sign in sign up
<<
Home , Linda Saltzman, Prevalence, Rape, Violence, Violence against women

NT OF ME J T US U.S. Department of Justice R T A I P C E E D

B

O J C S F A Office of Justice Programs V M F O I N A C I J S R E BJ G O OJJ DP O F PR National Institute of Justice JUSTICE

National Institute of Justice Centers for Control and Prevention R e s e a r c h i n B r i e f November 1998 , , and Consequences of Against Partners in Research on

The findings from this second in a Women: Findings From the National series of reports issued jointly by the National Institute of Justice and Violence Against Women Survey Centers for Disease Control and Pre- vention paint a picture of the vio- by Patricia Tjaden and Nancy Thoennes lence against women as seen daily by emergency room, enforcement, and court professionals. Violence against women first came to be was conducted from November 1995 to May viewed as a serious social problem in the 1996. To provide a context in which to place Accurate information of the kind early 1970s, in part because of the re- women’s experiences, the National Violence presented in this Research in Brief emergence of the Women’s Movement.1 Against Women (NVAW) Survey sampled and in the full report contributes to In unprecedented numbers, scholars both women and men. Thus the survey pro- our Nation’s efforts to stop violence trained in such diverse disciplines as vides comparable data on women’s and men’s against women and alleviate the suf- , literature, law, and experiences with violent victimization. fering of its victims, much of which began to examine violence against women occurs in the of homes rather 2 Respondents to the NVAW Survey were than in public. in the context of a feminist ideology. De- spite the resulting outpouring of research queried about: This information will help the U.S. on violence against women, particularly ● Physical they experienced as Department of Justice direct and in the areas of and intimate partner children by adult caretakers. evaluate efforts to work within the violence, many gaps remain in our under- legal system to protect Americans standing of violence against women.3 ● Physical assault they experienced as and provide justice for those who adults by any type of perpetrator. violate the rights of others. This Until now, empirical data on the relation- ● Forcible rape or they experi- knowledge also will help the U.S. ship between certain types of violence enced at any time in their by any type Department of and against women, such as childhood of perpetrator. Services direct efforts toward pre- victimization and subsequent adult venting the physical and psy- victimization, have been limited. Reli- chological that results from Respondents who disclosed victimization violence. Ultimately, the knowledge able information on minority women’s were asked detailed questions about the gained from surveys like this one has experiences with violence and on the characteristics and consequences of their the power to call the Nation’s atten- consequences of violence against women, victimization, including sustained tion to how pervasive the problem is including rates of injury and use of and use of medical services (see “Survey and how many women face dangers medical services, is also limited.4 Screening Questions,” page 13 and in our communities. “Survey Methodology,” page 14). To further an understanding of violence Jeremy Travis, Director against women, the National Institute of Several features set the NVAW Survey National Institute of Justice Justice (NIJ) and the Centers for Disease apart from other victimization surveys: Control and Prevention (CDC) jointly Mark Rosenberg, Director ● State-of-the-art techniques protected National Center for Injury sponsored—through a grant to the Center for Policy Research—a national telephone confidentiality and minimized the poten- Prevention and Control, Centers tial for retraumatizing victims of violence for Disease Control and Prevention survey on violence against women, which R e s e a r c h i n B r i e f

