Case 06 / Attribution risk
Wrong-person attribution
Study how crowdsourced investigations can harm innocent people when similarity, rumor, and limited official information are treated as proof.
Real-case basis: Wrong people were identified online after the Boston Marathon bombing.
Data package
Limited official imagery or public statements
Names, photos, and resemblance claims circulating online
Corrections from media or platforms
Risk signals involving private people and families
Task
Classify resemblance claims as unverified hypotheses, not evidence.
Stop name sharing when official or independent confirmation is missing.
Make the harm risk clear in the report.
Hint
A face that looks similar is not an identification.
Forum claims and scanner rumors are not official confirmation.
Sometimes the correct research action is not to publish.
Answer key
The correct result is to stop publication of names and faces.
The Boston case shows how crowdsourced research can become personal harm.
A responsible report labels the claim as unverified and closes it without identification.
Weak analysis example
Many forum users point to the same person and the face looks similar, so sharing the name will help the investigation.
Careful report example
The available material is limited to low-quality resemblance and social media rumor. Officials have not named this person, and there is no independent confirmation. Personal details should not be published.