News item | 06-07-2022 | 09:45
Every year, about 15,000 young people are subject to a Halt settlement. In the initial interview, their risk of recidivism and criminogenic needs are estimated via the Halt signaling instrument (Halt-SI). On the basis of this, the Halt settlement is completed and the young person is referred to assistance if necessary. By applying other weightings, the predictive value (predictive validity) of the Halt-SI can be improved. This is apparent from research by the University of Amsterdam, commissioned by the WODC.
The Halt settlement is an intervention for young people outside criminal law. This is intended to make young people aware of their behavior and to prevent the repetition of criminal behaviour. During the initial interview, the young people and their parents answer questions about subjects such as school performance, leisure activities, interaction with peers and substance use. On this basis, among other things, the risk of recidivism is estimated and it is determined which modules and/or learning assignments are appropriate.
Different weights, better estimate
To investigate how well these estimates are made, the predictive value (predictive validity) of the Halt-SI was investigated. This shows that although most of the subjects that are asked are significantly related to recidivism, there are generally (very) weak connections. By basing the weights of the Halt-SI on data collected from Dutch youngsters, instead of American ones, the predictive value of the instrument appears to improve. All four weighting methods examined have a significantly higher predictive value.
The researchers also make a number of recommendations for updating the Halt-SI. For example, to no longer distinguish between boys and girls and different age groups, because there are not many differences between these groups in the strength of the association between items and recidivism. Another recommendation is to reconsider the classification into risk groups (no/little, some and high risk).
Characteristics of current Halt participants
The researchers also mapped out the current Halt population. This shows that Halt participants are on average 15.7 years old and that 75% are boys. Participants are referred to Halt about as often for a felony as for an infraction: they are most often referred for a property crime. The most common risk factors are no leisure activities in association, impulsiveness and truancy. Of the Halt participants, 91% fall in the ‘no/little’ risk group and only 0.5% in the ‘high’ risk group (according to the current risk classification).
It is impossible to say on the basis of the results of this study whether the group of Halt participants has become ‘heavier’ in recent years. Halt participants have committed a crime more often in recent years and an offense less often, but the total score of the risk factors present has remained unchanged. Some risk factors have increased slightly and others decreased. Due to the corona measures and the transition to a new registration system at Halt, it is difficult to determine the exact cause of this.