Issues and Findings

Discussed in this Brief: Results of a na- time in their life and 0.3 percent said they were raped and/or physically assaulted tionally representative telephone survey of experienced a completed or attempted rape since age 18 were assaulted by a current 8,000 women and 8,000 men about their in the previous 12 months. Of the women or former husband, cohabiting partner, experiences with rape, physical assault, who reported being raped at some time in or date, compared with 18 percent of the and stalking cosponsored by the National their lives, 22 percent were under 12 years men. It is therefore imperative that strate- Institute of Justice and the Centers for old and 32 percent were 12 to 17 years old gies for preventing violence against Disease Control and Prevention and con- when they were first raped. Given these women should focus on ways of protect- ducted by the Center for Policy Research. findings, research and intervention strategies ing women from risks posed by current should focus on perpetrated against and former intimates. Key issues: This study provides empirical children and adolescents. data on the prevalence and incidence of ● Women are significantly more likely rape, physical assault, and stalking; the ● Among women of different racial and than men to be injured during an prevalence of male-to- and female- ethnic backgrounds, the difference in the assault: 32 percent of the women to-male intimate partner violence; the prevalence of reported rape and physical as- and 16 percent of the men who were prevalence of rape and physical assault sault is statistically significant: American In- raped since age 18 were injured during among women of different racial and eth- dian/Alaska Native women were most likely their most recent rape; 39 percent of nic backgrounds; the rate of injury among to report rape and physical assault victimiza- the women and 25 percent of the men rape and physical assault victims; and tion, while Asian/Pacific Islander women who were physically assaulted since age injured victims’ use of medical services. were least likely to report rape and physical 18 were injured during their most recent assault victimization. Hispanic women were physical assault. About one in three Key findings and policy implications: less likely to report rape victimization than women who were injured during a rape Although the survey sampled both women non-Hispanic women. More research is or physical assault required medical and men, this report focuses on women’s needed to determine how much of the care. To better meet the medical needs experiences with violence. Analysis of sur- difference can be explained by the of women who are victims of violence, vey data produced the following results: respondent’s willingness to report informa- medical professionals should receive tion to interviewers and how much by social, comprehensive training on the physical ● Using a definition of physical assault demographic, and environmental factors. consequences of violence against that includes a range of behaviors, from women and appropriate treatment slapping and hitting to using a gun, the ● Women experience significantly more strategies. survey found that physical assault is wide- partner violence than men do: 25 percent of spread among American women: 52 per- surveyed women, compared with 8 percent ● Using a definition of stalking that re- cent of surveyed women said they were of surveyed men, said they were raped quires the victim to feel a high level of physically assaulted as a by an adult and/or physically assaulted by a current or fear, the survey found that stalking is more caretaker and/or as an adult by any type former spouse, cohabiting partner, or date prevalent than previously thought: 8 per- of perpetrator; 1.9 percent of surveyed in their lifetime; 1.5 percent of surveyed cent of surveyed women and 2 percent of women said they were physically assaulted women and 0.9 percent of surveyed men surveyed men said they were stalked at in the previous 12 months. Based on said they were raped and/or physically as- some time in their life; 1 percent of sur- these estimates, approximately 1.9 million saulted by such a perpetrator in the previous veyed women and 0.4 percent of surveyed women are physically assaulted annually in 12 months. According to survey estimates, men said they were stalked in the previous the United States. More research is needed approximately 1.5 million women and 12 months. According to survey estimates, to understand the relationship between 834,700 men are raped and/or physically approximately 1 million women and physical assault experienced in childhood assaulted by an intimate partner annually 371,000 men are stalked annually in the and physical assault experienced in in the United States. Because women are United States. Given these findings, stalk- adulthood. also more likely to be injured by intimate ing should be considered a serious criminal partners, research aimed at understanding justice and concern. ● Using a definition of rape that includes and preventing partner violence against forced vaginal, oral, and anal intercourse, women should be stressed. Target audience: Criminal justice and the survey found that rape is a com- public health researchers and practitioners, mitted primarily against : 18 percent ● Violence against women is primarily part- legislators, policymakers, and intervention of women surveyed said they experienced ner violence: 76 percent of the women who planners at all levels of government. a completed or attempted rape at some

