Collection and Processing of forensic evidence

• Dror suggested that there are 4 types of bias:

  1. perceptual bias - perception may be distorted,
  2. conformity bias - we are likely to just agree with what others say (e.g. Brendon Mayfield),
  3. expectancy bias - our own stereotypes impact how we look for evidence,
  4. motivational bias - emotion of a crime making an analyst feel more pressured to find a match
• Dror researched university students and got them to analyse fingerprints either in a crime involving harm or no harm. In the harm condition the students were more likely to find a match. This shows that emotional context creates a motivational bias in fingerprint analysis.
• WWF - Hall & Player

  • Whether the written report of a crime impact a fingerprint experts interpretation of a poor quality mark and 2) whether the fingerprint experts emotionally affected by the circumstances of the case.
  • Studied 70 fingerprint experts working for the metropolitan police fingerprint bureau. Mean length of experience was 11 years.
  • 52% of the 30 who read the high emotion context felt affected by the information of the report compared to 6% of who read the low emotion context. No significant differences were found between the amount of matches found between the 2 groups.
• How - Hall & Player

  • Lab experiment but as realistic as possible - it took place in the New Scotland Yard but the task was artificial.
  • Randomly allocated to 2 groups: low emotion context (fraud)/ high emotion context (murder)
  • Volunteer's right forefinger was scanned and put onto a £50 note - partially obscured. Ppts had to say if the latent print was a match to the scan of the prints they were given.
  • Ppts were asked: if they read the crime info prior to analysis; if print was match, not match, or not enough evidence; and whether the ppt would use the analysis in court.
• Conclusions - Hall & Player

  Emotional context affects a fingerprint expert's confidence of the analysis but this does not have any impact on their final decision. Showing highly trained professionals are not impacted by bias.
• Improving education

  • cognitive experts should run training courses where they teach forensic experts about different types of bias, as identified by Dror, and given strategies to reduce bias. e.g. not read case info, and ongoing training on cognitive bias.
  • Dror identified a number of cognitive biases which can impact analysis of forensic evidence.
• Sequential unmasking

  • examiners can first analyse the forensic evidence without any information from the case, case managers can then screen case files to filter necessary information that can be given to the analysts.
  • Hall & Player's research found 52% of forensic analysts who read high emotional context (murder) felt emotionally impacted. This may lead to motivational bias where the analysts feel pressured to find a match.
• Falsification techniques

  • New protocols could be introduced to forensic teams where falsification is the last step of the forensic process. Experts should be asked to evaluate the possibility of other matches and prove there are no other possible matches.
  • This provides a scientific approach to analysis of forensic evidence as it means the analysis is more objective and is unbiased. Also reduces expectancy bias (stereotypes).