Cover Image

Handbook of forensic statistics /

Handbook of Forensic Statistics is a collection of chapters by leading authorities in forensic statistics. Written for statisticians, scientists, and legal professionals having a broad range of statistical expertise, it summarizes and compares basic methods of statistical inference (frequentist, lik...

Full description

Bibliographic Details
Corporate Author: Taylor & Francis
Other Authors: Banks, David L. (Editor), Kafadar, Karen (Editor), Kaye, D. H. (David H.), 1947- (Editor), Tackett, Maria (Editor)
Format: eBook
Language:English
Published: Boca Raton : Chapman & Hall/CRC, 2021.
Series:Chapman & Hall/CRC handbooks of modern statistical methods.
Subjects:
Forensic sciences > Statistical methods.
Forensic statistics.
Criminalistique > Méthodes statistiques.
Statistiques légales.
LAW > Criminal Law > General.
LAW > Forensic Science.
MATHEMATICS > Probability & Statistics > General.
Forensic sciences > Statistical methods
Forensic statistics
handbooks.
Handbooks and manuals.
Guides et manuels.
Online Access:Connect to the full text of this electronic book
Table of Contents:
  • Cover
  • Half Title
  • Series Page
  • Title Page
  • Copyright Page
  • Dedication
  • Contents
  • Foreword
  • Preface
  • Editors
  • Contributors
  • Section I: Perspectives on Forensic Statistics
  • 1. The History of Forensic Inference and Statistics: A Thematic Perspective
  • 1.1 Introduction
  • 1.2 Forensic Science and the Evaluation of Evidence
  • 1.3 The Need for an Interpretative Model
  • 1.4 Support of Judicial Disciplines for a Scientific Presentation of the Value of Evidence
  • 1.5 Probability of Proposition Given Evidence and of Evidence Given Proposition
  • 1.6 Quantification of the Value of Evidence Using Alternative Numerical Summaries
  • 1.7 Change from Two-Stage Approach to Continuous Approach
  • 1.8 Presentation of Evidence: New Challenges to Solve
  • 1.8.1 The Island Problem and Results of a Database Selection
  • 1.8.2 Profile Probability vs Conditional Profile Probability
  • 1.8.3 Evaluation by Taking Errors into Account
  • 1.9 A Minimum Value for the Profile Probability
  • 1.10 Propositions and Pre-Assessment
  • 1.10.1 The Choice of Propositions
  • 1.10.2 The Pre-Assessment
  • 1.11 Translation of a Numerical Value into a Verbal Equivalent
  • 1.12 Assessment of Performance
  • 1.13 Role for Likelihood Ratio as aMeasure for Investigation as Well as for Evaluation
  • 1.14 Probabilistic GraphicalModels
  • 1.14.1 Bayesian Networks
  • 1.14.2 Bayesian Networks to Manage 'Masses' of Evidence
  • 1.14.3 Bayesian Networks in Judicial Contexts
  • 1.14.4 Bayesian Networks in Forensic Science: Particular Case Modeling
  • 1.14.5 Bayesian Networks in Forensic Science: Generic Patterns of Inference
  • 1.15 Not Only Inference: The Way to Make a Decision
  • 1.15.1 The Objectives and Ingredients of Decision Theory
  • 1.15.2 Graphical Models
  • 1.16 The Existence or Otherwise of a True Value of the Evidence
  • Acknowledgments
  • References.
