Adjusting the National Crime Victimization Survey's Estimates of Rape and Domestic Violence for "Gag" Factors

• NCJ Number: 173061
• Author(s): A. L. Coker, E. A. Stasny
• Date Published: January 1994
• Length: 25 pages
• Agencies: NIJ-Sponsored

Annotation

Statistical modeling techniques were used to estimate rates of rape and domestic assault, adjusting for interviewing conditions under which the National Crime Victimization Survey (NCVS) was administered. Data set archived by the NIJ Data Resources Program at the National Archive of Criminal Justice Data, located at URL http://www.icpsr.umich.edu/nacjd.

Abstract

The research focused on the factors associated with reporting rape or domestic assault and factors associated with interview conditions (telephone or personal and who was present during the interview), and statistical modeling to adjust the estimates. The research used NCVS unweighted data for 1986-90 for females ages 16 years or older. Results revealed that crime reporting rates differed by the setting of the interview and the type of crime reported. Rape was reported 2.4 times less frequently and domestic violence 3.1 times less frequently in telephone compared with personal interviews, regardless of who was present during the interviews. The magnitude of this gagging effect was smaller for both assaults and breaking and entry. In addition, rape was reported 2.8 times less frequently when a spouse was present, but 1.35 times more frequently when others and not the spouse were present. Similarly, domestic assault was reported 5.6 times less frequently if a spouse was present for the interview, but 1.8 times more frequently if others were present. A saturated model fit the data exactly but did not make it possible to take other factors into account. An extended model provided a poor fit in all cases. However, the model allows for possible extensions that may provide better first to the data. Tables and 9 references