法科学者のための統計学(第2版)<br>Statistics and the Evaluation of Evidence for Forensic Scientists (Statistics in Practice) (2ND)

法科学者のための統計学(第2版)
Statistics and the Evaluation of Evidence for Forensic Scientists (Statistics in Practice) (2ND)

  • ただいまウェブストアではご注文を受け付けておりません。 ⇒古書を探す
  • 製本 Hardcover:ハードカバー版/ページ数 509 p.
  • 言語 ENG
  • 商品コード 9780470843673
  • DDC分類 363.23015195

Full Description


The first edition of Statistics and the Evaluation of Evidence for Forensic Scientists established itself as a highly regarded authority on this area. Fully revised and updated, the second edition provides significant new material on areas of current interest including: Glass Interpretation Fibres Interpretation Bayes' Nets The title presents comprehensive coverage of the statistical evaluation of forensic evidence. It is written with the assumption of a modest mathematical background and is illustrated throughout with up-to-date examples from a forensic science background. The clarity of exposition makes this book ideal for all forensic scientists, lawyers and other professionals in related fields interested in the quantitative assessment and evaluation of evidence. 'There can be no doubt that the appreciation of some evidence in a court of law has been greatly enhanced by the sound use of statistical ideas and one can be confident that the next decade will see further developments, during which time this book will admirably serve those who have cause to use statistics in forensic science.' D.V. Lindley

