LEC19 Evidence Evaluation PDF

Evidence Evaluation LEC19 Monday November 25th Foundations in Forensic Science  Labs Completed! Sets Available Online Labs 1 & 2 Course Update Due Wednesday December 4th at 2359hrs Final Exam D...

Evidence Evaluation LEC19 Monday November 25th Foundations in Forensic Science  Labs Completed! Sets Available Online Labs 1 & 2 Course Update Due Wednesday December 4th at 2359hrs Final Exam December 14th 1100hrs in Athletic Complex Verbal Statements “exclude” Exists in exclusions “consistent with” Scientists: unquantified positive evidential value Lay person: evidence supports hypothesis “does not exclude” & “fail to exclude” Lacks negative support Expert witness will place evidence in context of value using their experience Evidence Types Two categories: 1. Evidence where it is not an integral part of the offender and suspect Examples: clothing, possession of a vehicle Absence of characteristic does not mean not guilty Presence of characteristics does not mean guilty 2. Evidence is an integral part of the offender and suspect Examples: DNA, tattoo Absence of characteristic means not guilty Presence of characteristic does not mean guilty Evidence Value Example Eyewitness Tattoo on left hand Forensic Scientists Brown sweater based on fibre evidence Total population = 100,000 100 people tattoo on left hand 10,000 people brown sweaters Detain suspect fits both descriptions Total population = 1,000,000 Problems with this approach Changes with population size Becomes weaker with larger population Calculating the probability suspect being offender Probability of guilt given the evidence We aren’t examining actual VALUE of evidence How do we measure strength of evidence? Examine probability of guilt without any evidence Evidence Value Example Eyewitness Tattoo on left hand Forensic Scientists Brown sweater based on fibre evidence Total population = 100,000 100 people tattoo on left hand 10,000 people brown sweaters Detain suspect fits both descriptions Total population = 1,000,000 Problems with this approach Effect of evidence value on probability remains the same for all pops Assumptions: Every person in population equally likely Every person with shared characteristic equally likely Problem! Need the total population/characteristic pop How about frequencies?!  Evidence Value Example Eyewitness u Characteristic probability: Tattoo on left hand u 100 in 100,000 = 0.001 Forensic Scientists u Frequency of finding tattoo on left hand Brown sweater based on fibre evidence u Innocence probability: Total population = 100,000 u 99 in 99,999 = 0.00099 (or ≈ 0.001) 100 people tattoo on left hand 10,000 people brown sweaters u Are we measuring strength of evidence? Detain suspect fits both descriptions Total population = 1,000,000 Evidence Value Example Probability of seeing the evidence Tattoo on left hand = 0.001 Probability of seeing the evidence if suspect is offender Tattoo on left hand = 1 Only 1 left hand, that hand has a tattoo They both have a tattoo on their left hands Probability of evidence if suspect not offender Tattoo on left hand ≈ 0.001 1 is 1000 greater than 0.001 This is for evidence where suspect and offender characteristic matches! Evidence Value Example Eyewitness Tattoo on left hand Forensic Scientists Brown sweater based on fibre evidence Total population = 100,000 100 people tattoo on left hand 100 people brown sweaters Evidentiary value the same? 1/1000 Tattoo must carry greater weight  “…evidence evaluation is about ‘the probability of Evidence evidence’ in the light of competing hypotheses.” Evaluation -David Luci (2015) Significance Testing Can better measurements solve the problem? Significance Testing Can better measurements solve the problem? No! Still valuable Compares means of samples to population means (Not whether same pane) “Proximity of observations does not by necessity equal identity between objects.” This is the reason why “match” is incorrect! Relevance Evidence evaluation Consider common sense! What connects evidence to conclusions? RELEVANCE! ‘x is relevant to y for our purposes if x contributes towards proving or disproving y’ An inherent association between x and y Two outcomes = either relevant or not relevant Investigators create initial relevance Court must agree with relevance Relevance Constitutive facts Facts we have been structuring our hypothesis around Fact of innocence or guilt the ultimate issue Evidential Facts Material facts which by some leap of inference lead to a constitutive fact E.g. owning an item that is thought to have been involved in a crime (crowbar) Inferential leap to connect suspect to incident Does not mean they caused the incident Had to have been used & suspect owns one Statistics and the role of the courts Relevance Investigators create initial relevance Court must agree with relevance Measures of relevance (statistics) Shouldn’t this be done by courts? Varies greatly Depending upon evidence Investigators create statistics based on their expertise/science Explain to the courts Things to Remember! Evidence evaluation Uses probabilities Based on frequencies or population sizes Make sure to calculate correct probability Assumptions are made Significance testing Valuable for determining differences Cannot confirm “matches” Cannot assume meaning of comparative similarity