PROBABILITY AND STATISTICS IN THE LAW Philip Dawid University College London.

Slides:



Advertisements
Similar presentations
Forensic DNA Inference ICFIS 2008 Lausanne, Switzerland Mark W Perlin, PhD, MD, PhD Joseph B Kadane, PhD Robin W Cotton, PhD Cybergenetics ©
Advertisements

Juror Understanding of Random Match Probabilities Dale A. Nance Case Western Reserve University August, 2007.
Attaching statistical weight to DNA test results 1.Single source samples 2.Relatives 3.Substructure 4.Error rates 5.Mixtures/allelic drop out 6.Database.
In 1999, Sally Clark was convicted of the murder of her two sons. The data: In 1996, her first son died apparently of cot death at a few weeks of age.
How strong is DNA evidence?
What Went Wrong in the Case of Sally Clark?
Bayes’s Theorem and the Weighing of Evidence by Juries Philip Dawid University College London.
ICFIS, Leiden 21 August 2014 Norman Fenton Queen Mary University of London and Agena Ltd Limitations and opportunities of the likelihood.
How TrueAllele ® Works (Part 3) Kinship, Paternity and Missing Persons Cybergenetics Webinar December, 2014 Mark W Perlin, PhD, MD, PhD Cybergenetics,
Slide 1 PGM 2012 The Sixth European Workshop on Probabilistic Graphical Models Granada, Spain 20 September 2012 Norman Fenton Queen Mary University of.
Bayesian Health Technology Assessment: An Industry Statistician's Perspective John Stevens AstraZeneca R&D Charnwood Bayesian Statistics Focus Team Leader.
1 Statistical genetics and genetical statistics Thore Egeland, Rikshospitalet and Section of Medical Statistics Joint work with P. Mostad, NR, B. Olaisen,
1 UCL. 14 Mar 05 UCL Monday 14 Mar 2005 Forensic inference – is the law a ass?  The Forensic Science Service 2004 Ian Evett Forensic Science Service.
Bayesian inference “Very much lies in the posterior distribution” Bayesian definition of sufficiency: A statistic T (x 1, …, x n ) is sufficient for 
Trial By Probability Bayes’ Theorem in Court. Presented By... Dave Bucheger Jill Thompson Sally Danielson Justin Koplitz Eric Hartmann.
Evaluation and interpretation of crime forensic evidence Crime Trace recovery Potential sources of the traces scenarios producing the traces Evaluation.
Presented By: Syeda Saleha Raza. A young girl, Lulu, has been found murdered at her home with many knife wounds. The knife has not been found. Some bloodstains.
Statistics Introduction.
DNA fingerprinting.
Fundamentals of Forensic DNA Typing Slides prepared by John M. Butler June 2009 Appendix 3 Probability and Statistics.
Chapter Seventeen HYPOTHESIS TESTING
Representing and Solving Complex DNA Identification Cases Using Bayesian Networks Philip Dawid University College London Julia Mortera & Paola Vicard Università.
DNA Forensics MUPGRET Workshop. “DNA evidence…offers prosecutors important new tools for the identification and apprehension of some of the most violent.
Philip Dawid University of Cambridge TexPoint fonts used in EMF.
Introduction to Hypothesis Testing CJ 526 Statistical Analysis in Criminal Justice.
Teaching Probability and Statistics to Law Students Philip Dawid University College London TexPoint fonts used in EMF. Read the TexPoint manual before.
Forensics Chapter 8. Central Points  DNA testing can determine identity  DNA profiles are constructed in specialized laboratories  DNA profiles used.
Lecture 9: p-value functions and intro to Bayesian thinking Matthew Fox Advanced Epidemiology.
Forensic Statistics From the ground up…. Basics Interpretation Hardy-Weinberg equations Random Match Probability Likelihood Ratio Substructure.
The Inexpert Witness Born 1933 Distinguished paediatrician Famous for “Munchausen Syndrome by Proxy” Expert witness in cases of suspected child abuse and.
Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc Chapter 9 Introduction to Hypothesis Testing.
Inference in practice BPS chapter 16 © 2006 W.H. Freeman and Company.
CHP400: Community Health Program - lI Research Methodology. Data analysis Hypothesis testing Statistical Inference test t-test and 22 Test of Significance.
