P3 - C.P. - Collection of Evidence (BIO) Flashcards
What is Forensic Evidence?
Information that is collected at a Crime Scene. This can include:
* blood splatters
* shoe prints
* tyre tracks
* hair and fibre samples
* fingerprints
This is collected and presented as Evidence in Court.
Fingerprints especially are easy to collect and analyse, and every person has an individual fingerprint unique to them.
Forensic investigation should be Objective and Scientific, however it is not accurate 100% of the time.
The case of Mayfield
- In March 2004 there were terrorist-related bomb attacks on four public trains in Madrid.
- Very high Emotion and Profile due to number of victims
- International investigation including the FBI in the US.
- Using fingerprints found on Plastic Bag of the bomber, the FBI suspected Mayfield, a Muslim convert who had been of interest to the FBI since 9/11.
- A number of different experts all confirmed that it was Mayfield’s fingerprint on the bag.
- The Spanish police eventually found the real bomber, an Algerian terrorist.
Mayfield was cleared and offered a formal apology from the US government.
False Evidence
Hampikian (2011) conducted a review of trial transcripts and found that:
* 35% of blood analysis evidence was incorrect
* 22% of hair comparisons were incorrect
* 3% of bite marks were incorrect
* 2% of fingerprint analysis was incorrect
* Dror (2012) claims that despite Computers/Machine analysis, a Human makes the Judgement – Human Error responsible
The 2 Processes of Investigations
Dror (2012) describes that when analysing fingerprints, two processes take place:
* Bottom-Up - examination of the ridges and patterns of the fingerprints to identify the unique features of a person. However, if the prints are degraded or incomplete then the examiner must adopt the second process.
* Top-Down approaches use the expert’s previous experience and knowledge to make an assumption about the identity of the fingerprints.
Top-Down approach is open to mistakes from aspects of the expert’s personality, emotional state and expectations.
Cognitive Biases
- Expectancy bias – The Expert anticipates the Outcome and expects the Analysis to agree.
- Selective attention – Expectation of the Outcome leads to some aspects of the Analysis that could support an unexpected outcome being ignored.
- Conformity effect – If asked to Agree or Disagree with Others, the Expert won’t challenge them.
- Need-determination perception – If there is a Strong Desire to solve a crime, the Expert may make Conclusions with Insufficient Evidence
- Overconfidence bias – Overconfident Experts value hunch over Others’ Critiques.
- Confirmation Bias - Wanting an Explanation for something already Believed In
Mayfield Case - Confirmation Bias
Factors that cause bias
Dror (2006) tested what can cause Cognitive Biases in Fingerprint Analysts.
* 5 Experts with ave. 17 yrs of experience were studied for 12 months as part of their everyday work. They were unfamiliar with the Mayfield case so were less likely to show Demand Characteristics.
* Each examined a pair of fingerprints to see if there was a Match.
* Each had actually seen the fingerprints 5 yrs previously on a Real Case and Recorded it as a definite match.
* Told they were Wrongly Identified and No Match was expected.
* 4/5 changed their decision that the fingerprints were matched. 3 claimed it a Definite Mismatch.
Shows that Contextual information stops experts from giving Reliable Information; Confirmation Bias.
Dror (2005) - Emotional states
In another study, Dror et al. (2005) manipulated the participants’ Emotions, rather than expectations.
* 27 Uni Students (non-fingerprint experts) were given 96 pairs of fingerprints. Half were easy to match, the other half ambiguous (so Top-Down processes would become a factor).
* The fingerprints were related to crimes – either Low or High Emotion
* In some of the Emotional Cases, subliminal messages including the words ‘same’ or ‘guilty’ were flashed up on the computer screen participants were looking at the fingerprints on.
* Participants had to press either a button for ‘Same’ or ‘Different.
* More likely to find a Match in High Em. (58%) than Low Em. (49%).
* When the subliminal messages were added to the High Em. (66%).
Not involving Experts
Hall & Player - Aims
1) If a written report of a crime would affect a fingerprint expert’s analysis
2) If an emotional context would also affect the fingerprint expert.
Hall & Player - Participants
70 Volunteer fingerprint Experts from the Metropolitan Police Fingerprint Bureau. Experience Mean of 11 yrs (3 months - 30 yrs)
The majority worked in Active Teams, but some in Managerial Roles - all still registered as fingerprint Experts.
Hall & Player - Materials
- A fingerprint from a right forefinger was inked onto paper, then scanned into a computer. That fingerprint was then transposed onto a £50 note, with the detail of the note obscuring parts of the fingerprint.
- A sample of independent practitioners confirmed that the fingerprint was ambiguous and not clear.
- Magnifine Glass
- Russel Comparator (tool for analysing fingerprints)
Hall & Player - Procedure
- Part of their normal working day
- Told to investigate if a set of fingerprints matched the fingerprint on a £50 note
- No time limit and could leave the Bureau (but no discussion of the study)
- Along with the fingerprints and £50 note, participants were supplied with a routine crime scene examiner’s report that detailed aspects of the crime.
- 3 Groups: Low Em. (Forgery), High Em. (double gunshot into a victim), Chose Not to Read the Report
After Analysis: had to state either: 1) matched with the suspect’s fingerprints 2) did not match the fingerprints 3) did not have enough detail to analyse or 4) some agreement but not enough to match. - After, Questionnaire about how they came to their decision, whether they read the crime report and whether the crime report affected their analysis.
Hall & Player - Results
There was no significant difference found between the high and low emotional context groups.
* 57/70 participants read the crime scene report.
* 30 of these read the high emotional context report.
* 52% w/ High Em. thought the context affected their decision - 6% w/ Low Em.
Hall & Player - Conclusions
- Emotional context does not reduce a fingerprint expert’s ability to make a final decision.
- Unlike previous research, they concluded Em. Context didn’t effect Experts, however previous research used Students.
- Some of the experts chose not to read the crime report at all as this was not necessary in establishing a match.
This suggests that experts are capable of analysing fingerprints in a non-detached objective manner.
APPLICATION - How to reduce bias in the collection of forensic evidence
Although experts have been shown to be immune to bias in the Key Research, there is still a need to ensure it doesn’t affect them in the field.
Dror (2012) and Kassin et al. (2013) identified several ways to reduce bias:
* Educate Judges and Juries into the strengths and weaknesses of Forensic Science so evidence from something like fingerprints is not taken as the Ultimate Truth.
* Train Forensic Examiners in bias and how it can affect them.* If they are consciously aware that they might be biased they might be more objective*.
* Blind testing – do not tell Forensic Examiners information about the case (such as an emotional context) that could influence them.
* If the Expert is verifying a prior decision, do not tell them what the previous expert declared to avoid Conformity Bias.
APPLICATION - Falsification
Any good scientist tries to falsify ideas to leave them with the truth.
* Forensic Examiners could try to alter the way they examine evidence like fingerprints, trying to eliminate all of the people who didn’t commit the crime - reduces need-determination bias
* Overconfidence bias could be tackled by asking Experts, no matter how experienced, to always consult with a team to discuss their findings.
