U.S. Department of Justice Office of Justice Programs

Bureau of Justice Statistics Special Report

August 1995, NCJ-154348 National Crime Victimization Survey Violence against Women: Estimates from the Redesigned Survey

By Ronet Bachman, Ph.D. is interested in a broad spectrum of increased the range of incident types Bureau of Justice Statistics, Statistician incidents, not just those involving that are being reported to interviewers. Linda E. Saltzman, Ph.D. weapons, severe violence, or violence And behavior-specific wording has Centers for Disease Control and perpetrated by strangers. New meth- replaced criminal justice terminology Prevention, Senior Scientist ods of cuing respondents about poten- to make the questions more under- tial experiences with victimizations standable. Estimating rates of violence against women, particularly sexual assault and other incidents which are perpe- Highlights trated by intimate offenders, continues to be a difficult task. Many factors Women age 12 or older annually about equally vulnerable to violence inhibit women from reporting these sustained almost 5 million violent by an intimate. victimizations both to police and to victimizations in 1992 and 1993. Women age 19 to 29 and women interviewers, including the private About 75% of all lone-offender vio- in families with incomes below nature of the event, the perceived lence against women and 45% of $10,000 were more likely than other stigma associated with one's victimi- violence involving multiple-offenders women to be victims of violence by zation, and the belief that no purpose was perpetrated by offenders whom an intimate. wil be served in reporting it. the victim knew. In 29% of all vio- lence against women by a lone Among victims of violence commit- The redesign of the National Crime offender, the perpetrator was an ted by an intimate, the victimization Victimization Survey intimate (husband, ex-husband, rate of women separated from their boyfriend, or ex-boyfriend). husbands was about 3 times higher After an extensive 10-year redesign than that of divorced women and project, the National Crime Victimiza- Women were about 6 times more about 25 times higher than that tion Survey (NCVS) has been revised. likely than men to experience vio- of married women. Because the A redesigned questionnaire was in lence committed by an intimate. NCVS reflects a respondent's mari- wide use by January 1992. One goal Women annually reported about tal status at the time of the interview, of the redesign was to produce more 500,000 rapes and sexual assaults which is up to 6 months after the in- accurate reporting of incidents of rape to interviewers. Friends or acquaint- cident, it is possible that separation and sexual assault and of any kind of ances of the victims committed over or divorce followed the violence. crimes committed by intimates or fam- half of these rapes or sexual as- Female victims of violence by an ily members. saults. Strangers were responsible intimate were more often injured by for about 1 in 5. the violence than females victimized The new NCVS questionnaire encour- by a stranger. ages reporting of incidents in several Women of all races and Hispanic ways. Questions were added to let and non-Hispanic women were respondents know that the interviewer A side-by-side comparison of the new patterns of victimization differ by the Rates of violence for women and old screening questions as they sex of the victim, the first section of and men, 1992-93 relate to types of activities and types the report provides rates of violence of situations and offenders is on for both women and men. Women annually reported about page 8. Later sections contain more detailed 500,000 rapes and sexual assaults, information about the specific types almost 500,000 robberies, and about This report presents the first release and contextual characteristics of vio- 3.8 million assaults to NCVS inter- of 1992-93 estimates of violence lence against women and the types viewers (table 1). (As used in this against women resulting from the new of offenders most likely to victimize report, women and men refer to per- NCVS methodology. To illustrate how women. Definitions of terminology sons age 12 or older.) Per capita appear in Methodology on pages 6-7. rates of reported rapes and other sexual assaults against women were about 10 times higher than equivalent Table 1. Average annual rate of violent crime and number of violent victimizations, by sex of victim and type of crime, 1992-93 rates against men. The rates of rob- bery and aggravated assault against Average annual rate men were about double those against per 1,000 persons Average annual age 12 or older number of incidents women. Type of crime Female Male Female Male

