File #2839: "2020_Book_TargetingDomesticAbuseWithPoli.pdf"
Testo
1|Preface|6
1|Acknowledgements|7
1|Contents|8
1|About the Authors|11
1|Contents|12
1|Contents|14
1|Chapter 1: Introduction|16
2|1.1 Targeting Domestic Abuse|16
2|References|19
1|Chapter 2: Domestic Abuse in England and Wales|21
2|2.1 How Domestic Abuse Is Defined|21
2|2.2 Characteristics and Terms Used|22
2|2.3 The Main Problems Facing Responders|23
2|2.4 Police Responses to Domestic Abuse|25
3|2.4.1 Initial Response: Mandatory Police Attendance|25
3|2.4.2 Arrest Policy|26
3|2.4.3 Alternatives to Prosecution|27
4|2.4.3.1 Civil Orders|27
4|2.4.3.2 Cautions|28
4|2.4.3.3 Restorative Justice|28
3|2.4.4 Risk Assessment|29
3|2.4.5 Advocates|30
3|2.4.6 Multi-agency Meetings|30
3|2.4.7 Perpetrator Management|31
2|2.5 A Summary|31
2|References|32
1|Chapter 3: Key Questions That Police Data Might Help Us Answer|36
2|3.1 About Police Data|36
2|3.2 Victim Surveys|37
2|3.3 Police Records|38
2|3.4 What Problems to Measure?|40
2|3.5 Summary|45
2|References|45
1|Chapter 4: The Existing Evidence|48
2|4.1 Introduction|48
2|4.2 Evidence on Repeat Domestic Abuse|49
2|4.3 Evidence on Escalation|50
2|4.4 Evidence on the Concentration of Harm|52
2|4.5 Evidence on Serial Domestic Abuse|52
3|4.5.1 Typologies of Domestic Batterers|53
3|4.5.2 Serial Perpetrators|55
3|4.5.3 Prevalence of Serial Perpetrators of Domestic Abuse|56
2|4.6 Evidence on Forecasting Domestic Abuse|57
3|4.6.1 Actuarial Instruments in Criminal Justice Forecasts|57
3|4.6.2 Machine Learning Techniques|61
3|4.6.3 Previous Use of Random Forests for Criminal Justice Forecasting|63
3|4.6.4 Criticisms and Problems|67
2|4.7 A Summary of the Evidence|69
2|References|70
1|Chapter 5: Measuring Harm|76
2|5.1 What Is Harm and How Is It Measured?|76
3|5.1.1 Review of Harm Measurement Tools|78
4|5.1.1.1 Public Perception–Based Tools|78
4|5.1.1.2 Economic Harm–Based Tools|81
4|5.1.1.3 Sentence-Based Tools|82
4|5.1.1.4 Theoretical-Framework Tools|85
3|5.1.2 Assessing Which Tool to Use|85
4|5.1.2.1 Cambridge Crime Harm Index|86
4|5.1.2.2 Crime Survey for England and Wales Victim Seriousness Judgement|87
4|5.1.2.3 The Home Office Economic and Social Costs of Crime Tool|88
4|5.1.2.4 Office for National Statistics Crime Severity Score|89
2|5.2 Summary|90
2|References|92
1|Chapter 6: Repeat Domestic Abuse, Escalation and Concentration of Harm|96
2|6.1 Important Questions on Repeat Abuse|96
3|6.1.1 Important Questions on Escalation|97
3|6.1.2 Important Questions on the Concentration of Harm|98
3|6.1.3 What Data Did We Have and How Did We Analyse It?|98
4|6.1.3.1 Procedure: Repeat Abuse|100
4|6.1.3.2 Procedure: Escalation|100
4|6.1.3.3 Procedure: Concentration of Harm|101
3|6.1.4 Findings|101
4|6.1.4.1 Prevalence of Repeat Domestic Abuse Among Victims|101
4|6.1.4.2 Conditional probability of Further Crimes for Victims|103
4|6.1.4.3 Prevalence of Repeat Domestic Abuse Among Offenders|103
4|6.1.4.4 Conditional Probability of Further Crimes for Offenders|105
