Lec 11

Question 1

Problems with forensics
Forensics = interpretation and evaluation of forensic evidence in criminal investigations

Answer 1

• Interpretation: often subjective and the evidence is ambiguous
• Influence of expectations = confirmation bias during interpretation of evidence (reason to believe the suspect guilty)
• Lack of scientific validation: many techniques have not been rigorously tested (false confidence in reliability and validity)

Question 2

Pattern matching
What are four examples of unreliable or unclear reliability techniques

Answer 2

• Non-dna hair matching
• Bullet marks
• Shoe prints
• Bite marks

Question 3

Forensics vs Medicine
Is the standardized?
Incentive/motive?
Restriction to access?
Evidence is open?
Feedback about errors?

Answer 3

• Training and certification procedure varies / standardized, rigorous training
• Labs affiliated, there is an incentive for how careful the tests are because prosecution may be the end goal / peer-reviewed validation studies required
• Restricted access to info (used in court cases) because public should not know all the techniques / open publication of results
• Evidence may be confidential or destroyed / data from many studies
• No feedback – we don’t know if a mistake occurs / feedback about success of treatments

Question 4

Fingerprint ID
Biasing statements e.g. a suspect confessed => 16.6% reversed earlier judgments of the fingerprints they had analyzed earlier in their careers
Note: only recently studied scientifically

Answer 4

This indicates fingerprint ID still has some ambiguity/uncertainty bc some responses changed
Biased statements were leading info
 DNA testing has revealed mistakes, and we’re unsure where/when the error was made
 Biases: neglect of missing info + confirmation bias
 But there’s a reluctance to question reliability because of the legal ramifications

Question 5

Lime Street Fire case
Arson = intentional fire, the evidence is pour patterns of the accelerant

Answer 5

• In all the cases they investigated that were NOT arson, they did not see this pattern, therefore concluded this distinctive pattern was a sign of arson
  Issue? We have to know which observed fires actually were not arson
  => The defendant was almost prosecuted due to flawed analysis
  => Issue: they don’t question if the evidence is true, and they are missing info

Question 6

Problem: effect of contextual info
What contextual info should an examiner know before evaluating evidence?

Answer 6

• The dilemma: examiners may need some contextual info to evaluate evidence (where the sample was taken), but some info not needed because could bias the evaluation (any prior arrests)
  Strategy => linear sequential unmasking

Question 7

Linear sequential masking

Answer 7

• Hierarchy of potentially biasing info
• Presented sequentially
• Begin with trace evidence (not that biasing), evaluated before comparing to reference samples ( level 2) or reviewing case info
• Note: limited support, 49% agree

Question 8

Bias blind spot
Forensic scientists may underestimate the problem

Answer 8

• 71% believed cognitive bias is an issue in forensic science, but only 26% believed/was aware their own judgments are influenced by cognitive bias
• => overconfidence in ability to resist bias
• => makes it harder to correct these issues