3 Person Mixture #4 (Or is it 2?) The Hardest Mixture I Know.
DNA Identification Science: The Search for Truth Cybergenetics © Summer Research Symposium Duquesne University July, 2010 Mark W Perlin, PhD,
Chapter 9 Power. Decisions A null hypothesis significance test tells us the probability of obtaining our results when the null hypothesis is true p(Results|H.
DNA evidence The DNA Double Helix Consists of so-called nucleobases always in pairs A-T, C-G. One part of the pair is inherited from the mother, the other.
Florida State University College of Law Research Center Statistics Concepts for Law Students Fall ‘08 Workshop Jon R. Lutz.
Florida State University College of Law Research Center Statistics Concepts for Law Students Spring 2012 Workshop Jon R. Lutz
Exeter University 10 October 2013 Norman Fenton Director of Risk & Information Management Research (Queen Mary University of London) and CEO of Agena Ltd.
Thinking About DNA Database Searches William C. Thompson Dept. of Criminology, Law & Society University of California, Irvine.
TrueAllele ® Genetic Calculator: Implementation in the NYSP Crime Laboratory NYS DNA Subcommittee May 19, 2010 Barry Duceman, Ph.D New York State Police.
Statistical Inference An introduction. Big picture Use a random sample to learn something about a larger population.
Bayesian Inference, Review 4/25/12 Frequentist inference Bayesian inference Review The Bayesian Heresy (pdf)pdf Professor Kari Lock Morgan Duke University.
Cybergenetics Webinar January, 2015 Mark W Perlin, PhD, MD, PhD Cybergenetics, Pittsburgh, PA Cybergenetics © How TrueAllele ® Works (Part 4)
Ku San! (greetings in Girawa) Do Now: – Take out HW Agenda: – Identification and Comparison – Wayne Williams Case HW – p103 #11-14 and p104 #1-3.
Unleashing Forensic DNA through Computer Intelligence Forensics Europe Expo Forensic Innovation Conference April, 2013 London, UK Mark W Perlin, PhD, MD,
Statistics and the Law The case of the negligent nurse Willem R. van Zwet, University of Leiden Bahadur lecture Chicago 2005.
Reasoning with Probs How does evidence lead to conclusions in situations of uncertainty? Bayes Theorem Data fusion, use of techniques that combine data.
Slide 1 UCL JDI Centre for the Forensic Sciences 21 March 2012 Norman Fenton Queen Mary University of London and Agena Ltd Bayes and.
Sally Clark. Sally Clark (August 1964 – 15 March 2007) was a British lawyer who became the victim of a miscarriage of justice when she was wrongly convicted.
Murder in McKeesport October 25, 2008 Tamir Thomas.
AP Statistics Section 11.2 A Inference Toolbox for Significance Tests.
- 1 - Outline Introduction to the Bayesian theory –Bayesian Probability –Bayes’ Rule –Bayesian Inference –Historical Note Coin trials example Bayes rule.
Individual Identity and Population Assignment Lab. 8 Date: 10/17/2012.
Introduction to Hypothesis Testing Chapter 10. What is a Hypothesis? A tentative explanation for an observation, phenomenon, or scientific problem that.
BY K.UDAY KIRAN REDDY. CONTENTS Historical background What is DNA Fingerprinting Structure of DNA Making DNA Fingerprints Practical applications of DNA.
DNA Profiling LOs All pupils will: Recall the uses of DNA profiling.
DNA Forensics Bio Interpret how DNA is used for comparison and identification of organisms.
Validating TrueAllele® genotyping on ten contributor DNA mixtures
Error in the likelihood ratio: false match probability
DNA Forensics Bio Interpret how DNA is used for comparison and identification of organisms.
Explaining the Likelihood Ratio in DNA Mixture Interpretation
Probability.
Forensic inference – is the law a ass?
Daniela Stan Raicu School of CTI, DePaul University
Forensic match information: exact calculation and applications
DNA Identification: Mixture Interpretation
David W. Bauer1, PhD Nasir Butt2, PhD Jeffrey Oblock2
Reporting match error: casework, validation & language
Presentation transcript:

