Read on Twitter
And now we start day 2 of @LegalAidNYC DNA Unit's 2021 Questioning Forensic Conference. First up, my colleague (and LAS Homicide Unit DNA Attorney) Allison Lewis and Duke Law Professor Brandon Garrett discuss forensic science litigation myths and jury research #QF2021 #DNA
Professor @brandonlgarrett wrote Convicting the Innocent and Autopsy of a Crime Lab: Exposing the Flaws in Forensics #DNA #QF2021
For litigators: challenging these myths is important not only for juries, but also for preserving appeal. Appellate courts love saying "harmless error" based on preconceptions about what they think would or wouldn't have mattered to juries. #DNA #QF2021
Myth #1: forensic result is too strong to challenge. In (non-DNA) pattern disciplines, experts give certainty without any objective measure of accuracy. The analysis and conclusions are circular. #DNA #QF2021
Read about Brandon Mayfield's case, where the FBI/DOJ falsely prosecuted Mr. Mayfield for the Madrid train bombings based on what it claimed was a 100% fingerprint match (which was "fabricated and concocted by the FBI and DOJ"). http://america.aljazeera.com/opinions/2014/2/the-terrifying-surveillancecaseofbrandonmayfield.html
Prof. Garrett "For all prosecutors complain about the CSI effect, the truth is exactly the opposite." People assume that forensic evidence is infallible, and this has to be challenged. #DNA #QF2021
Judges instructing juries on potential error rates in forensic science can have a meaningful effect on jury outcomes. #DNA #QF2021
Inconclusive error rates are "really high" in firearms toolmarks analysis and fingerprint analysis. Lawyers shouldn't stop looking into evidence because the analyst called it "inconclusive," because the evidence might actually be exculpatory. #DNA #QF2021
Myth #2: "control the language, control the verdict." Idea that by limiting language you can prevent jury from being misled about evidence. In reality, jurors might still have common misconceptions about strength of forensic evidence even with limiting language. #DNA #QF2021
(note: media spends a lot more time convincing the public that forensic evidence solves crimes than it does on cautionary stories like Brandon Mayfield or Lukis Anderson, who were falsely implicated by forensic evidence)
When juries hear the possibility of error, especially if for the first time on cross, it can have a profound impact. #DNA #QF2021
Juries really want to know "how good is this expert?" Juries assume a match is a match unless they're told otherwise. #QF2021 #DNA
Myth #3: no defense expert at trial, I can get whatever I need on cross (and assume a burden). In studies, to hear from defense expert - even to say it's inconclusive - had a huge effect. #DNA #QF2021
Even an expert who didn't examine evidence, but testified about methods and what can go wrong, had a huge impact in mock juror studies in reducing conviction rates. #QF2021 #DNA
Myth #4: shouldn't bore jury to death cross by going too deep into complicated forensic science issues. Studies show that the more lawyer unpacks these techniques, the more the jury realizes there's room for error. #DNA #QF2021
Myth #5: lab too big to fail, I can't take on the entire lab because it has fancy accreditation. Counterpoint: some of the biggest labs in the country (in just about every state) have had scandals involving thousands of cases, and it's not revealed by accreditation. #DNA #QF2021
Verification by supervisors comes with bias: work is only "verified" after analyst has found a match, so the review isn't blind. #DNA #QF2021
Myth #6: cognitive bias is irrelevant if the result is too strong. Cognitive bias is always important! #DNA #QF2021
Myth #7: the science has always been admitted, so it should be admitted in this case. Reality: courts are moving toward a better understanding of reliability and other issues with forensic evidence, and that they should be scrutinized. #DNA #QF2021
Keith Harward was falsely convicted after six prosecution "bite mark experts" claimed that he matched bite marks found on a murder victim. He was exonerated and released after spending more than half his life in jail. Listen to his story here #DNA #QF2021 https://www.law.upenn.edu/live/profiles/1453-keith-harward
Back from break with Manfred Kayser, Professor of
Forensic Molecular Biology at Erasmus University Rotterdam, to discuss forensic DNA phenotyping - predicting a person's appearance based on DNA. #DNA #QF2021
Forensic Molecular Biology at Erasmus University Rotterdam, to discuss forensic DNA phenotyping - predicting a person's appearance based on DNA. #DNA #QF2021
Currently, can predict categorical eye color, hair color, and skin color from forensic DNA, because genes are known with enough trait variance explained, DNA markers are known with predictive effects, and sufficiently accurate tests and models exist. #DNA #QF2021
Predictions models exist for eyebrow color, hair structure, tall stature, freckles, male hair loss (lol) but don't exist for a wide variety of characteristics. #QF2021 #DNA
Unresolved issue: some appearance traits depend on ancestry and/or age (like hair color), requiring combination of appearance, ancestry, and age in a single test. #DNA #QF2021
Cannot predict face from DNA (like a composite sketch) based on current DNA phenotyping tests #DNA #QF2021
Two parts of our genome are very useful for determining ancestry, because they don't change from generation to generation: mitochondrial DNA (maternal side) and Y-DNA (paternal side) #DNA #QF2021
(this science is really dense, but really interesting! mtDNA can lead us to the "biological Eve" - the common maternal ancestor of all humans on Earth) https://www.livescience.com/mitochondrial-eve-first-human-homeland.html