  • Section II: General Concepts andMethods
  • 2. Frequentist Methods for Statistical Inference
  • 2.1 Introduction
  • 2.2 Definitions and Notation
  • 2.2.1 Data and Evidence
  • 2.3 Random Variables and Probability Distributions
  • 2.3.1 Sampling from a Distribution or Population
  • 2.4 Estimation
  • 2.4.1 Properties of Point Estimators
  • 2.4.2 Estimating Allele Proportions 2.4.2.1 A Point Estimate
  • 2.4.2.2 Constructing a Confidence Interval
  • 2.4.2.3 Choosing a Confidence Coefficient
  • 2.4.3 Estimating a False Positive Probability Through an Experiment 2.4.3.1 The Design of Experiments to Test Categorical Source
  • 2.4.3.2 An Experiment to Test Categorical Judgments of Latent Print Examiners
  • 2.4.3.3 Constructing Confidence Intervals
  • 2.4.4 Interpreting Confidence Intervals
  • 2.5 p-Values
  • 2.5.1 p-Values in a Comparison of Glass Fragments
  • 2.5.2 Interpreting p-Values
  • 2.6 Hypothesis Tests
  • 2.6.1 Classical Hypothesis Tests for Refractive Index Matching 2.6.1.1 Type I Errors and the Size of a Test
  • 2.6.1.2 Type II Errors and the Power of a Test
  • 2.6.2 Hypothesis Testing with p-Values
  • 2.6.3 Hypothesis Testing with Confidence Intervals
  • 2.7 Issues in Interpreting the Results of Hypothesis Tests, p-Values, and Confidence Coefficients
  • 2.7.1 Transposition
  • 2.7.2 Multiple Tests: Proof of the Null Hypothesis and Adjusted p-Values
  • 2.7.3 Arbitrary Lines
  • 2.7.4 Alternatives and Likelihoods
  • 2.8 Resampling Methods
  • 2.8.1 Bootstrap Estimates
  • 2.8.2 Permutation Tests
  • Acknowledgments
  • References
  • 3. Bayesian Methods and Forensic Inference
  • 3.1 Introduction
  • 3.2 The Basics
  • 3.2.1 A Beta-Binomial Mock Example
  • 3.2.2 A Gamma-Poisson Mock Example
  • 3.3 Markov Chain Monte Carlo
  • 3.4 Broad Applications
  • 3.5 Summary
  • Acknowledgments
  • References
  • 4. Comparing Philosophies of Statistical Inference.
  • 4.1 Inferential Philosophies
  • 4.1.1 Frequentist Inference
  • 4.1.2 Bayesian Inference
  • 4.1.3 Other Approaches to Inference
  • 4.1.3.1 Fiducial Inference
  • 4.1.3.2 Likelihood Inference
  • 4.1.3.3 Confidence Distributions
  • 4.2 Comparing the Approaches
  • 4.2.1 Planning Studies Using Frequentist Inference
  • 4.2.2 Challenges for Frequentist Inference
  • 4.2.3 Flexible Inference with Bayesian Methods
  • 4.2.4 Model Modifications and Adjustments
  • 4.2.5 The Prior Distribution and the Definition of Probability
  • 4.3 Relevance to Forensic Statistics
  • 4.3.1 Likelihood Ratios and Bayes Factors
  • 4.3.2 Two-Stage Procedures in Forensic Science
  • 4.3.3 Forensic Evidence as Expert Opinion and Error Rates
  • 4.4 Summary
  • References
  • 5. Decision Theory
  • 5.1 Introduction
  • 5.2 Concepts of Statistical Decision Theory
  • 5.2.1 Preliminaries: Basic Elements of Decision Problems
  • 5.2.2 Utility Theory
  • 5.2.3 Implications of the Expected Utility Maximisation Principle
  • 5.2.4 The Loss Function
  • 5.2.5 Particular Forms of the Expected Utility Maximisation Principle
  • 5.2.6 Likelihood Ratios in the Decision Framework
  • 5.3 Decision Theory in the Law and Forensic Science
  • 5.3.1 Legal Applications
  • 5.3.2 Forensic Science Applications 5.3.2.1 Forensic Identification
  • 5.3.2.2 Understanding Probability Assignment as a Decision: The Use of Proper Scoring Rules
  • 5.3.2.3 Other Forensic Decision Problems: Consignment Inspection
  • 5.4 Discussion and Conclusions
  • 5.5 Further Readings
  • 5.5.1 Forensic Science
  • 5.5.2 General
  • Acknowledgments
  • References
  • 6. Association Does Not Imply Discrimination: Clarifying When Matches Are (and Are Not) Meaningful
  • 6.1 Introduction
  • 6.2 Association and Discrimination
  • 6.2.1 Quality of Test: Sensitivity and Specificity
  • 6.2.2 Sources of Error
  • 6.2.3 Weight of Evidence: The Likelihood Ratio.