4|6.1.4.5 Escalation Among Victims|105
4|6.1.4.6 Escalation Among Offenders|106
4|6.1.4.7 Concentration of Harm – The Power Few|108
4|6.1.4.8 Never Called Before (or Again)|109
3|6.1.5 Theoretical Implications|110
3|6.1.6 Research Implications|111
3|6.1.7 Policy Implications|112
2|References|113
1|Chapter 7: Serial Domestic Abuse|116
2|7.1 Important Questions on Serial Abuse|116
2|7.2 What Data Did We Have and How Did We Analyse It?|116
3|7.2.1 Procedure: Serial Abuse|119
2|7.3 Findings|120
3|7.3.1 Prevalence|120
3|7.3.2 Types of Abuse|121
3|7.3.3 Harm|122
3|7.3.4 Other Crimes|125
3|7.3.5 Subclassifications of Cohorts|127
2|7.4 Theoretical Implications|128
2|7.5 Research Implications|130
2|7.6 Policy Implications|132
2|References|135
1|Chapter 8: Forecasting Risk|137
2|8.1 Important Questions on Forecasting Risk|137
2|8.2 What Data Did We Have and How Did We Analyse It?|138
2|8.3 Procedure: Forecasting|141
3|8.3.1 How Random Forest Algorithms Work|142
3|8.3.2 Model Parameters|145
4|8.3.2.1 Instrument Selection|145
4|8.3.2.2 Definition of Unit of Prediction and Follow Up Period|145
4|8.3.2.3 Outcome Variable|146
4|8.3.2.4 Predictors|146
4|8.3.2.5 Relative Costs of Errors|147
4|8.3.2.6 Presentation of Findings|147
4|8.3.2.7 Baseline Considerations|148
2|8.4 Findings|148
3|8.4.1 What Proportion of All Arrestees Go on to Commit Domestic Abuse?|148
3|8.4.2 What Proportion of Domestic Abuse Arrestees Have Prior Domestic Records?|149
3|8.4.3 Can Antecedent Inputs Predict Future Domestic Abuse Cases to a High Degree of Accuracy?|151
2|8.5 Which Predictors Have the Greatest Impact on Accuracy?|153
3|8.5.1 Age First Arrested for Domestic Abuse|155
3|8.5.2 Years Since Last Arrest for Domestic Abuse|157
3|8.5.3 Presenting Offence Was Domestic Abuse|157
3|8.5.4 Number of Prior Domestic Arrests|157
3|8.5.5 Age at First Arrest for a Sexual Offence|158
2|8.6 Theoretical Implications|158
2|8.7 Research Implications|160
3|8.7.1 Using Targeting Research Alongside Testing|163
3|8.7.2 Integrating Harm Measurement Tools|164
2|8.8 Policy Implications|165
3|8.8.1 Political Implications|165
3|8.8.2 Practical Implications: The Need for More Information|166
3|8.8.3 A Working Model for Prediction|166
3|8.8.4 Building Police Forecasting Capabilities|168
2|References|169
1|Chapter 9: Conclusions: Integrating Research into Practice|172
2|9.1 What We Set Out to Analyse|172
2|9.2 More Data Are Needed in This Fight|174
2|9.3 How to Improve Our Analysis|176
3|9.3.1 Police Records Are Not the Whole Story|176
4|9.3.1.1 The CCHI Is Imperfect|176
4|9.3.1.2 Intimate Partner Abuse Versus Domestic Abuse|176
4|9.3.1.3 Limited Scope for Prediction|177
4|9.3.1.4 Data Quality|177
4|9.3.1.5 The Black Box Nature of Random Forests|178
2|9.4 Final Conclusions|179
2|A. Appendices|179
3|Appendix I: Technical Information Relating to Random Forest Modelling|179
4|About This Appendix|179
4|Model Tuning|180
4|Partial Response Plots|182
3|Appendix II: Cambridge Crime Harm Index – Selected Values|185