2 R e s e a r c h i n B r i e f and jeopardizing the safety of respon- and 1 of 33 U.S. men has experienced experienced a completed or attempted dents. In addition to lessening the an attempted or completed rape as a rape at some time in their life (see ex- possibility that respondents would be child and/or an adult; specifically, hibit 1). These findings are similar to harmed by their participation in the 18 percent of surveyed women and findings from the National Health and survey, these techniques increased 3 percent of surveyed men said they Social Life Survey, which found that the quality of the information being gathered. Other Publications in the Series ● Information about both the preva- To obtain other publications in the series ● Final Report on Prevalence, Incidence, and lence (lifetime and annual) and from the National Violence Against Women Consequences of Violence Against Women incidence of violence was gathered. Survey: by Patricia Tjaden and Nancy Thoennes, Thus victimization estimates from the Atlanta, GA: Centers for Disease Control NVAW Survey can be compared with ● Stalking in America: Findings From the National Violence Against Women Survey, and Prevention, National Center for Injury victimization estimates from a wide Prevention and Control, forthcoming 1998. variety of other surveys. Research in Brief, by Patricia Tjaden and Nancy Thoennes, Washington, D.C.: U.S. Learn about the availability of the Final ● Behaviorally specific questions Department of Justice, National Institute of Report and other CDC and screened respondents for rape, physical Justice, April 1998 (NCJ 169592). assault, and stalking victimization. intimate violence prevention activities These questions were designed to leave Visit NIJ’s Web site at http:// by visiting the National Center for Injury little doubt in the respondents’ minds as www.ojp.usdoj.gov/nij or contact the Prevention and Control’s Web site at to the type of information being sought. National Criminal Justice Service at P.O. http://www.cdc.gov/ncipc/dvp/fivpt. Box 6000, Rockville, MD 20849–6000, ● Additional reports in the series are Detailed information about the (800) 851–3420 or (301) 519–5500, or send forthcoming. characteristics and consequences of an e-mail message to [email protected]. victimization for each type of perpetra- tor identified by the respondents was gathered. While this approach created Exhibit 1: Persons Raped or Physically Assaulted in Lifetime by Sex of Victim a very complicated data set, it also created the opportunity to track vic- Percentage Numbera timizations by the same perpetrator Women Men Women Men (e.g., the victim’s first ex-husband). Type of Assault (n=8,000) (n=8,000) (100,697,000) (92,748,000) This Research in Brief summarizes Total rapeb 17.6 3.0 17,722,672 2,782,440 the survey’s findings on the prevalence Completedb 14.8 2.1 14,903,156 1,947,708 and incidence of rape, physical as- Attempted onlyb 2.8 0.9 2,819,516 834,732 sault, and stalking; the rate of injury Total physical assaultb 51.9 66.4 52,261,743 61,584,672 among rape and physical assault vic- Threw somethingb 14.0 22.4 14,097,580 20,775,552 tims; and injured victims’ use of medi- Pushed, grabbed, shovedb 30.6 43.5 30,813,282 40,345,380 cal services. Complete details about Pulled hair 19.0 17.9 19,132,430 16,601,892 men’s and women’s experiences with Slapped, hitb 43.0 53.7 43,299,710 49,805,676 stalking are contained in an earlier Kicked, bitb 8.9 15.2 8,962,033 14,097,696 report. This Research in Brief therefore Choked, tried to drownb 7.7 3.9 7,753,669 3,617,172 focuses primarily on rape and physical Hit with objectb 21.2 34.7 21,347,764 32,183,556 assault. The full report detailing the Beat upc 14.1 15.5 14,198,277 14,375,940 prevalence, incidence, and conse- Threatened with gunb 6.2 13.1 6,243,214 12,149,988 quences of violence against women will Threatened with knifeb 5.8 16.1 5,840,426 14,932,428 be available through the CDC (see Used gunb 2.6 5.1 2,618,122 4,730,148 “Other Publications in the Series”). Used knifeb 3.5 9.6 3,524,395 8,903,808 Rape and/or physical assaultb 55.0 66.8 55,383,350 61,955,664 Rape: Prevalence and incidence a. Based on estimates of women and men in the United States aged 18 years and older, Using a definition of rape that includes U.S. Bureau of the Census, Current Population Survey, 1995. forced vaginal, oral, and , the b. Differences between women and men are statistically significant: p-value ≤ .001. survey found that 1 of 6 U.S. women c. Differences between women and men are statistically significant: p-value ≤ .01.

