Week 12 reading 2: Kassin 2013; the forensic cofnriamtion bias; problems, perspective and proposed solutions

Question 1

 Brandon Mayfield and Madrid Bomber Case

Answer 1

is an example of the forensic sciences being subject to contextual bias and fraught with error
* Madrid Bombings in 2004 – series of bombs on 4 commuter trains in madrid
o Kills 191 people wounded 1800
* Identified a finderprint on the bag with detonating devises
o The FBI from this positively identified brnadon Mayfield and American Muslim
o Brandon had been on watch since 9/11 and was taken to court for the bombings
o Later the Spanish authories matched the prints to the real bomber Ouhnane Daoud
o FBI did an internal investigation and concluded confirmation bias was the cause of the erroneous identification

Question 2

 Forensice science errors

Answer 2

 Forensice science errors have also surfaced with alarming frequency in DNA exoneration cases and other wrongful convictions
* In cases where trial transcripts or reliable forensic science data were available for review, 38% contained incorrect serology testimony, which is highly regarded. In addition 22% involved hair comparisons; 3% incolved bite mark comparisons and 2% fingerprint comparisons

Question 3

• The criticism of forensics is twofold

Answer 3

o The realisation that too often the stimulus does not compel a perceptual judgement that is objective, hence there is concern for both inter-rater reliability across experts and for intra-test reliability over time within experts

• Often visual patterns are being compared for sufficient similarity however, determining this has no criteria or quantification
  o 10% of the time the same expert will produce different outcomes over time

Question 4

 Contextual influences

Answer 4

• Peer review processes
• Cross-communication
• Etc
• Can impinge forensic examinationation
• Data should be examined in isolation with no additional information

Question 5

o Accuracy and Error

Answer 5

 More than 300 DNA exonerations through the innocence project have found 2 issues
* 1. Forensic science judgements are often derived from inadequate testing and analysis if not outright fabrication
* 2. Experts often give imprecise or exaggerated testimony, drawing conclusion not supported in the data drawing charges of misconduct in some cases

Question 6

 The CSI effect

Answer 6

• TV programes communicate a false belief in the powers of forensic science which could be exacerbated when forensic experts overstate the strength of evidence
  o This is common especially considering the followsing
   1. Experits tend to be overconfident in their abilities
   2. The courts tend to blindly accept forensic science evidence
   3. Many forensic examinors work for police and advocate the prosecution
   4. Errors are not apparent in forensic science and the gorund truth is often not known
   5. Many forensic examiners consider themselves objective and immune to bias

Question 7

o Confirmation Bias

Answer 7

 Different types of confirmation biases whereby people seek, perceive, interpret and create new evidence in ways that verify their preexisting bliefs.
* Studies showed that prior exposure to images of a face or abody, an animal or a human, or letter or number, can bias what people see in an ambiguous figure.
* Classic research suggests the perception of a stimulus is not solely a function of the stimulus itself, but is also shaped by the qualities of the observer
o Eg. Bruner and Goodman 1947 asked children to estimate the size of coins from emmory and found that children to estimate the size of coins from memory and found that children of low-SES overestimation size of the coins to a greater degree than did the children of high SES
* To sum up; a wealth of evidence indicates that an observer’s expectation can impact visual and auditory perception – although similar effects can be driven by motivation – confirmation biases are a natural and automatic feature of human cognition that can occur in the absence of self-interest and operate without conscious awareness.

Question 8

 Social Perception Effects
* Strong expectancy can contaminate social perception

Answer 8

o Eg. Primacy effects; in impression formation by which information about a person presented early in a sequence is weighted more heavily that information presented later which is ignored, discounted or assimilated into early formed perceptions
o Perseverance effect then suggests that beliefs once they take food can persist even when evidence on which they are formed is discredited
o The fact people are jaded by existing beliefs has consequence in forensic settins;
 Eg. One study (Obrien, 2009) has participants review a mock police file of a crime investigation that contained weak circumstantial evidence ointing to a possible suspect. Some participants were asked to form an initial hypothesis and the other half were not. Those who made a hypothesis tended to then seek out new evidence to confirm this so a weak suspect became the prime suspect.
 Eg. Kassin Goldstein and Savitsky 2003 had some participants and not others commit a mock crime – everyone was then interviewed by interrogators who had been led to presume guilt or innocence. Those who presumed guilt asked more incriminating questions, had more coercive interrogations and tried harder to get. Aconfessions, this led to more defensive suspects and let to observors who later listened to interrogation tapes judge them as guilty, even when they were innocent.

