Published In

Journal of Experimental Criminology

Document Type


Publication Date



Children and violence -- Research, Victims of crimes -- Services for -- Evaluation, Children -- Violence against -- Case studies



The study tests whether participation in interventions offered by a subset of sites from the National Safe Start Promising Approaches for Children Exposed to Violence initiative improved outcomes for children relative to controls.


The study pools data from the nine Safe Start sites that randomized families to intervention and control groups, using a within-site block randomization strategy based on child age at baseline. Caregiver-reported outcomes, assessed at baseline, six and 12 months, included caregiver personal problems, caregiver resource problems, parenting stress, child and caregiver victimization, child trauma symptoms, child behavior problems, and social-emotional competence.


Results revealed no measurable intervention impact in intent-to-treat analyses at either six- or twelve-month post-baseline. In six-month as-treated analyses, a medium to high intervention dose was associated with improvement on two measures of child social-emotional competence: cooperation and assertion. Overall, there is no reliable evidence of significant site-to-site effect variability, even in the two cases of significant intervention effect.


Since families in both the intervention and control groups received some degree of case management and both groups improved over time, it may be advantageous to explore the potential impacts of crisis and case management separately from mental health interventions. It may be that, on average, children in families whose basic needs are being attended to improve substantially on their own.


Pre-publication version of an article accepted for publication in the Journal of Experimental Criminology. The version of record may be found at

Persistent Identifier