PROBABILITY AND STATISTICS IN THE LAW Philip Dawid University College London

STATISTICS = LAW Interpretation of evidence Hypothesis testing Decision-making under uncertainty

INGREDIENTS  Prosecution Hypothesis  Defence Hypothesis  Evidence

– or posterior odds: BAYESIAN APPROACH FREQUENTIST APPROACH and Find posterior probability of guilt: Look at & effect on decision rules

SALLY CLARK Sally and Stephen Clark’s sons Christopher and Harry died suddenly at ages 11 and 8 weeks, in Sally’s care The Clarks claimed that their children had died from natural causes (SIDS??) Contested prosecution medical evidence of maltreatment –SALLY CONVICTED OF MURDER

A paediatrician testified that, for a family like the Clarks, the probability of one child dying from SIDS is 1 in 8,543 At Trial: He was asked if the report calculated “the risk of two infants dying in that family by chance.” Answer: Yes, you have to multiply 1 in 8,543 times 1 in 8,543 …. [the CESDI study] points out that it’s approximately a chance of 1 in 73 million

WHAT TO THINK? Clear intuitive argument against independence (and thus calculation of “1 in 73 million”) BUT probability of 2 natural deaths remains very small HOW TO CONSIDER?

Prosecutor’s Fallacy = 1 in 73 million Probability of deaths arising from natural causes is 1 in 73 million = 1 in 73 million Probability of innocence is 1 in 73 million

Alternatively… P(2 babies die of SIDS) = 1/73 million P(2 babies die of murder) = 1/2000 million BOTH figures are equally relevant to the decision between the two possible causes

BAYES: POSTERIOR ODDS = LIKELIHOOD RATIO  PRIOR ODDS If prior odds = 1/2000 million posterior odds = m ??

IDENTIFICATION EVIDENCE Assume “match probability” Individual i Criminal SuspectEvidence: Match

PROSECUTOR’S ARGUMENT The probability of a match having arisen by innocent means is 1/10 million. So= 1/10 million – i.e.is overwhelmingly close to 1 – CONVICT

DEFENCE ARGUMENT Absent other evidence, there are 30 million potential culprits 1 is GUILTY (and matches) ~3 are INNOCENT and match Knowing only that the suspect matches, he could be any one of these 4 individuals So –ACQUIT

BAYES  POSTERIOR ODDS = (10 MILLION)  “PRIOR” ODDS  PROSECUTOR’S argument OK if Only BAYES allows for explicit incorporation of B  DEFENCE argument OK if

The Island Problem N+1 on island: N (100) innocent, 1 guilty Match, probability = P (0.004) Prosecution: Defence: (0.996) (0.714)

Other Arguments Let number of individuals i having I i = x be M – need distribution of M given Note: Initially So

Argument 1 Evidence tells us So (0.902)

Argument 2 Evidence tells us 1 (guilty) individual has x Our of remaining N innocents, number with x is ; while So (0.824)

Argument 3 Evidence E is equivalent to 2 successes on 2 Bernoulli trials with replacement So Then (0.714 – as for defence)

DENIS ADAMS –Match probability = 1/200 million 1/20 million 1/2 million  Doesn’t fit description  Victim: “not him”  Unshaken alibi  No other evidence to link to crime Sexual assault DNA match

BAYES’S THEOREM POSTERIOR ODDS on guilt = LIKELIHOOD RATIO  PRIOR ODDS = 2 million  (1 / 200,000) = 10 (10:1)  Posterior probability of guilt = 10/11 = 91% Reasonable doubt – ACQUIT

WHAT ABOUT OTHER EVIDENCE? Didn’t fit description Victim: “not him” Unshaken alibi LR = 0.1 / 0.9 = 1/9 LR = 0.25 / 0.5 = 1/2 Apply Bayes’s Theorem again: Final odds on guilt = 10  1/9  1/2 } = 5/9 (5:9) (probability of guilt = 5/14 = 35%)

Dependence on Match Probability Match probability1/200m1/20m1/2m Posterior probability of guilt 98%85%35% – number of noughts does matter!