Table of Contents

        List of tables                             xiii
List of figures xix
Foreword xxii
Preface to the first edition xxv
Preface to the second edition xxviii
Uncertainty in forensic science 1 (34)
Introduction 1 (1)
Statistics and the law 2 (3)
Uncertainty in scientific evidence 5 (7)
The frequentist method 6 (1)
Stains of body fluids 7 (2)
Glass fragments 9 (3)
Terminology 12 (3)
Types of data 15 (1)
Probability 16 (19)
Introduction 16 (2)
A standard for uncertainty 18 (2)
Events 20 (1)
Subjective probability 21 (2)
Laws of probability 23 (2)
Dependent events and background 25 (4)
information
Law of total probability 29 (3)
Updating of probabilities 32 (3)
Variation 35 (34)
Populations 35 (2)
Samples and estimates 37 (3)
Counts 40 (12)
Probabilities 40 (1)
Summary measures 41 (2)
Binomial distribution 43 (1)
Multinomial distribution 44 (1)
Hypergeometric distribution 45 (3)
Poisson distribution 48 (3)
Beta-binomial distribution 51 (1)
Measurements 52 (17)
Summary statistics 52 (1)
Normal distribution 53 (7)
Student's t-distribution 60 (2)
Beta distribution 62 (1)
Dirichlet distribution 63 (1)
Multivariate Normal and correlation 64 (5)
The evaluation of evidence 69 (50)
Odds 69 (3)
Complementary events 69 (1)
Examples 70 (1)
Definition 70 (2)
Bayes' Theorem 72 (6)
Statement of the theorem 72 (1)
Examples 73 (5)
Errors in interpretation 78 (17)
Fallacy of the transposed conditional 79 (2)
Source probability error 81 (1)
Ultimate issue error 82 (1)
Defender's fallacy 82 (1)
Probability (another match) error 83 (1)
Numerical conversion error 84 (1)
False positive fallacy 85 (1)
Uniqueness 86 (1)
Other difficulties 87 (2)
Empirical evidence of errors in 89 (6)
interpretation
The odds form of Bayes' theorem 95 (6)
Likelihood ratio 95 (4)
Logarithm of the likelihood ratio 99 (2)
The value of evidence 101 (17)
Evaluation of forensic evidence 101 (4)
Summary of competing propositions 105 (2)
Qualitative scale for the value of the 107 (4)
evidence
Misinterpretations 111 (1)
Explanation of transposed conditional 112 (4)
and defence fallacies
The probability of guilt 116 (2)
Summary 118 (1)
Historical review 119 (38)
Early history 119 (3)
The Dreyfus case 122 (3)
Statistical arguments by early 125 (1)
twentieth-century forensic scientists
People v. Collins 126 (3)
Discriminating power 129 (12)
Derivation 129 (1)
Evaluation of evidence by 130 (3)
discriminating power
Finite samples 133 (2)
Combination of independent systems 135 (1)
Correlated attributes 136 (5)
Significance probabilities 141 (8)
Calculation of significance 141 (3)
probabilities
Relationship to likelihood ratio 144 (3)
Combination of significance 147 (2)
probabilities
Coincidence probabilities 149 (4)
Introduction 149 (2)
Comparison stage 151 (1)
Significance stage 151 (2)
Likelihood ratio 153 (4)
Bayesian inference 157 (22)
Introduction 157 (3)
Bayesian inference for a Bernoulli 160 (2)
probability
Estimation with zero occurrences in a 162 (3)
sample
Estimation of products in forensic 165 (1)
identification
Bayesian inference for a Normal mean 166 (4)
Interval estimation 170 (5)
Confidence intervals 170 (2)
Highest posterior density intervals 172 (1)
Bootstrap intervals 172 (1)
Likelihood intervals 173 (2)
Odds ratios 175 (4)
Sampling 179 (26)
Introduction 179 (3)
Choice of sample size 182 (8)
Large consignments 182 (4)
Small consignments 186 (4)
Quantity estimation 190 (7)
Frequentist approach 190 (1)
Bayesian approach 191 (6)
Misleading evidence 197 (8)
Interpretation 205 (40)
Concepts and court cases 205 (9)
Relevant population 205 (1)
Consideration of odds 206 (2)
Combination of evidence 208 (1)
Specific cases 208 (6)
Pre-assessment and relevant propositions 214 (7)
Levels of proposition 214 (3)
Pre-assessment of the case 217 (3)
Pre-assessment of the evidence 220 (1)
Assessment of value of various evidential 221 (18)
types
Earprints 221 (2)
Firearms and toolmarks 223 (3)
Fingerprints 226 (2)
Speaker recognition 228 (1)
Hair 229 (2)
Documents 231 (2)
Envelopes 233 (2)
Handwriting 235 (4)
Paint 239 (1)
Pre-data and post-data questions 239 (6)
Transfer evidence 245 (38)
The likelihood ratio 245 (9)
Probability of guilt 245 (1)
Justification 246 (2)
Combination of evidence and comparison 248 (6)
of more than two propositions
Correspondence probabilities 254 (1)
Direction of transfer 255 (16)
Transfer of evidence from the criminal 255 (5)
to the scene
Transfer of evidence from the scene to 260 (1)
the criminal
Transfer probabilities 261 (9)
Two-way transfer 270 (1)
Presence of non-matching evidence 271 (1)
Grouping 271 (3)
Relevant populations 274 (9)
Discrete data 283 (36)
Notation 283 (1)
Single sample 283 (5)
Introduction 283 (3)
General population 286 (1)
Particular population 286 (1)
Examples 286 (2)
Two samples 288 (4)
Two stains, two offenders 288 (3)
DNA profiling 291 (1)
Many samples 292 (3)
Many different profiles 292 (1)
General cases 293 (2)
Relevance of evidence and relevant 295 (9)
material
Introduction 295 (1)
Subjective probabilities 296 (1)
Association propositions 296 (1)
Intermediate association propositions 297 (1)
Examples 298 (4)
Two stains, one offender 302 (2)
Summary 304 (4)
Stain known to have been left by 304 (1)
offenders
Relevance: stain may not have been left 305 (2)
by offenders
Relevance and the crime level 307 (1)
Missing persons 308 (4)
Case 1 (Kuo, 1982) 309 (1)
Case 2 (Ogino and Gregonis,1981) 309 (1)
Calculation of the likelihood ratio 310 (2)
Paternity: combination of likelihood 312 (7)
ratios
Likelihood of paternity 314 (3)
Probability of exclusion in paternity 317 (2)
Continuous data 319 (40)
The likelihood ratio 319 (2)
Normal distribution for between-source 321 (9)
data
Sources of variation 322 (1)
Derivation of the marginal distribution 322 (2)
Approximate derivation of the 324 (2)
likelihood ratio
Lindley's approach 326 (1)
Interpretation of result 327 (1)
Examples 328 (2)
Estimation of a probability density 330 (7)
function
Kernel density estimation for 337 (5)
between-source data
Application to medullary widths of cat 339 (1)
hairs
Refractive index of glass 339 (3)
Probabilities of transfer 342 (11)
Introduction 342 (1)
Single fragment 342 (3)
Two fragments 345 (4)
A practical approach to glass evaluation 349 (3)
Graphical models for the assessment of 352 (1)
transfer probabilities
Approach based on t-distribution 353 (4)
Appendix Derivation of V when the 357 (2)
between-source measurements are assumed
normally distributed
Multivariate analysis 359 (22)
Introduction 359 (1)
Description of example 360 (2)
Univariate t-tests 362 (1)
Hotelling's T2 363 (2)
Univariate Normality, two sources of 365 (1)
variation
Multivariate Normality, two sources of 366 (5)
variation
Caveat lector 371 (1)
Summary 372 (1)
Appendix 373 (8)
Matrix terminology 373 (4)
Determination of a likelihood ratio 377 (4)
with an assumption of Normality
Fibres 381 (18)
Introduction 381 (1)
Likelihood ratios in scenarios involving 381 (11)
fibres
Fibres evidence left by an offender 382 (5)
Comments on the fibres scenario 387 (1)
Fibres evidence not left by the offender 388 (1)
Cross-transfer 389 (3)
Pre-assessment in fibres scenarios 392 (4)
The case history 392 (1)
Formulation of the pairs of 392 (2)
propositions and events
Assessment of the expected value of the 394 (2)
likelihood ratio
The relevant population of fibres 396 (3)
DNA profiling 399 (30)
Introduction 399 (2)
Hardy-Weinberg equilibrium 401 (3)
DNA likelihood ratio 404 (1)
Uncertainly 405 (1)
Variation in sub-population allele 405 (4)
frequencies
Related individuals 409 (3)
More than two propositions 412 (2)
Database searching 414 (5)
Search and selection effect (double 418 (1)
counting error)
Island problem 419 (2)
Mixtures 421 (3)
Error rate 424 (5)
Bayesian networks 429 (26)
Introduction 429 (1)
Bayesian networks 430 (7)
The construction of Bayesian networks 431 (6)
Evidence at the crime level 437 (2)
Preliminaries 437 (1)
Description of probabilities required 437 (2)
Missing evidence 439 (3)
Preliminaries 439 (1)
Determination of a structure for a 440 (2)
Bayesian network
Comments 442 (1)
Error rates 442 (2)
Preliminaries 442 (1)
Determination of a structure for a 443 (1)
Bayesian network
Transfer evidence 444 (3)
Preliminaries 444 (1)
Determination of a structure for a 444 (2)
Bayesian network
Comment on the transfer node 446 (1)
Combination of evidence 447 (2)
Cross-transfer evidence 449 (3)
Description of nodes 451 (1)
Probabilities for nodes 452 (1)
Factors to consider 452 (1)
Parameter choice 453 (1)
Summary 453 (2)
References 455 (30)
Notation 485 (4)
Cases 489 (2)
Author index 491 (8)
Subject index 499