2|References|187
1|Index|189
1|Acknowledgements|7
1|Contents|8
1|About the Authors|11
1|Contents|12
1|Contents|14
1|Chapter 1: Introduction|16
2|1.1 Targeting Domestic Abuse|16
2|References|19
1|Chapter 2: Domestic Abuse in England and Wales|21
2|2.1 How Domestic Abuse Is Defined|21
2|2.2 Characteristics and Terms Used|22
2|2.3 The Main Problems Facing Responders|23
2|2.4 Police Responses to Domestic Abuse|25
3|2.4.1 Initial Response: Mandatory Police Attendance|25
3|2.4.2 Arrest Policy|26
3|2.4.3 Alternatives to Prosecution|27
4|2.4.3.1 Civil Orders|27
4|2.4.3.2 Cautions|28
4|2.4.3.3 Restorative Justice|28
3|2.4.4 Risk Assessment|29
3|2.4.5 Advocates|30
3|2.4.6 Multi-agency Meetings|30
3|2.4.7 Perpetrator Management|31
2|2.5 A Summary|31
2|References|32
1|Chapter 3: Key Questions That Police Data Might Help Us Answer|36
2|3.1 About Police Data|36
2|3.2 Victim Surveys|37
2|3.3 Police Records|38
2|3.4 What Problems to Measure?|40
2|3.5 Summary|45
2|References|45
1|Chapter 4: The Existing Evidence|48
2|4.1 Introduction|48
2|4.2 Evidence on Repeat Domestic Abuse|49
2|4.3 Evidence on Escalation|50
2|4.4 Evidence on the Concentration of Harm|52
2|4.5 Evidence on Serial Domestic Abuse|52
3|4.5.1 Typologies of Domestic Batterers|53
3|4.5.2 Serial Perpetrators|55
3|4.5.3 Prevalence of Serial Perpetrators of Domestic Abuse|56
2|4.6 Evidence on Forecasting Domestic Abuse|57
3|4.6.1 Actuarial Instruments in Criminal Justice Forecasts|57
3|4.6.2 Machine Learning Techniques|61
3|4.6.3 Previous Use of Random Forests for Criminal Justice Forecasting|63
3|4.6.4 Criticisms and Problems|67
2|4.7 A Summary of the Evidence|69
2|References|70
1|Chapter 5: Measuring Harm|76
2|5.1 What Is Harm and How Is It Measured?|76
3|5.1.1 Review of Harm Measurement Tools|78
4|5.1.1.1 Public Perception–Based Tools|78
4|5.1.1.2 Economic Harm–Based Tools|81
4|5.1.1.3 Sentence-Based Tools|82
4|5.1.1.4 Theoretical-Framework Tools|85
3|5.1.2 Assessing Which Tool to Use|85
4|5.1.2.1 Cambridge Crime Harm Index|86
4|5.1.2.2 Crime Survey for England and Wales Victim Seriousness Judgement|87
4|5.1.2.3 The Home Office Economic and Social Costs of Crime Tool|88
4|5.1.2.4 Office for National Statistics Crime Severity Score|89
2|5.2 Summary|90
2|References|92
1|Chapter 6: Repeat Domestic Abuse, Escalation and Concentration of Harm|96
2|6.1 Important Questions on Repeat Abuse|96
3|6.1.1 Important Questions on Escalation|97
3|6.1.2 Important Questions on the Concentration of Harm|98
3|6.1.3 What Data Did We Have and How Did We Analyse It?|98
4|6.1.3.1 Procedure: Repeat Abuse|100
4|6.1.3.2 Procedure: Escalation|100
4|6.1.3.3 Procedure: Concentration of Harm|101
3|6.1.4 Findings|101
4|6.1.4.1 Prevalence of Repeat Domestic Abuse Among Victims|101
4|6.1.4.2 Conditional probability of Further Crimes for Victims|103
4|6.1.4.3 Prevalence of Repeat Domestic Abuse Among Offenders|103
4|6.1.4.4 Conditional Probability of Further Crimes for Offenders|105
4|6.1.4.5 Escalation Among Victims|105