Crimes of violence 43.7 64.9 4,748,500 6,602,100 Offenders acted alone in most violent victimizations (table 2). A greater per- Homicide .04 .16 4,700 17,100 Rape/Sexual assault 4.6 .5 500,200 48,500 centage of victimizations against Robbery 4.4 8.6 475,900 870,800 males than females were committed Aggravated assault 8.0 16.9 863,000 1,715,400 by multiple offenders. Multiple offend- Simple assault 26.7 38.8 2,904,700 3,950,400 Note: Average annual numbers have been rounded to the nearest 100. ers committed 27% of the violent Homicides are not measured in the victimization survey; see the box victimizations against men and 16% on page 4 for the source. of those against women. For both sexes, robberies were more Table 2. Violent victimization of females and males likely than other offenses to involve by lone or multiple offenders, by type of crime, 1992-93 multiple offenders: 34% of the robber- Average annual percent of victimizations ies of women and 54% of the robber- Female victims Male victims Lone Multiple Lone Multiple ies of men. Among women, the Type of crime offender offenders offender offenders lowest percentages of multiple offend- Crimes of violence 84% 16% 73% 27% ers occurred for rape or sexual assault Rape/Sexual assault 90 10 93 * (10%) and for simple assault (13%). Robbery 66 34 46 54 A fifth of the simple assaults of males Aggravated assault 79 21 69 31 involved more than one offender. Simple assault 87 13 80 20 Note: Excludes homicide; see the box on page 4. Excludes incidents Violent victimizations against men in which the number of offenders was not ascertained. *Ten or fewer sample cases. involving multiple offenders were predominately committed by strangers (table 3). There was no significant Table 3. Multiple-offender violent victimization of females difference between the extent to which and males involving known or unknown offenders, by type of crime, 1992-93 multiple-offender victimizations against women involved known and Average annual percent of multiple-offender victimizations Female victims Male victims unknown offenders. When considered Victim knew Victim knew Victim knew Victim knew by offense and the sex of the victim, at least one none of the at least one none of the simple assaults against women by Type of crime offender offenders offender offenders multiple offenders involved a higher Crimes of violence 45% 55% 31% 69% percentage of known offenders (57%) Rape/Sexual assault 37 63 * * than strangers (43%). In contrast, Robbery 20 80 20 80 simple assaults against men involving Aggravated assault 46 54 29 71 Simple assault 57 43 39 61 multiple offenders were more likely to Note: Excludes homicide; see the box on page 4. be perpetrated by strangers (61% ver- *Ten or fewer sample cases. sus 39% known).

2 National Crime Victimization Survey Victim-offender relationship ized by known offenders than by or ex-boyfriend  an intimate. Com- for lone-offender victimizations strangers. About three-quarters of all pared to men, women were about 6 lone-offender violence against women times more likely to experience vio- Among victims of offenders acting was perpetrated by an offender whom lence committed by an intimate. alone, men were just as likely to be the victim knew. In 29% of all vio- victimized by a stranger as by some- lence against women by a lone of- Violence at the hands of an intimate one they knew (table 4). By contrast, fender, the perpetrator was a involved about 9 in 1,000 women women were more likely to be victim- husband, ex-husband, boyfriend, annually (table 5). This rate translates into about 1 million women who be- Table 4. Victim-offender relationship and sex of victim, by type of violent came the victims of such violence victimization committed by lone offenders, 1992-93 every year.

Average annual percent of victimizations Men were about twice as likely as Rape/Sex- Aggravated Simple women to experience acts of violence Victim-offender relationship Total ual assault Robbery assault assault by strangers. Men were victims of Female victims almost 2 million acts of stranger- Intimate 29% 26% 28% 28% 29% perpetrated violence annually, while Spouse 9 5 6 5 11 Ex-spouse 4 5 5 5 4 women experienced about 800,000. Boy/girlfriend (or ex-) 16 16 18 17 15 Other relative 9 3 5 7 11 Acquaintance/friend 40 53 19 36 41 Stranger 23 18 48 30 19 Male victims Intimate 4% * 3% 5% 3% Spouse 1 * * 1 1 Ex-spouse 1 * * * 1 Boy/girlfriend (or ex-) 2 * 2 3 1 Other relative 3 * 2 4 3 Acquaintance/friend 44 54 26 40 49 Stranger 49 46 69 51 45 Note: Excludes homicide. Intimate includes spouse or ex-spouse, boyfriend or girlfriend, and ex-boyfriend or ex-girlfriend. Detail may not add to total because of rounding. *Ten or fewer sample cases.

