Development research runs on people — their bodies, their households, their grief, their poverty. When you survey, interview, photograph or trial something with them, you take something from them. Research ethics is the discipline of taking responsibly.

Ethical research treats participants as the authors of their own lives, not as raw material. They consent to take part, shape what is asked, and retain a stake in what is found. Consent of the governed applies to knowledge as much as to politics.

Modern research ethics did not arrive as abstract philosophy. Each major code was a response to a documented abuse — people harmed, deceived or experimented on without consent. To understand the rules, know what they answer.

After the Second World War, the Nuremberg trials exposed brutal experiments performed on prisoners without consent. The Nuremberg Code (1947) was the response — the first international statement that research on humans requires consent.

Adopted by the World Medical Association in 1964 and revised many times since, the Declaration of Helsinki turned broad principles into working guidance for medical research, and introduced the idea of independent ethics review.

In the U.S. Public Health Service’s Tuskegee Syphilis Study (1932–1972), hundreds of poor Black men with syphilis were observed for decades and deliberately left untreated — even after penicillin became a standard cure — and were never told the truth about their condition.

Issued in the United States in 1979 in the wake of Tuskegee, the Belmont Report distilled research ethics into three principles that still anchor ethics review worldwide.

The early codes came from Europe and North America. But ethics in a context of deep poverty, low literacy and stark power gaps needs more than imported rules. The CIOMS guidelines and India’s own ICMR framework adapt these principles to low-resource realities.

Respect for persons has two parts: honour the choices of those who can decide for themselves, and protect those whose autonomy is diminished — children, people with cognitive impairment, those under duress.

Justice is not abstract here. The same marginalised communities — Dalit, Adivasi, slum, migrant — are surveyed again and again, often with no feedback and no change. Meanwhile the convenient and powerful are rarely subjects at all.

Real fieldwork pits the principles against each other. Respecting one person’s autonomy may expose another; beneficence may require limiting choice; protecting confidentiality may clash with a duty to report harm.

Large numbers of participants in South Asian development research cannot read a printed consent form. A written form they cannot read is a ritual, not consent — and a thumbprint on it can mask total incomprehension.

Voluntariness is fragile where power is unequal. A villager may agree because the researcher arrived with officials, because they fear losing a ration or a programme, or simply because refusing a visitor feels rude or risky.

For minors, you generally need both a parent’s or guardian’s informed consent and the child’s own assent, in language suited to their age. A child’s refusal should be respected even if a parent agrees.

For someone living on the edge, a payment, a meal, or even the hope of future help can override free choice. The poorer the participant, the smaller the sum needed to compromise consent — and the harder it is to walk away.

Research on caste, manual scavenging, untouchability or atrocity touches deep stigma. A finding that brands a community — not just an individual — can entrench prejudice and cause collective harm.

For refugees, undocumented migrants and the displaced, the very act of being recorded can be dangerous. Data on identity, location or movement could expose them to detention, eviction or violence.

The answer to vulnerability is not to exclude these groups — that would erase exactly the people development research exists to serve. The answer is stronger safeguards.

Privacy is about boundaries: people decide what of themselves to reveal, to whom, and in what setting. Barging into a home, questioning in front of neighbours, or probing beyond what was agreed all violate it.

Anonymity is harder than deleting a name. A combination of quasi-identifiers — village, age, caste, occupation, number of children — can pick out one person, especially in a small area where few share those traits.

Information on health (HIV, mental illness), caste, religion, sexuality, immigration status, criminal involvement or experiences of violence carries higher stakes — disclosure can bring stigma, eviction, violence or prosecution.

It is tempting to reassure participants with blanket promises of secrecy. But you may not control every risk — a court order, a funder, a small sample that gives people away. Promising the impossible is its own breach of trust.

Development research rarely involves needles or drugs — so it is easy to assume it is harmless. It is not. Asking questions, recording answers and publishing findings can injure people in ways that never show on a body.

Researchers and field staff who hear accounts of violence, abuse and death can suffer vicarious or secondary trauma. Ignoring their wellbeing is both an ethical failure and a threat to data quality.

Researchers are poorly placed to judge their own ethics — they are invested in their study going ahead. Independent ethics review exists so that a body with no stake in the result checks that participants are protected before any data is collected.

A well-constituted committee includes scientists, a clinician, a legal expert, an ethicist and crucially a lay member who represents the community’s perspective.

The Indian Council of Medical Research’s National Ethical Guidelines for Biomedical and Health Research Involving Human Participants (2017) are India’s principal ethics framework — an update of ICMR’s earlier guidance and the reference point for ethics committees.

Clinical trials of drugs and devices carry an extra layer of regulation. In India the Drugs Controller General of India (DCGI), under the CDSCO, regulates clinical trials, and committees overseeing them must be registered with the national authority.

Textbook ethics meets messy reality the moment you reach the village. Gatekeepers, crowds, raw emotion and impossible trade-offs do not appear in the protocol. This section rehearses the dilemmas you will actually face.

We met this under consent; in the field it sharpens. A crowd gathers when a payment is known; people answer to qualify; gatekeepers expect a cut. What was meant as respect becomes a market — and a source of pressure.

A participant discloses ongoing child abuse, a serious threat, or imminent danger. Confidentiality, normally sacred, now collides with a duty to prevent harm — and sometimes a legal mandatory-reporting obligation.

A photograph is identifiable data and can travel far beyond your intent — into reports, websites, donor decks and social media. A recognisable face attached to a story of poverty, illness or violence can stigmatise for years.

How you portray people in outputs — the framing, the caption, the crop — carries the same ethical weight as how you collected the data. Respect does not end at the camera.

Sitting with grief, violence and destitution day after day leaves a mark. Enumerators and researchers carry secondary trauma, and exhausted, distressed field teams also collect worse data and make worse ethical calls.

The interview ends; the ethical duty does not. How you store, analyse, own, share and report the data — and whether anything flows back to the community — are ethical questions every bit as serious as consent.

Communities are routinely extracted from — surveyed repeatedly while the knowledge, the publications and the value flow to outside institutions. The people who generated the data rarely own, control or benefit from it.

Reciprocity means the people studied learn what was found, in a form they can use — not a 60-page English PDF, but a meeting, a poster, a radio spot, a conversation in their language.

The DPDP Act, 2023 is India’s first comprehensive data-protection law. It sets duties for anyone handling personal digital data — researchers and NGOs included.

Open data advances science — but releasing micro-data on marginalised people can enable re-identification, surveillance or stigma. The push to share must be balanced against the duty to protect.