• Reassure at the end – participants must be given a final opportunity to withdraw their data after debriefing.

---

Avoiding Pressure

Participants must never feel pressured or coerced into continuing against their will.

• No coercion – they should not feel obligated to continue, even if they worry about “spoiling” the research.

• Incentives clarified – if payment or course credit is offered, participants must be reassured they will still receive it even if they withdraw.

• Example of misuse – in Milgram’s obedience study, participants were pressured to continue despite distress, a clear breach of modern ethical standards.

---

Ethical Issues in Psychology & Socially Sensitive Research

Why Standard Ethics Are Not Enough

Traditionally, psychologists assumed that if they complied with the BPS or APA codes – ensuring participants were not deceived, humiliated, or harmed, and that debriefing and confidentiality were provided – then their work was ethically unproblematic.

However, socially sensitive research raises issues that extend beyond the treatment of participants.

Even when no direct harm is caused to individuals, the way findings are interpreted, publicized, and applied can have far-reaching consequences for groups, communities, and society.

Illustrative Examples

These examples show that the main ethical concern is not always the treatment of participants, but the social consequences of research outcomes.

Childcare and Cognitive Development

Caughy et al. (1994) found that middle-class children in early daycare scored lower on cognitive tests than those raised at home.

No participants were harmed, consent was obtained, and confidentiality preserved.

Yet, once published, the findings risked stigmatizing daycare, causing anxiety for parents, and influencing public opinion or policy on childcare provision.

Race and Intelligence

IQ testing in the US has consistently shown Black Americans scoring lower on average than white Americans.

Although participants were not harmed directly, the publication of such results has historically been weaponized to justify discriminatory hiring, educational segregation, and eugenics.

This demonstrates how SSR can unintentionally reinforce systemic inequality.

Sieber and Stanley (1988) outline four groups that may be affected by psychological research. The first is the group most directly concerned.

1. The social group being studied: Findings may stigmatize or disadvantage an ethnic, gender, or social group. For example, early research on IQ was used to discriminate against US Blacks.
2. Friends and relatives of those participants: Case studies may reveal private information or attach stigma. Example: Genie’s mother, who became infamous after her daughter’s case study.
3. The research team. There are examples of researchers being intimidated because of the line of research they are in.
4. The institution in which the research is conducted. Universities or funding bodies may be discredited if the research is controversial or misused.

Four Core Ethical Concerns in SSR

1. The Research Question – framing a question (e.g., “Are some races less intelligent?”) may itself perpetuate harmful assumptions.

2. Treatment of Participants – individuals must still be protected from harm, embarrassment, or coercion.

3. Institutional Context – the motives of sponsors (e.g., government, military, corporations) may bias research goals.

4. Interpretation & Application – findings may be misrepresented by the media, policymakers, or public, leading to real-world harm.

Ethical Guidelines For Carrying Out SSR

Sieber and Stanley suggest the following ethical guidelines for carrying out SSR. There is some overlap between these and research on human participants in general.

• Privacy – questions about sexuality, trauma, or family life can feel intrusive and offensive, even if data are anonymized.

• Confidentiality – leakage of sensitive information (e.g., HIV status, criminal history) could cause reputational, legal, or health consequences.

• Sound Methodology – SSR demands high methodological rigor. Flawed methods can still influence policy because lay audiences cannot easily detect weaknesses.

  • Example: Bowlby’s maternal deprivation studies, despite limitations, shaped childcare policy for decades.

• Deception – misleading findings (or careless interpretation) may create false public beliefs, e.g., overstating parental responsibility for child outcomes.

• Informed Consent – participants should understand potential wider implications of their involvement, not just immediate risks.

• Justice & Equity – avoid research that exacerbates prejudice or denies treatment to some groups for comparison.

• Scientific Freedom vs. Responsibility – researchers have a right to investigate. This must be balanced with social responsibility.

• Ownership of Data – sensitive data may be misused if fully public. Questions remain: should politically charged findings be freely accessible, or restricted until properly contextualized?

• Practice reflexivity – acknowledge your own values and biases, and those of participants. These shape design, analysis, and conclusions.

• Cost–Benefit Analysis – costs often extend beyond individuals to entire communities. Since participants rarely benefit directly, researchers must ensure wider social benefits clearly outweigh risks.

Arguments for SSR

• Reduces prejudice and discrimination – by studying gender, race, sexuality, or disability, psychology can challenge stereotypes and inform inclusive policies.
• Societal benefits – research on eyewitness testimony (EWT) revealed memory fallibility, shaping fairer trial procedures and highlighting children’s reliability as witnesses.
• Corrects research bias – much psychological research historically relied on white, middle-class American samples (Henrich et al., 2010). SSR diversifies knowledge and makes findings more representative.
• Higher scrutiny – ethics boards review SSR with more rigor, often rejecting weak or harmful proposals. This safeguards quality.
• Broader accountability – Sieber and Stanley’s (1988) framework names four groups whose interests are at stake (participants, their families, the research team, and sponsoring institutions), giving reviewers a structured, four-part way to weigh wider social consequences beyond the individual participant.