Table 5. Average annual rate of violent victimization, by sex of victim, victim-offender relationship, and type of crime committed by lone offenders, 1992-93

Average annual rate per 1,000 persons age 12 or older Other Acquaintance/ Type of crime Intimate relative friend Stranger Female victims Crimes of violence 9.3 2.8 12.9 7.4 Rape/Sexual assault 1.0 .1 2.0 .7 Robbery .7 .1 .5 1.2 Aggravated assault 1.5 .4 2.0 1.6 Simple assault 6.1 2.2 8.5 3.9 Annual average number of violent crimes 1,008,000 304,500 1,402,500 802,300

Male victims Crimes of violence 1.4 1.2 17.2 19.0 Rape/Sexual assault * * .2 .2 Robbery .1 .1 .9 2.4 Aggravated assault .5 .4 3.8 4.8 Simple assault .8 .7 12.4 11.6 Annual average number of violent crimes 143,400 122,000 1,754,000 1,933,100 Note: Excludes homicide. Intimate includes spouse or ex-spouse, boyfriend or girlfriend, and ex-boyfriend or ex-girlfriend. Average annual numbers have been rounded to the nearest 100. *Ten or fewer sample cases.

Violence against Women: Estimates from the Redesigned Survey 3 Demographic and contextual Table 6. Average annual rate of violent victimizations of women by a lone offender, by victim characteristics and victim-offender characteristics of violence relationship, 1992-93 against women

Average annual rate of violent victimizations per 1,000 females age 12 or older Violence against women perpetrated Victim Other Acquaint- by intimates was consistent across characteristic Total Intimate relative ance/friend Stranger racial and ethnic boundaries. No sta- Crimes of violence 36.1 9.3 2.8 12.9 7.4 tistically significant differences existed Race between these groups. Black and White 35.2 9.1 2.6 12.5 7.1 white women and Hispanic and non- Black 44.6 10.9 3.5 17.2 9.5 Other 27.8 6.5 4.5 8.4 5.7 Hispanic women sustained about the Ethnicity same amount of violence by intimate Hispanic 33.9 7.3 3.2 10.0 9.0 partners (table 6). Non-Hispanic 36.3 9.4 2.8 13.2 7.2 Age Compared to all other age groups, 12-18 74.6 9.6 6.1 39.1 11.9 women age 19 to 29 reported more 19-29 63.7 21.3 4.7 18.2 13.9 violence by intimates. Women age 30-45 37.5 10.8 2.8 12.4 7.4 46-64 12.5 2.2 1.2 4.1 3.7 12 to 18 were more likely than women 65 or older 4.8 1.2 0.3 1.1 1.2 older than 18 to report violence Education against them by friends or acquaint- Some high school or less 47.7 9.9 4.7 20.5 7.5 tances. In general, women age 65 High school graduate 27.9 9.2 2.2 8.8 4.9 or older were the least likely to experi- Some college or more 35.9 8.7 2.1 11.6 9.6 ence an act of violence. Annual family income $9,999 or less 57.1 19.9 6.1 18.5 7.8 Women with an annual family income $10,000-$14,999 46.8 13.3 4.0 14.1 9.1 $15,000-$19,999 42.2 10.9 3.1 17.3 7.0 under $10,000 were more likely to re- $20,000-$29,999 38.0 9.5 2.7 14.8 7.9 port having experienced violence by $30,000-$49,999 30.6 5.4 1.8 11.5 8.4 an intimate than those with an income $50,000 or more 24.8 4.5 1.8 9.7 6.3 of $10,000 or more. Marital status Married 16.9 2.7 1.6 5.7 4.9 Among victims of violence committed Widowed 10.4 1.9 0.6 3.6 2.5 Divorced 61.8 23.1 4.2 19.5 10.2 by an intimate, the victimization rate Separated 123.5 82.2 10.0 19.9 7.4 of women separated from their hus- Never married 63.9 12.0 4.6 27.2 12.9 bands was about 3 times higher than Location of residence that of divorced women and about 25 Urban 45.4 10.7 3.0 15.9 10.8 Suburban 33.6 9.2 2.7 11.5 7.1 times higher than that of married Rural 29.5 7.7 2.7 12.2 3.7 women. Because the NCVS reflects Note: Rates of violence for this table include rapes, sexual assaults, robberies, and a respondent's marital status at the aggravated and simple assaults from the NCVS. Rates exclude homicide victimizations. time of the interview, it is not possible Relationship-specific rates do not add to the total because some victims did not identify to determine whether a woman was their relationship to the offender. separated or divorced at the time of

