Participation should be informed and voluntary.
Information must be understandable enough for the decision actually being requested.
Side 124
The ethical architecture of knowledge production: who bears risk, who receives benefit, how consent and integrity are protected, and what obligations survive after data are collected.
Consent, capacity and vulnerability shape what participation can legitimately mean.
Information must be understandable enough for the decision actually being requested.
Safeguards should increase when participants cannot independently evaluate or authorize participation.
Leaving a study should not require accepting penalties that undermine voluntariness.
Collection, linkage and sharing create risks even when names are removed.
Risk is not only physical; social, economic, psychological and informational harms can be consequential.
Low-probability harms still matter when severity is high.
A useful design should not impose more risk than the research question requires.
Benefit to society does not automatically justify concentrated burden on participants.
Unexpected harms or efficacy signals can change the ethical balance during a study.
Convenience sampling can concentrate burdens while limiting who benefits from the resulting knowledge.
Vulnerability should not become an easy source of accessible participants.
Evidence may fail populations systematically excluded from research.
Research value is partly shaped by whether affected groups can realistically benefit.
Stigmatization or collective identification can occur even when individuals consent.
Fabrication is only the obvious edge; selective analysis, undisclosed conflicts and distorted publication also alter the evidence environment.
Untraceable edits make both error correction and accountability difficult.
Honorary and ghost authorship obscure responsibility.
Disclosure does not eliminate bias but allows readers to evaluate relevant incentives.
Selective publication can systematically distort what later researchers believe.