4|6.1.4.6 Escalation Among Offenders|106
4|6.1.4.7 Concentration of Harm – The Power Few|108
4|6.1.4.8 Never Called Before (or Again)|109
3|6.1.5 Theoretical Implications|110
3|6.1.6 Research Implications|111
3|6.1.7 Policy Implications|112
2|References|113
1|Chapter 7: Serial Domestic Abuse|116
2|7.1 Important Questions on Serial Abuse|116
2|7.2 What Data Did We Have and How Did We Analyse It?|116
3|7.2.1 Procedure: Serial Abuse|119
2|7.3 Findings|120
3|7.3.1 Prevalence|120
3|7.3.2 Types of Abuse|121
3|7.3.3 Harm|122
3|7.3.4 Other Crimes|125
3|7.3.5 Subclassifications of Cohorts|127
2|7.4 Theoretical Implications|128
2|7.5 Research Implications|130
2|7.6 Policy Implications|132
2|References|135
1|Chapter 8: Forecasting Risk|137
2|8.1 Important Questions on Forecasting Risk|137
2|8.2 What Data Did We Have and How Did We Analyse It?|138
2|8.3 Procedure: Forecasting|141
3|8.3.1 How Random Forest Algorithms Work|142
3|8.3.2 Model Parameters|145
4|8.3.2.1 Instrument Selection|145
4|8.3.2.2 Definition of Unit of Prediction and Follow Up Period|145
4|8.3.2.3 Outcome Variable|146
4|8.3.2.4 Predictors|146
4|8.3.2.5 Relative Costs of Errors|147
4|8.3.2.6 Presentation of Findings|147
4|8.3.2.7 Baseline Considerations|148
2|8.4 Findings|148
3|8.4.1 What Proportion of All Arrestees Go on to Commit Domestic Abuse?|148
3|8.4.2 What Proportion of Domestic Abuse Arrestees Have Prior Domestic Records?|149
3|8.4.3 Can Antecedent Inputs Predict Future Domestic Abuse Cases to a High Degree of Accuracy?|151
2|8.5 Which Predictors Have the Greatest Impact on Accuracy?|153
3|8.5.1 Age First Arrested for Domestic Abuse|155
3|8.5.2 Years Since Last Arrest for Domestic Abuse|157
3|8.5.3 Presenting Offence Was Domestic Abuse|157
3|8.5.4 Number of Prior Domestic Arrests|157
3|8.5.5 Age at First Arrest for a Sexual Offence|158
2|8.6 Theoretical Implications|158
2|8.7 Research Implications|160
3|8.7.1 Using Targeting Research Alongside Testing|163
3|8.7.2 Integrating Harm Measurement Tools|164
2|8.8 Policy Implications|165
3|8.8.1 Political Implications|165
3|8.8.2 Practical Implications: The Need for More Information|166
3|8.8.3 A Working Model for Prediction|166
3|8.8.4 Building Police Forecasting Capabilities|168
2|References|169
1|Chapter 9: Conclusions: Integrating Research into Practice|172
2|9.1 What We Set Out to Analyse|172
2|9.2 More Data Are Needed in This Fight|174
2|9.3 How to Improve Our Analysis|176
3|9.3.1 Police Records Are Not the Whole Story|176
4|9.3.1.1 The CCHI Is Imperfect|176
4|9.3.1.2 Intimate Partner Abuse Versus Domestic Abuse|176
4|9.3.1.3 Limited Scope for Prediction|177
4|9.3.1.4 Data Quality|177
4|9.3.1.5 The Black Box Nature of Random Forests|178
2|9.4 Final Conclusions|179
2|A. Appendices|179
3|Appendix I: Technical Information Relating to Random Forest Modelling|179
4|About This Appendix|179
4|Model Tuning|180
4|Partial Response Plots|182
3|Appendix II: Cambridge Crime Harm Index – Selected Values|185
2|References|187
1|Index|189