Dashing Hopes? the Predictive Accuracy of Domestic Abuse Risk Assessment by Police

Abstract The Domestic Abuse, Stalking and Honour Based Violence (DASH) form is a standardized risk assessment implemented across most UK police forces. It is intended to facilitate an officer’s structured professional judgment about the risk a victim faces of serious harm at the hand of their abuser...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:British journal of criminology Ročník 59; číslo 5; s. 1013 - 1034
Hlavní autoři: Turner, Emily, Medina, Juanjo, Brown, Gavin
Médium: Journal Article
Jazyk:angličtina
Vydáno: UK Oxford University Press 01.09.2019
Témata:
domestic abuse
risk assessment
police
machine learning
predictive policing
ISSN:0007-0955, 1464-3529
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Abstract The Domestic Abuse, Stalking and Honour Based Violence (DASH) form is a standardized risk assessment implemented across most UK police forces. It is intended to facilitate an officer’s structured professional judgment about the risk a victim faces of serious harm at the hand of their abuser. Until now, it has been an open question whether this tool works in practice. Here, we present the largest scale European study, making the case that the risk assessment tool is underperforming. Each element of the DASH questionnaire is, at best, weakly predictive of revictimization. Officer risk predictions based on DASH are little better than random and a logistic regression model that predicts the same outcome using DASH only provides modest improvement in performance.
ISSN:0007-0955
1464-3529
DOI:10.1093/bjc/azy074