The impact of length of domestic violence treatment on the patterns of subsequent intimate partner violence

Randomized clinical trials (RCTs) have been used for several decades in criminal justice research to assess the benefits of interventions. RCTs have been employed with some frequency in evaluating batterer intervention programs (BIPs); between 1990 and 2005, there were five RCTs of BIPs. While most of these studies failed to find positive findings, one conducted in Brooklyn, New York, appeared to detect a positive effect of a BIP on domestic violence re-offending. Subsequent inspection of the data, however, revealed several complicating factors that led us to reanalyze these data. Our reanalysis focused on whether the intervention temporarily suppresses abusive behavior during the time that offenders are under court control, or whether the program produces lasting changes in offender behavior. With improved data analysis techniques, additional outcome data, and several alternative analytical models, we arrive at an alternative conclusion than is reported in the original publication. While we still find evidence of a reduction in the number of new incidents across a 15-month period due to treatment assignment , we now find that the effect of attending treatment does not likely persist beyond the treatment period once the treatment measure is specified as a time-dependent parameter rather than a fixed effect. This outcome is more consistent with a suppression/supervision explanation rather than a therapeutic outcome explanation. The implications of these results are discussed.