Homicide of men and women Percent of all Because the NCVS is a survey of Female victims of homicide homicides in 1992 Female Male individuals about their victimization were significantly more likely Homicide victims victims experiences, it does not gather data to be killed by a husband, Victim-offender relationship on homicides. The Uniform Crime ex-husband, or boyfriend Spouse/ex-spouse 18.0% 2.2% Reports (UCR), collected by the Fed- than male victims were to Boy/girlfriend 10.3 1.4 eral Bureau of Investigation, provides be killed by their wife, ex- Other relative 10.2 5.5 Acquaintance/friend 22.0 34.6 the number of homicides known to wife, or girlfriend. In 1992 Stranger 8.6 15.0 police. approximately 28% of female Relationship not identified 30.9 41.3 victims of homicide (1,414 Number of incidents The 1992 UCR reported the victim- women) were known to have Relationship identified 3,454 10,351 offender relationship for 61% of the been killed by their husband, Relationship not identified 1,547 7,824 homicide incidents. The reported ex-husband, or boyfriend. Total number of incidents 5,001 17,635 patterns of intimate perpetration In contrast, just over 3% of Note: Because in 41% of male homicides and 31% for men and women were similar to male homicide victims (637 of female homicides the victim-offender relationship those for other types of victimizations men) were known to have was not identified, readers are urged to use caution collected by the NCVS. been killed by their wife, in interpreting these estimates. Source: FBI, Uniform Crime Reports, 1992. ex-wife, or girlfriend.

4 National Crime Victimization Survey the violence or whether separation or Women's injury and reporting Rapes and other sexual assaults divorce followed the violence. to police against women In general, there was little variation Compared to violence without injury, a The redesigned NCVS now obtains in the extent to which women living higher percentage of violence against information on a broad scope of sex- in urban, suburban, and rural locations women involving injury was reported ual assaults, ranging in severity from experienced violence by intimates. to police. Victimizations that resulted a completed rape to a verbal threat of However, urban women were more in injury were equally likely to be re- sexual assault. Sexual assaults other likely than either suburban or rural ported to police regardless of the rela- than rape were not measured in the women to experience violence by tionship between the victim and earlier victimization survey. strangers. offender. A completed rape is a report of a re- Percent of violent Weapons and violence victimizations spondent physically forced or psycho- against women against women logically coerced to engage in sexual reported to police intercourse. Intercourse is an act of About a fifth of all lone-offender All lone-offender vaginal, anal, or oral penetration by violence against women involved violent crimes the offender(s), including penetration Injured 54% a weapon. Compared to known of- Not injured 39 by a foreign object. (For more infor- fenders, violent offenders who were Victim-offender mation on NCVS methodology as it strangers to the female victim were relationship relates to rape and sexual assault, more likely to carry or use a weapon. Intimate see Methodology on pages 6-7. Also Injured 55% see the forthcoming Criminal Victimi- Percent of violent Not injured 46 victimizations zation in the United States, 1993, Other relative against women NCJ-151657, Appendix.) Injured 52 involving weapons Not injured 39 Victimizations not involving completed All lone-offender violent crimes 20% Acquaintance/friend or attempted sexual intercourse but Injured 50 Victim-offender Not injured 33 having some form of sexual behavior relationship Stranger forced on the victim were categorized Intimate 18% Injured 56 as sexual assault. These crimes in- Other relative 17 Not injured 42 Acquaintance/friend 16 Stranger 30 Table 7. Number and rate of rapes and sexual assaults Injured female victims of crime of female victims age 12 or older, by type of assault, 1992-93 Average Percent of Average annual Women were more likely to be injured annual all rapes/ rate per 1,000 in violent incidents committed by inti- number of sexual females age 12 Type of victimization incidents assaults or older mates than in incidents committed by strangers. However, there was no Rape/Sexual assault 500,200 100% 4.6 significant difference across relation- Completed rape 172,400 34 1.6 Attempted rape 141,200 28 1.3 ship categories in the extent to which Sexual assault with serious injury 23,600 5 0.2 injured victims required medical care. Sexual assault with minor injury 20,700 4 0.2 This lack of difference may reflect less Sexual assault without injury 75,800 15 0.7 severe injuries for intimate victims or Verbal threat of rape 29,200 6 0.3 Verbal threat of sexual assault 37,300 7 0.3 may reflect factors which keep some Note: Detailed numbers and percentage distribution may not add to total women from seeking medical care. because of rounding. Average annual numbers have been rounded to the nearest 100. This table does not include sexual contact without force. Female victims Percent of injured who Table 8. Victim-offender relationship in rapes and sexual assaults required of female victims age 12 or older, by type of assault, 1992-93 Sustain- medical ing injury care Percent of lone-offender victimizations against females Other Acquaint- All lone-offender violent crimes 33% 41% Type of victimization Total Intimate relative ance/friend Stranger Victim-offender Rape/Sexual assault 100% 26% 3% 53% 18% relationship Completed rape 100 39 * 50 8 Intimate 52% 41% Attempted rape 100 24 * 49 23 Other relative 38 35 Sexual assault with injury 100 23 * 42 28 Acquaintance/friend 26 43 Sexual assault without injury 100 13 * 75 * Stranger 20 37 Verbal threat of rape or sexual assault 100 * * 53 42 Note: This table does not include sexual contact without force. Percentage distributions may not total 100% because some victims did not identify the offender's relationship to them and because of rounding. *Ten or fewer sample cases.