DATABASE SEARCH Crime trace, frequency (match probability) 1 in 1 million Search Police DNA database (D) of size 10,000 Find unique match: “John Smith” (S) No other evidence

Defence Case Probability of finding a match in database if innocent ~ 10,000  (1/1,000,000) = 1/100 Match probability of 1/100 is not convincing evidence Evidence against John Smith is (significantly) weakened by virtue of database search – ACQUIT

Prosecution Case We have examined 10,000 individuals Of these, 9,999 found not to match This has reduced the pool of potential alternative culprits Evidence against John Smith is (marginally) strengthened by virtue of database search – CONVICT

Which likelihood ratio? Hypothesis H S : “John Smith did it” is data- dependent Replace by hypothesis H D : “Someone in database D did it” –equivalent after search identifies S (but not before) LR = 1/(match probability) is now only 100 –weak evidence? But H D is a priori 10,000 times more probable than H S –posterior odds the same! –agrees with prosecution argument

Multiple Stains 2 DNA stains –1 on sheet, 1 on pillow –assume 2 perpetrators, 1 stain from each John Smith (S) matches pillow stain –associated “match probability” P What are appropriate hypotheses, likelihoods, inferences?

Hypotheses S left one of 2 stains S left pillow stain S left neither stain S didn’t leave pillow stain (  = prior probability S is guilty)

What to present in Court? Hypotheses equivalent (only) after data Different prior odds Identical posterior odds

Mixed Stains Crime trace containing DNA from more than 1 contributor –Rape –Scuffle etc

O. J. SIMPSON Crime    OJS   RG   A B C Marker DQ-  Frequency 13% 20% 28% “MATCH” to OJS Allele

MATCH PROBABILITY? PROSECUTION: Frequency of OJS type AB: 5% DEFENCE: Combined frequency of all matching types AA, AB, AC, BB, BC, CC:39% LR approach assuming Goldman (AC) in mixture: AB, BB, BC:21% LR approach not assuming Goldman in mixture: (more complex calculation) ~ 21%

MISSING DNA DATA What if we can not obtain DNA from the suspect ? (or other relevant individual?) Sometimes we can obtain indirect information by DNA profiling of relatives But analysis is complex and subtle…

HANRATTY James Hanratty convicted and executed in 1962 DNA profile from crime items analysed in 1998 Population frequency less than 1 in 2.5 million DNA profiles from mother and brother – “consistent with” crime DNA being from Hanratty (“A6” murder and rape, 1961)

PRESS REPORTS “There is a 1 in 2.5 million chance that Hanratty was not the A6 killer” “The DNA is 2.5 million times more likely to belong to Hanratty than anyone else” Likelihood Ratio based on profiles of mother and brother (complex calculation): 440 –even though no direct match to Hanratty!

DISPUTED PATERNITY MOTHER (m1) of CHILD (c1) claims that PUTATIVE FATHER (pf) is its TRUE FATHER (tf) But DO have DNA profiles from: Two full BROTHERS (b1, b2) of PUTATIVE FATHER undisputed child disputed child brothers His UNDISPUTED CHILD (c2) and its MOTHER (m2) DNA profiles from MOTHER and CHILD  No profile from PUTATIVE FATHER

DECISION AID “PROBABILISTIC EXPERT SYSTEM” – embodies probabilistic relationships (between inherited genes)

ANALYSIS Measurements for 12 DNA markers on all 6 individuals Enter data, “propagate” through system Overall Likelihood Ratio in favour of paternity: ~1300

FURTHER COMPLEX DNA CASES Contamination Laboratory errors, mix-up, fraud Relatives –…

Statistics Law Crime Science Psychology Economics Philosophy of Science Geography Medicine Ancient History Computer Science Education … EVIDENCE, INFERENCE AND ENQUIRY

EVIDENCE SCIENCE Subject- and substance-blind approach Inference, explanation, causality Recurrent patterns of evidence Narrative, argumentation, analysis, synthesis Cognitive biases Formal rules Decision aids Interdisciplinary studies …