1 / 7

Gianluca Baio Evidence Project, UCL

Comments on: “The impact of discredited evidence” by David Lagnado & Nigel Harvey. Gianluca Baio Evidence Project, UCL. EVIDENCE 2 Neighbour 2 says S was outside house on night of crime.

akina
Download Presentation

Gianluca Baio Evidence Project, UCL

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Comments on: “The impact of discredited evidence” by David Lagnado & Nigel Harvey Gianluca BaioEvidence Project, UCL

  2. EVIDENCE 2 Neighbour 2 says S was outside house on night of crime Scenario: House burglary, local suspect S apprehended (1) HYPOTHESIS: Suspect S did it EVIDENCE 1 Neighbour 1 says that S often loiters in area ? Neighbour 2 is lying because he dislikes S (Alleged) direct influence of knowing that Evidence 2 is discredited on Evidence 1

  3. Scenario: House burglary, local suspect S apprehended (2) HYPOTHESIS: Suspect S did it EVIDENCE 1 Neighbour 1 says that S often loiters in area EVIDENCE 2 Neighbour 2 says S was outside house on night of crime Is Neighbour 1 lying too? Neighbour 2 is lying because he dislikes S When Evidence 1 & Evidence 2 are observed, knowing that Evidence 2 is discredited influences indirectly the credibility of Evidence 1

  4. Scenario: House burglary, local suspect S apprehended (3) Representation in terms of a BN consistent with the fact that ‘corroborating evidence’ is integrated (link between Reliability Evidence 1 & Reliability Evidence 2, when Evidence 1 & Evidence 2 are observed Does not accommodate differences related to temporal ordering

  5. Intervention model (agent-based causality) HYPOTHESIS: Suspect S did it EVIDENCE 1 Neighbour 1 says that S often loiters in area EVIDENCE 2 Neighbour 2 says S was outside house on night of crime Is Neighbour 1 lying too? Neighbour 2 is lying because he dislikes S If Evidence 2 is known to be “manipulated” (intervened on), the link between Hypothesis S & Evidence 2 is removed. Evidence 2 becomes essentially irrelevant

  6. Intervention model (agent-based causality) (2) • 1. Observe Evidence 1 • Update Hypothesis S & Reliability of Evidence 1 • 2a. Observe Evidence 2 • Update Hypothesis S, Reliability of Evidence 1 & Reliability of Evidence 2 • 2b. Observe Manipulation (discredit Evidence 2) • Update only Hypothesis S & Reliability of Evidence 1

  7. Intervention model (agent-based causality) (3) Including a Control Evidence (i.e. check an alibi) could be a good alternative to qualify the extent to which Evidence 2 is discredited. Probably, the Hypothesis S & Reliability of Evidence 1 would be affected in a more appropriate way

More Related