Violence against Women: Estimates from the Redesigned Survey 5 cluded assaults and threats, and they This report includes data on series In addition, explicit cuing for rape and generally involved sexual contact, victimizations. A series crime incident other sexual assault is included in the such as the offender's grabbing or is defined by the NCVS as a crime in new screening instrument. fondling of the victim. which a respondent experienced at Definitions of relationships least six similar victimization incidents Sexual assaults also included inci- during the given reference period (pre- Intimates: Includes spouses or ex- dents in which the offender's motive vious 6 months) but could not report spouses, boyfriends and girlfriends, was not clear. If victims reported that the date and details of each clearly or ex-boyfriends and ex-girlfriends. they had been sexually attacked but enough to report them separately. could not or would not say that it was Other relatives: Parents or step- These incidents are recorded as one a rape or an attempted rape, the inci- parents, children or stepchildren, series incident and included in all dent was classified as a sexual as- brothers or sisters, or some other tables. The characteristics of the inci- sault. (For more information about relative. dent are based on those represented screening questions, see Methodology in the last incident in the series. and the comparisons on page 8.) Acquaintances/friends: Friends or The redesigned screening instrument former friends, roommates or board- Each year an estimated 500,000 ers, schoolmates, neighbors, some- women were the victims of some form A goal of the NCVS redesign was to one at work, or some other known of rape or sexual assault (table 7). produce more accurate reporting of nonrelative. Thirty-four percent of these victimiza- incidents of rape and sexual assault Strangers: Anyone not known previ- tions were completed rapes, and an and other crimes committed by inti- additional 28% were attempted rapes. mates and family members. The new ously by the victim. NCVS methodology encourages re- Women were more likely to report Definitions of violent crimes spondents to report incidents of this being raped or sexually assaulted by nature in a number of ways. Ques- Rape: Carnal knowledge through the friends or other acquaintances than tions were added to let respondents use of force or threat of force, includ- by intimates, relatives, or strangers know that the interviewer is interested ing attempts; attempted rape may (table 8). Friends and acquaintances in a broad spectrum of incidents, not consist of verbal threats of rape. It in- committed about half of all rapes and just those involving weapons, severe cludes male as well as female victims. sexual assaults. Intimates committed violence, or violence perpetrated by an additional 26%. Altogether, offend- The definition from the NCVS inter- strangers. In addition to the new ers known to the victim accounted for viewer's manual is as follows: "Rape questions, new methods of cuing re- about three-quarters of all rapes and is forced sexual intercourse and in- spondents about potential experiences sexual assaults against women. cludes both psychological coercion as with victimizations have also been Strangers committed 18% of such well as physical force. Forced sexual added. For example, instead of the assaults. intercourse means vaginal, anal, or yes/no question-and-answer format oral penetration by the offender(s). Methodology of the previous screener, the new This category also includes incidents screener provides an extended list Except for homicide data provided by where the penetration is from a foreign of cues regarding crime victimizations the Uniform Crime Reports, the tables object such as a bottle." and situations in which victimizations in this report include data from the re- might have occurred. (See page 8.) Respondents are asked a series of designed National Crime Victimization Another general change in the new questions about attacks, rapes, at- Survey (NCVS) for 1992 and 1993. screener has been the replacement tempted rapes, sexual attacks, and The NCVS obtains information about of criminal justice terms with more forced or coerced unwanted sex. crimes, including incidents not re- behavior-specific language. These questions are asked directly ported to police, from a continuous, and are accompanied by cues about nationally representative sample of The redesign's enhanced ability the offender (casual acquaintance, households in the United States. to measure sexual assaults and co-worker, neighbor, friend, relative, Approximately 50,000 households and intimate-perpetrated violence and so on). If the respondent replies 100,000 individuals age 12 or older To more accurately capture sexual in the affirmative to any of these ques- are interviewed for the survey annu- assaults and intimate-perpetrated tions or cues, an Incident Report is ally. References in this report to violence, two frames of reference filled out to gather details. In gather- "women" or "females" include adoles- have been added or more explicitly ing these details, interviewers are in- cents, but not children under age 12. defined in the NCVS screener. The structed to ask "Do you mean forced For more information about the NCVS first relates to crimes being committed or coerced sexual intercourse?" to de- sample, see Criminal Victimization in by someone the respondent knows, termine whether the incident should the United States, 1993, NCJ-151657, and the second relates to possible be re-corded as a rape. If the respon- forthcoming, published in an annual locations of a crime or activities the dent requests clarification of the terms series by the Bureau of Justice respondent may have been involved Statistics. in.

