Reported Incidents Between Intimate Partners: An Analysis of Data from the Michigan Information Crime Reporting System

Prepared by the Michigan Justice Statistics Center, Michigan State University

Analytic Technique

Michigan Incident Crime Reporting (MICR), Michigan's incident-based reporting system, provides data to the FBI through the National Incident-Based Reporting System. Like NIBRS, MICR compiles information by incident on offense type, offender and victim characteristics and relationship, arrests, type of property lost or seized, and drug involvement. An incident is defined as one or more offenses committed by the same offender or group of offenders acting in concert, at the same time or place . Unlike summary reporting that collects single-level data by offense, NIBRS allows data to be collected on a number of offenders, offenses, and victims within one incident. Because of the hierarchical structure of the data, incident data are stored in six separate files: administrative, offender, offense, victim, arrest, and property information. Each agency assigns separate incidents a unique incident number that is then used to identify each element within the six data files and link separate elements of the data files.

Although it is possible to examine MICR data separately by segment, there are benefits to linking data segments so that a linear file can be constructed. Because the goal of this analysis was both to demonstrate the utility of MICR and to examine all possible characteristics of a given incident, all segments of the data were linked by incident number. We also constructed a unique identifier for each incident by linking the incident number and the originating agency identification number for each incident.

Once the identifier was constructed, data segments were read into SPSS for analysis. Because MICR data is structured by incident, it was possible to collect multiple data characteristics--up to 20 offenders, 5 victims, 7 arrests, and 10 property or drug elements-- within a given incident. Using the lead forward commands in SPSS, data were brought up to one line by incident number. The lag procedure was conducted separately with each of the five data segments. The incident number was then used to link all data segments. By creating this linear file that included all elements of the original hierarchical files, it was possible to analyze all characteristics of the incident as a whole instead of examining data elements separately. Overall, 708,823 unique incidents were reported to the Michigan State Police under MICR in 1998. Of those incidents, 236,552 were reported by agencies within the sample of counties included in this study. Incidents that occurred between intimate partners accounted for 11,312 incidents in the sample jurisdictions.