Responsibility

Participants' wellbeing comes before the study's results — people before data.

Agreeing to take part knowing what the study involves (a guardian consents for children).

Not exposing participants to lasting physical or psychological harm.

Anonymity: nobody can be identified from the data or the report. Confidentiality: the data are kept private and used only as agreed.

Being able to stop and leave, and remove your data, at any time without penalty.

Telling participants the true aim after the study and checking they are okay.

When it is justified, causes no real harm, and is followed by a full debrief.

People may act differently if they know the aim, so mild deception keeps behaviour natural.

A parent or legal guardian.

Consent, Anonymity, Right to withdraw, Deception, Undue stress (protection from harm), Debrief: CARDUD.

Information, comprehension, voluntariness.

A parent or guardian consents for a child, and the child agrees (assents) too.

Deliberate misleading (e.g. a confederate), deceptive instructions, and deception by omission.

Those who agree may differ from those who refuse (self-selection bias); knowing the aim can create demand characteristics.

Weighing possible harm to participants against the value of what the study could find.

Participants dropping out of a study before it ends. If those who leave differ from those who stay, the results are biased.

People answer more honestly about sensitive topics, so the data are more valid. But the researcher cannot follow up.

In public places where they expect to be seen by strangers, with nobody identified and no risk of harm.

To control conditions and study processes that would be impossible or unethical in humans.

Replace (use something else), Reduce (use fewer animals), Refine (cause less suffering).

Use a method without animals wherever possible: models, cells, scans of human brains.

Use the smallest number of animals that still gives valid results.

Adjust procedures to minimise pain and distress (housing, pain relief, humane handling).

Benefits must clearly outweigh the animals' suffering, with ethics-committee approval.

They cannot understand or agree to a study, so the responsibility rests entirely with the researchers.

It can reveal mechanisms that would be impossible or unethical to study in humans.

Animals cannot consent and can suffer, so their use must be strictly limited.

How good is the research? How likely is the benefit? How much will the animals suffer?

The duty to report findings honestly and to think about how the knowledge is used in society.

Presenting results accurately, with limitations, without exaggeration or hiding findings.

Average group differences are not fixed truths about any individual.

Anticipating how findings could justify harm and framing them to limit that risk.

Because how findings are reported and used can help or harm society.

A measure valid in one culture may unfairly lower another group's scores.

A mark of shame attached to a group of people.

No. It is an average, and many people in the "lower" group score higher than many in the other.

State the real size and limits of a finding so it is not exaggerated.

Pressing for change that the evidence supports, for example in a school or a law.

Its findings could harm the group it is about, especially a group that is already treated unfairly.