The purpose of the study was to corroborate the
findings of the Minneapolis Domestic Violence Experiment (1984) that
arrest is an effective deterrent against continued domestic assaults.
The study addressed the following questions: (1) To what extent does
arrest decrease the likelihood of continued violence, as assessed by
the victim? (2) To what extent does arrest decrease the likelihood of
continued complaints of crime, as assessed by police records? (3) What
are the differences in arrest recidivism between cases that involved
arrest versus cases that involved mediation, separation, warrant
issued, or no warrant issued?
Cases included in the study were required to meet
the following eligibility criteria: (1) Probable cause for an arrest
for misdemeanor assault was established. Cases lacking probable cause
and felony assault cases were excluded. (2) The case involved at least
two people, a victim and a suspect. (3) Both parties to the assault
were at least 18 years of age. (4) Both parties lived together
sometime during the year preceding the assault. (5) If the suspect was
present, a check of police records indicated no arrest warrant was on
file. If officers determined that the case met the eligibility
criteria, a treatment was randomly assigned by a computer routine
operated by the Information Unit of the Omaha Police Department. Cases
in which both the victim and the suspect were present when the police
arrived were assigned a treatment of either mediation, separation, or
arrest. Cases in which no suspects were present at the time police
arrived were assigned treatments of warrant issued or no warrant
issued. Interviews with victims were conducted one week, six months,
and 12 months after the domestic violence incident. Additional data
were collected from police officers who completed Domestic Violence
Report forms for each case. Police record searches were conducted on
suspects six and twelve months after the incident. Continued
complaints of crime, arrest recidivism, and victim reports of repeated
violence (fear of injury, pushing/hitting, and physical injury) were
used as outcome measures to assess the extent to which the various
treatments prevented subsequent violence.
Two-stage random sampling design.
Domestic violence cases in Omaha, Nebraska during
Incidents of misdemeanor domestic assault.
personal interviews and police records
event/transaction data, and survey data,
Data were collected regarding the nature of the
domestic violence incident, subsequent incidents, the extent of the
victim's injury, extent of drug use by the victim and the suspect, and
the suspect-victim relationship. During the victim interviews
additional data were collected regarding the victim's level of fear,
self-esteem, locus of control, and welfare dependency, and changes in
the relationship between suspect and victim. Demographic information
includes age, sex, income, employment, and marital status. Arrest,
charge, and complaint data were collected on the suspects at six- and
twelve-month intervals following the original domestic violence
A total of 577 domestic violence cases were
selected for inclusion in the study. The number of respondents who
completed interviews for Waves I, II, and III are 477 (81 percent),
438 (76 percent), and 416 (72 percent), respectively.
A modified Conflict Tactic Scale was used in the study.
Hollingshead and Duncan Socioeconomic Indices are also included.