Question 9

 3 step behavioural confirmation process

Answer 9

• A perceiver forms and impression of a target person, interacts in a manner that is consistent with that impression and causes the target person unwittingly to adjust his or her behaviour. The net result; a process that transorms expectations into reality

Question 10

 Cognitive and Motivational Sources of Bias

Answer 10

• Heurtistic thinking, whilst generally beneficial, but can produce systematic errors on judgement – especially where strong prior expectations exist.
  o Over time, and across a range of domains, basic psychological research has shown that strong expectations provide

Question 11

• Motivational Goals
  o Kunda 1990

Answer 11

 Argued that motivation influences reasoning indirectly as a result of 2 goals
* 1. Accuracy goals – strive to form accurate beliefs or judgements
* 2. Directional goals – where individuals seek a particular desired conclusion
o In this people maintain an illusion of objectivity that prevents them from recognising that their cognition has been tained by preference or desire
 Motivated reasoning is pervasice, so some being show self-serving positivity biases

Question 12

Balcetis and Dunning 2006

Answer 12

 Showed people an ambiguous figure – B or number 13
* Depending on the stimulus picked they’d get to drink orange juice
* Those told the letter would assign them to OJ – 72% saw the letter and those who were told the number would get them juice – 61% saw the number
* Was subconscious – wishful seeing

Question 13

• The forensic confirmation bias

Answer 13

o Tversky and Kahneman 1974 reasoned that confirmation bias effects could extend to the legal system – this has since been corroborates
 So have forensic confirmation bias as a term to summarise the class of effects through which an individual’s pre-existing beliefs, epxectations, motives and situational content influence the collection, perception and interpretation of evidence during the course of a criminal case.

Question 14

• Context effects

Answer 14

o Potential biaseing effect of expectation and concept on perceptual judgements made by forensic examiners
o Corroboration inflation
 Kassin 2012
* Suggests other strong contextual cues may bias forensic judgements in the crimal justice system
o Dror, Charlton and Peron 2006
 Shows fingerprint judgements may be subject to bias
* Asked 5 experienced experts to asses pairs of fingerprints, that unbeknownst to them, they had examined years earlier and declared to be a match
o Before stimuli were re-presented these examiners were told that the fingerprints were taken from a high-profile case of errornous identification, implying one was not a match.
o Given this biasing info, only one of the 5 experts judged the fingerprints to be a match, indicated that context undermined reliability
o This is particularly concerning as the change as a function of context was seen in experts
o Dror and Charlton 2006
 Followed up and presented 6 latent fingerprint experts with 8 pairs of prints from a crime scene and suspect in an actual case in which they had previously made a match or exclusion judgement.
 Prints were accompanies with either
* No extraneous info (control)
* Info the suspect confessed, suggesting a match
* Info the suspect was in custody at the time of the crime, suggesting exclusion
 Results showed contextual info produced a change in results in 17%
o Rosenthal 2008 estimated the reliability of fingerprint experts judgements over tiem likely falls in the range of 0.33 – 0.9 implying considerable degree of subjectivity

Question 15

 Elasticity of Forensic Evidence

Answer 15

• People are more likely to exhibit bias when evaluating ambiguous evidence
  o Fingerprints are always going to be slightly different due to variations in skin elasticity, the amount of pressure applied, the material the print was left on and how the prints were recovered etc
  o Therefore, fingerprint data is often ambiguous and incomplete – meaning experts are susceptible to bias in processing this
  o This occurseven in DNA evidence

Question 16

 Bias and self-insight

Answer 16

• Although confirmation bias is typically unconscious – there can be motivational, cognitive and emotional factors that guide job performance
• Charlton et al., 2010 showed in a semi-structured interview that many examiners have an interest in catching criminals and solving crimes – showed a strong need for closure – and the joy that comes with finding a print match. They also exchibited a fear of making erroneous judgements or committing a false-positive error that could implicate an innocent person

Question 17

o Implications for Accuracy and Error

Answer 17

 The fact that confressions and other strong bases for presumption of guilt can bias the search, collection, perception and interpretation of subsequently obtained evidence undermines a silent bu basic tenet of the judicial system – namely that the items of evidence presented at a trial are independent of one another
* Corroboration inflation = when one witness influences another and a strong bias is greated which creates momentum for more bias, aka the bias snowball effect by Dror 2012
* The influence of one witness or item of evidence on another witness or item of evidence consistues a biasing process of confirmation, one that can increase the likelihood or erros
* Eg. In a Texas arson-murder cas eyewitnesses changed their account once being told about forensic evidence suggesting the fire wasn’t accidental – although the forensic conclusion was later found to be erroneous

Question 18

o How to reduce bias

Answer 18

 There are 2 levels at which it is necessary to reduce bias and its consequences
* 1. At the crime scene or in the forensic lab where evidence is collected and sometimes analysed
o Examiners should work linear rather than circular – thus initially examining evidence from the crime scene and documenting their findings before making comparisons against a targe. This would eliminate the ptoentail influence of the target on how information is processed and the weight placed on it (so data cannot be re-assessed to fit the target)
o Eg. Blind testing – removing contextual variable – removing extraneous information that may taint conclusions
 Does not eliminate the base rate expectation that a sample is the perpetrators
o Verification of forensics decision
 Verifier should not know the initial conclusion or who the examiner was
 Cross-lab verifications are advisable – so independent corrobortation
o Technology
* 2. In the trial and appellate courts where evidence is evaluated
o Cofnrimation bias spawns three froblems
 1. Corrupts conclusions and testimony of forensic examiners
 2. Influences other lines of evidence eg. Witnessness, indcuging false confessions etc
 3. Biased sources of info are presented to judges, juries and appeal courts which rely heavily on forensic science to support decision making
o The harmless error doctring – that an erroneously admitted confession can prove harmless if other evidence is sufficient to support conviction – rests on the tacit and often icorrect assumption that alleged other evidence was independent of the erroneously admitted item