BIAS

Question 1

Expectation bias

Answer 1

• examiners look for indicators that confirm the outcome they were expecting
• disbelieve or minimise significance of opposing results

Question 2

Confirmation bias

Answer 2

• people test hypotheses by looking for confirming evidence rather than conflicting evidence
• can be overcome by trying to contradict yourself

Question 3

Contextual bias

Answer 3

-tendency for a consideration to be influenced by other information aside from that being considered

Question 4

Anchoring effects

Answer 4

• individual relies too heavily on an initial piece of information when making subsequent judgements
• conclusions then interpreted based on the anchor

Question 5

Role effects

Answer 5

• tendency for individual to identify as part of a team with common goals
• e.g. identify as prosecution trying to find a perpetrator

Question 6

How is bias seen in forensic science?

Answer 6

• methods are subjective and are especially vulnerable to human error and cognitive bias
• tackled by a robust experimental design:
• clearly define research question
• factor in practical and legal restraints
• design experiment within those restraints

Question 7

What is the problem with bias in fingerprint analysis?

Answer 7

• longstanding belief that it is evidence of fact, not opinion
• presented with 100% confidence

Question 8

What de-biasing techniques did the fingerprint community establish?

Answer 8

• sequential unmasking
• independence of multiple examinations
• training
• cognitive risk assessments
• ACE-V system

Question 9

What is the ACE-V system?

Answer 9

A - analysis - information gathering stage
C - comparison - side by side with original
E - evaluation - identification/exclusion/inconclusive
V - verification - reviewed by 2 or 3 other analysts

Question 10

McKie and Ross 1997

Answer 10

• fingerprint found in murder victims home, identified as belonging to detective constable on case
• McKie denied being inside the house, stated categorically that it wasn’t her print
• later charged and tried unsuccessfully for perjury
• changed methodology of fingerprint examination in Scotland
• lead to Independent Fingerprint Enquiry

Question 11

Independent Fingerprint Enquiry

Answer 11

• fingerprint evidence should be recognised as opinion evidence, not factual
• results should stop being presented with 100% confidence
• ACE-V process should be detailed and enforced
• features should be demonstrable to a lay person with normal eyesight
• match should not be made if there are unexplained differences

Question 12

Ask and Granhag 2005

Answer 12

• conducted experiment with Swedish police officers to investigate effect of confirmation bias on evaluation processes
• case involving murder of female psychiatrist, another woman (Eva) found injured at scene
• Eva’s husband was a client of psychiatrist, had expressed suspicion of an affair, evidence strongly indicated Eva as perpetrator
• police officers asked to rate reliability and credibility of witness statement
• 2 versions: two loud female voices heard shouting upstairs, one male and one female voice heard shouting upstairs
• police officers who received statement involving man rated it a lot lower than those receiving female statement
• led to believe Eva was perpetrator, statement which opposes this was given less credit because it did not conform with previous judgement

Question 13

Powell et al. 2012

Answer 13

• investigated if giving prior information to interviewers impacted their interview technique and if this changed in skilled and non-skilled interviewers
• interviews with children to discuss an event they participated in at school
• half of interviewers given details about event and activities undertaken
• non-skilled interviewers with prior knowledge asked leading questions to retrieve the information they already knew or expected = confirmation bias
• skilled interviewers used some leading questions but less than non-skilled
• shows that contextual information can affect interviewing style and encourage use of leading questions
• also shows trained interviewers were less bias, so training is important to reduce affect

Question 14

Van den Eeden et al. 2017

Answer 14

• investigates if information CSIs are given before entering scenes affects their analysis technique
• experienced Dutch CSI assessed mock crime scene in which female victim found dead in home hanging from stairwell
• before entering, some were told it was suicide, some told it was murder
• CSI investigating thinking it was murder collected significantly more evidence than in condition 1 and found certain things more significant
• shows prior knowledge and information influences how extensive investigation is = contextual bias
• e.g. CSIs who believe the victim has committed suicide wont conduct as thorough examination and may not recover evidence that could indicate a murder
• contextual bias influences what CSIs believe to be important indicators or forms of evidence
• e.g. if believed to be suicide they will consider the presence of anti-depressant drugs very important, if believed to be murder a near by knife could be the murder weapon and is very important