6 National Crime Victimization Survey "rape" or "sexual intercourse," inter- Urban: The largest city, or grouping were qualified by phrases such as viewers may read the above definition of cities, in an MSA. In this report, city "somewhat" or "some evidence of a from the interviewer's manual. areas are categorized as those por- difference." Comparisons that failed tions of metropolitan areas located in the 90-percent hypothesis test were Assault: An unlawful physical attack, "central cities." not considered statistically significant. whether aggravated or simple, on a person. It includes attempted assaults Suburban: A county or group of coun- Although the data in this report were with or without a weapon, but ex- ties containing a central city, plus any collected over a 2-year period, some cludes rape, attempted rape, and at- contiguous counties that are linked estimates were based on a relatively tacks involving theft or attempted theft socially and economically to the cen- small number of sample cases, par- (classified as robbery). Severity of tral city. In this report, suburban areas ticularly for certain demographic assaults are classified into two major are categorized as those portions of groups. The data tables note when subcategories: Metropolitan Statistical Areas situated estimates were based on 10 or fewer "outside central cities." unweighted sample cases. Because 1. Simple assault: An attack without standard errors cannot be accurately a weapon resulting either in minor Rural: A county or group of counties computed for such estimates, it is injury (that is, bruises, black eyes, not located inside a Metropolitan Sta- inadvisable to compare them to other cuts, scratches, or swelling) or in un- tistical Area. This category includes estimates. Furthermore, caution determined injury requiring less than a variety of localities, including smaller should be used when comparing 2 days of hospitalization. It also in- cities with populations less than estimates not discussed in the text cludes attempted assault without a 50,000; however, it is primarily com- because since seemingly large differ- weapon and verbal threats of assault. prised of sparsely populated areas. ences may not be statistically signifi- 2. Aggravated assault: An attack Definition of weapons cant at the 95-percent or even the or attempted attack with a weapon 90-percent confidence level. Items such as guns (pistols, revolvers, regardless of whether an injury occur- rifles, and shotguns) and knives are red as well as an attack without a always considered weapons. Other The Bureau of Justice Statistics weapon when serious injury results. objects are considered weapons when is the statistical agency of the Serious injury includes broken bones, they are used as weapons or when U.S. Department of Justice. loss of teeth, internal injuries, loss of respondents felt threatened by them. Jan M. Chaiken, Ph.D., is the consciousness, and any injury requir- director. ing 2 or more days of hospitalization. Calculation of NCVS rates BJS Special Reports address a Sexual assault: A wide range of vic- The rates in this report were annual specific topic in depth from one timizations, separate from rape or at- average rates for 1992-93. The nu- or more datasets that cover many tempted rape. These crimes include merator of a given rate was the sum topics. attacks or attempted attacks generally of the estimated victimizations that involving (unwanted) sexual contact occurred both years for each respec- Ronet Bachman, Ph.D., BJS, and between victim and offender. Sexual tive demographic group; the denomi- Linda E. Saltzman, Ph.D., Family assaults may or may not involve force nator was the sum of the annual and Intimate Violence Prevention and include such things as grabbing population totals for these same years Team, Division of Violence Preven- or fondling. Sexual assault also in- and demographic groups. tion, National Center for Injury cludes verbal threats. Prevention and Control, Centers Application of standard errors Robbery: Completed or attempted for Disease Control and Prevention, The results presented in this report theft, directly from a person, of prop- wrote this report. Tom Hester and were tested to determine whether the erty or cash by force or threat of force, Tina Dorsey edited the report. observed difference between groups with or without a weapon. Marilyn Marbrook, assisted by was statistically significant. Most Jayne Robinson and Yvonne Definitions of locations comparisons mentioned in the report Boston, administered production. passed a hypothesis test at the .05 To define the location categories, this August 1995, NCJ-154348 level of statistical significance (or the report utilizes the Office of Manage- 95-percent confidence level), meaning ment and Budget's concept of Metro- that the estimated difference between politan Statistical Areas (MSA's) and comparisons was greater than twice classifies units of analysis into three the standard error of that difference. segments based on their relationship However, some comparisons were to an MSA: central city, outside cen- significant at the 90-percent confi- tral city, and nonmetropolitan area  dence level only. These comparisons labeled in this report as urban, subur- ban, and rural. A more detailed description of these areas follows:

