Using the National Incident-Based Reporting System (NIBRS) to Test Estimates of Arrestee and Offender Characteristics

National-Incident Based Reporting System (NIBRS) information for 1993 iscompared with two other sets of homicide data to assess the acuracy ofprocedures for estimating age-, sex-, and race-specific arrest counts fromtraditional Uniform Crime Report (UCR) data. The simultaneous age, race, andsex characteristics of offenders provided in the NIBRS arrest and offender segmentsare compared with estimates of the same characteristics derived from summaryUCR arrest data. The results suggest that using UCR marginal totals toestimate arrest counts by race and sex produces reasonably accurateestimates of the number of black and nonblack male arrests for murder andother offenses. Estimates of arrests of females by race and sex are lessaccurate, probably because of the relatively small number of arrests ofwomen and girls. Estimating male arrest counts for specific age groups alsoproduces reasonably accurate estimates. The results suggest that previousresearch employing such estimates may have been misleading in the estimatesof female arrests by race but accurate in the estimates of male arrests byrace. Although the use of summary UCR-based estimates in futurecross-sectional research will become increasingly unnecessary as the NIBRSreplaces the current UCR program, time series research designs on arrest andcrime trends will continue to require estimates.