  • 6.2.4 Useful Databases for Ascertaining Discriminatory Power
  • 6.2.5 Conflating Conditional Statements: The Prosecutor's Fallacy
  • 6.3 Examples: The Discriminatory Power of Forensic Evidence
  • 6.3.1 Arson Investigation
  • 6.3.2 Other Types of Forensic Evidence: DNA, Fingerprints, and Shoe Prints
  • 6.3.3 Abusive Head Trauma
  • 6.4 Conclusion
  • References
  • 7. Validation of Forensic Automatic Likelihood Ratio Methods
  • 7.1 Introduction
  • 7.1.1 Scope
  • 7.1.2 Aim
  • 7.1.3 Structure
  • 7.2 Validation Process
  • 7.2.1 Standardization
  • 7.2.2 Validation of Theoretical and Empirical Aspects
  • 7.2.3 Performance Characteristics for Automatic LR Methods
  • 7.2.4 Empirical Validation
  • 7.2.5 Validation Protocol
  • 7.3 Primary Performance Characteristics
  • 7.3.1 Performance of Probabilities by Proper Scoring Rules
  • 7.3.2 Discrimination and Calibration of Probabilities
  • 7.3.3 Performance of Likelihood Ratios
  • 7.3.4 Properties of Well-Calibrated Likelihood Ratios
  • 7.3.5 Examples with Primary Performance Characteristics
  • 7.4 Secondary Performance Characteristics
  • 7.4.1 Robustness
  • 7.4.2 Monotonicity
  • 7.4.3 Generalization
  • 7.5 Conclusion
  • References
  • 8. Bayesian Networks in Forensic Science
  • 8.1 Introduction
  • 8.2 Probability Logic
  • 8.3 Simple Bayesian Networks for Forensic Problems
  • 8.4 Object-Oriented Bayesian Networks
  • 8.5 Forensic Genetics
  • 8.5.1 Bayesian Networks for Simple Criminal Identification
  • 8.5.2 Simple Disputed Paternity
  • 8.5.3 Bayesian Networks for Complex Criminal Cases Involving Family Relationships
  • 8.5.4 Mutation
  • 8.6 Bayesian Networks for Analysing Mixed DNA Profiles
  • 8.6.1 Discrete Features
  • 8.6.2 Continuous Features
  • 8.7 Analysis of Sensitivity to Assumptions on Founder Genes
  • 8.7.1 Uncertainty in Allele Frequencies
  • 8.7.2 Heterogeneous Reference Population
  • 8.8 Conclusions.
  • Appendix 8A: Bayesian Network Basics
  • 8A.1 Qualitative Structure
  • 8A.2 Independence Properties
  • 8A.3 Quantitative Structure
  • 8A.4 Computation
  • References
  • Section III: Legal and Psychological Dimensions
  • 9. How Well Do Lay People Comprehend Statistical Statements from Forensic Scientists?
  • 9.1 Methodological Overview
  • 9.2 Consistency
  • 9.2.1 Framing
  • 9.2.2 Format
  • 9.3 Sensitivity
  • 9.4 (In)Coherence
  • 9.4.1 Prosecutor's Fallacy
  • 9.4.2 Defense Attorney's Fallacy
  • 9.4.3 Directional Errors
  • 9.4.4 Aggregation Errors
  • 9.5 Ability
  • 9.6 Orthodoxy
  • 9.7 Discussion
  • 9.8 Conclusion
  • References
  • 10. Forensic Statistics in the Courtroom
  • 10.1 The Purpose, Form, and Prerequisites of Expert Testimony
  • 10.1.1 Lay and Expert Testimony
  • 10.1.2 Qualifications for Statistical Experts (and Experts Who Use Statistics)
  • 10.1.3 Forms of Statistical Expert Testimony
  • 10.1.4 Reasonable Scientific or Statistical Certainty
  • 10.2 Special Rules for Scientific Expert Testimony*
  • 10.2.1 The General-Acceptance Standard
  • 10.2.2 The Scientific-Validity Standard
  • 10.3 Selected Evidentiary Issues in Forensic Statistics
  • 10.3.1 Two Uses of Statistical Analysis as Evidence
  • 10.3.2 Theory and Application
  • 10.3.3 Error Rates in Determining Admissibility
  • 10.3.4 Error Rates, Likelihood Ratios, and Bayes Factors for Quantifying Probative Value
  • 10.4 Conclusion
  • References
  • Cases and Rules
  • Section IV: Applications of Statistics to Particular Fields in Forensic Science
  • 11. DNA Frequencies and Probabilities
  • 11.1 Introduction
  • 11.2 Likelihood Ratios
  • 11.3 Population Genetics
  • 11.3.1 Single Loci
  • 11.3.2 Multiple Loci
  • 11.3.3 Population Structure
  • 11.3.4 Lineage Markers
  • 11.4 Mixtures
  • 11.4.1 Semi-Continuous Model
  • 11.4.2 Continuous Model
  • 11.5 DNA Sequence Data
  • 11.6 Future Directions.