Violence against Women: Estimates from the Redesigned Survey 7 Appendix: Comparison of new and old screener questions

Screener questions for violent crimes

New (beginning January 1992) Old (1972-92*) 1. Did anyone take something directly from you  1. Has anyone attacked or threatened you in any of these ways by using force, such as by a stickup, mugging, a. With any weapon, for instance, a gun or knife or threat? b. With anything like a baseball bat, frying pan, scissors, 2. Did anyone TRY to rob you by using force or stick  or threatening to harm you? c. By something thrown, such as a rock or bottle 3. Did anyone beat you up, attack you, or hit you d. Include any grabbing, punching, or choking,  with something, such as a rock or bottle? e. Any rape, attempted rape or other type of sexual assault 4. Were you knifed, shot at, or attacked with some  f. Any face to face threats other weapon by anyone at all? OR 5. Did anyone THREATEN to beat you up or g. Any attack or threat or use of force by anyone at all? THREATEN you with a knife, gun, or some other Please mention it even if you were not certain it was weapon, NOT including telephone threats? a crime. 6. Did anyone TRY to attack you in some other way? 2. Incidents involving forced or unwanted sexual acts are often difficult to talk about. Have you been forced or *During 1992 half of the sampled households responded coerced to engage in unwanted sexual activity by to the old questionnaire, and half to the redesigned survey. a. Someone you didn't know before b. A casual acquaintance OR c. Someone you know well.

Screener questions for all types of crimes

New Old 1. Was anything stolen from you while you were away 1. Were you attacked or threatened OR did you have from home, for instance, at work, in a theater or something stolen from you restaurant, or while traveling. a. At home including the porch or yard b. At or near a friend's relative's, or neighbor's home 2. Did you call the police to report something that c. At work or school happened to YOU that you thought was a crime? d. In place such as a storage shed or laundry room, 3. Did anything happen to YOU that you thought  a shopping mall, restaurant, bank or airport was a crime, but did NOT report to the police? e. While riding in any vehicle f. On the street or in a parking lot g. At such places as a party, theater, gym, picnic area, The table below compares results from this report, based bowling lanes, or while fishing or hunting. on the redesigned NCVS methodology, with results from an OR earlier report, using data from the old questionnaire (Violence h. Did anyone ATTEMPT to attack or attempt to steal against Women, NCJ-145325, January 1994). anything belonging to you from any of these places? The new survey instrument captures more incidents of violence 2. People often don't think of incidents committed by across all categories than did the old questionnaire. The extent someone they know. Did you have something stolen from you OR were you attacked or threatened by of increase, however, does vary according to the sex of the vic- a. Someone at work or school tim and the victim- b. A neighbor or friend Average annual offender relationship. c. A relative or family member rate of violent The rates presented in d. Any other person you've met or known? victimizations per 1,000 persons this table are based on 3. Did you call the police to report something that Female Male different time periods; happened to YOU which you thought was a crime? Old survey methodology, however, analysis of 4. Did anything happen to you which you thought 1987-91 data for overlapping was a crime, but did NOT report to the police? Intimate 5.4 .5 periods confirms these Other relative 1.1 .7 Acquaintance/friend 7.6 13.0 patterns. (See Criminal Stranger 5.4 12.2 Victimization in the New NCVS methodology, United States, 1993, 1992-93 NCJ-151657, Appendix, Intimate 9.3 1.4 Other relative 2.8 1.2 forthcoming, by the Acquaintance/friend 12.9 17.2 Bureau of Justice Stranger 7.4 19.0 Statistics.)

Appendix table

8 National Crime Victimization Survey