Arguments against SSR

• Policy misuse – flawed studies have justified eugenics, forced sterilization, or exclusionary immigration laws (e.g., early 20th-century US policies).
• Reinforcement of stereotypes – findings may unintentionally confirm prejudices, even if the researcher’s intent was neutral.
• Weak regulation – current ethical guidelines may lack the power to prevent harmful or politically driven SSR.
• Harm to vulnerable groups – even anonymized findings can stigmatize entire communities.
• Impeccable methods, harmful impact – a study on race and cognitive ability can have flawless consent and full debriefing, yet still damage the standing of the group it studied once findings enter public debate. Procedural compliance alone cannot prevent that harm.
• Compliant research can still stigmatise – research on childcare arrangements can follow a code of ethics to the letter and still stigmatise the parents whose choices it describes, because the harm operates through public interpretation of findings rather than through how participants were treated.

Critical Evaluation of the Ethics Framework

Seven decades of codes, committees and case law have changed the field. Outright exploitation of participants is now rare in mainstream psychology. Tuskegee-style withholding of treatment and Nuremberg-style non-consensual experimentation are no longer tolerated.

That progress took decades to earn.

Prospective ethics review gives a genuine check on individual researchers’ judgement. Investigators are rarely neutral about their own work. An independent committee catches conflicts they might otherwise miss. Sieber and Stanley’s framework extended this protection beyond individual participants to the wider social groups research can affect.

Real limitations remain, though. Procedural focus can crowd out substantive questions. A committee may approve a study that ticks every consent-form box while asking a question that should not have been asked at all.

Committee decisions are also inconsistent. The same protocol can receive very different treatment at different institutions, which is unfair to researchers without making participants any safer.

Consistency remains a work in progress.

A deeper problem is structural. The framework assumes discrete studies with clear starts and ends. Data collected for one study, though, is increasingly reused, combined, and re-analysed by other researchers years later.

Consent is a further weak point. Participants often cannot foresee the real burden or benefit of taking part, especially in longitudinal or AI-driven studies.

The codes also keep a US and European focus. That sits awkwardly with indigenous data-sovereignty protocols and other frameworks where the individual is not the primary unit of consent.

Contemporary Research

Nebeker, Torous and Bartlett Ellis (2019) reviewed how classical ethics principles hold up against digital-health research: studies using passive sensing, mobile apps, wearables, and machine-learning models.

They found consent breaking down first. Participants cannot foresee, at the start of a study, every analysis that will later run on years of sensor data. The authors proposed dynamic consent, letting participants update their permissions as new uses of their data arise, instead of a single up-front agreement.

Dynamic consent is not yet standard practice.

They also flagged a fairness risk. A machine-learning model trained mostly on one population, such as college students, can perform worse for the populations it is later used on. The researcher who never checked carries that ethical burden.

Their prescription was practical. Ethics committees should recruit data-science expertise and build in ongoing monitoring after a study launches. Consent, on this view, is something that evolves rather than a box ticked once.

American Psychological Association. (2017). Ethical principles of psychologists and code of conduct. https://www.apa.org/ethics/code

Baumrind, D. (1964). Some thoughts on ethics of research: After reading Milgram’s “Behavioral study of obedience.” American Psychologist, 19(6), 421–423. https://doi.org/10.1037/h0040128

Caughy, M. O. B., DiPietro, J. A., & Strobino, D. M. (1994). Day‐care participation as a protective factor in the cognitive development of low‐income children. Child development, 65(2), 457-471.

Harris, B. (1988). Key words: A history of debriefing in social psychology. In J. Morawski (Ed.), The rise of experimentation in American psychology (pp. 188-212). New York: Oxford University Press.

Rosenthal, R., & Rosnow, R. L. (1984). Applying Hamlet’s question to the ethical conduct of research: A conceptual addendum. American Psychologist, 39(5), 561.

Sieber, J. E., & Stanley, B. (1988). Ethical and professional dimensions of socially sensitive research. American Psychologist, 43(1), 49–55. https://doi.org/10.1037/0003-066X.43.1.49

British Psychological Society. (2021). Code of human research ethics (2nd ed.). British Psychological Society.

Further Information

• MIT Psychology Ethics Lecture Slides
• Ethical Concerns In Qualitative Research

BPS Documents

• Code of Ethics and Conduct (2018)
• Good Practice Guidelines for the Conduct of Psychological Research within the NHS
• Guidelines for Psychologists Working with Animals
• Guidelines for ethical practice in psychological research online

APA Documents

• APA Ethical Principles of Psychologists and Code of Conduct