3 R e s e a r c h i n B r i e f

other forms of , such as Exhibit 2: Persons Raped or Physically Assaulted in Previous 12 Months grabbing, fondling, and verbal threats by Sex of Victim of a sexual nature. Percentage Numbera In addition, the screening questions Women Men Women Men used by the two surveys differ substan- Type of Violence (n=8,000) (n=8,000) (100,697,000) (92,748,000) tially. The NVAW Survey uses five Rapeb 0.3 0.1 302,091 92,748 questions to screen respondents Physical assaultb 1.9 3.4 1,913,243 3,153,432 for rape victimization, while the NCVS 8 Rape and/or physical assaultb 2.1 3.5 2,114,637 3,246,180 uses only two questions. Although empirical data on this issue are lim- a. Based on estimates of men and women in the United States aged 18 and older, U.S. Bureau of ited, some researchers assume that the Census, Current Population Survey, 1995. increasing the number of screening b. Differences between men and women are statistically significant: p-value ≤ .001. questions increases disclosure rates.9 Furthermore, the NVAW Survey 22 percent of surveyed women and recent year for which estimates are screening questions are more explicit 2 percent of surveyed men had been available—are 432,100 rapes and than those used by the NCVS.10 “forced to do something sexual” at sexual of women age 12 and some time in their lifetime.5 older and 32,900 rapes and sexual as- Moreover, published NCVS estimates saults of men age 12 and older.7 How- count series victimizations—reports of The survey also found that 0.3 percent ever, direct comparisons between the six or more within a 6-month of surveyed women and 0.1 percent of NVAW Survey and the NCVS are diffi- period for which the respondent can- surveyed men said they experienced to make. Estimates reported by the not recall details of each crime—as a a completed or attempted rape in the two surveys refer to somewhat different single victimization. Thus, published previous 12 months.6 These estimates populations and sexual victimizations. NCVS estimates of the number of rape equate to approximately 302,100 Moreover, the two surveys differ sub- and sexual assault victimizations are women and 92,700 men who are forc- stantially with respect to other method- lower than would be obtained by in- ibly raped each year in the United ological issues. cluding all incidents reported to its sur- States (see exhibit 2). vey interviewers. To produce NCVS Specifically, the NVAW Survey esti- estimates for direct comparison with Because some rape victims experi- mates include victimizations against the NVAW Survey estimates, each enced more than one rape in the persons age 18 and older, while the crime in a series of victimizations 12 months preceding the survey, the NCVS estimates include victimizations reported to the NCVS interviewers incidence of rape (number of separate against persons 12 years of age and would have to be counted separately. victimizations) exceeded the preva- older. The NVAW Survey estimate for lence of rape (number of rape victims). forcible rape is being compared to an Finally, the sampling errors associated Specifically, women who were raped in NCVS estimate that includes rape and with the estimates from the NVAW the previous 12 months averaged 2.9 rapes, while men averaged 1.2 rapes. According to survey estimates, approxi- Exhibit 3: Number of Rapes and Physical Assaults Perpetrated in Previous mately 876,100 rapes were perpetrated 12 Months by Sex of Victim against women and approximately 111,300 rapes were perpetrated against Average Number men in the United States during the Number of of Victimizations Total Number Type of Violence Victims per Victima of Victimizations 12 months preceding the survey (see exhibit 3). Women Rape 302,091 2.9 876,064 The annual rape victimization estimates Physical assault 1,913,243 3.1 5,931,053 generated by the NVAW Survey are higher than comparable victimization Men estimates generated by the Bureau of Rape 92,748 1.2 111,298 Justice Statistics (BJS) National Crime Physical assault 3,153,432 2.5 7,883,580 Victimization Survey (NCVS). The a. The standard error of the mean for each estimate in this column is 1.431, 0.167, 0.539, 0.225, NCVS estimates for 1994—the most respectively.

4 R e s e a r c h i n B r i e f

exhibit 3). These estimates probably Exhibit 4: Women Raped or Physically Assaulted in Lifetime by Race/ underestimate the number of physical Ethnicity of Victim assaults perpetrated against women Women Raped or Physically Assaulted in Lifetime (%) and men annually because the NVAW Survey categorized victimizations Asian/ American involving both rape and physical as- African- Pacific Indian/ Mixed sault only as rapes. Total American Islander Alaska Native Race (n=7,850) (n=6,452) (n=780) (n=133) (n=88) (n=397) The NVAW Survey estimates of annual

a physical assaults are greater than the Rape 18.2 17.7 18.8 6.8 34.1 24.4 NCVS estimates. The NCVS estimate Physical assaultb 51.8 51.3 52.1 49.6 61.4 57.7 for 1994 is 4.1 million aggravated and Rape and/or simple assaults of women age 12 and physical assaultb 55.0 54.5 55.1 51.9 64.8 61.2 older and 5.7 million aggravated and simple assaults of men age 12 and a. Differences between racial/ethnic groups are statistically significant: p-value ≤ .001. older.12 Comparisons between the b. Differences between racial/ethnic groups are statistically significant: p-value ≤ .05. NVAW Survey and NCVS estimates of physical assault are confounded by the Survey and the NCVS would have to surveyed men said they were physi- same methodological differences dis- be compared. This is important given cally assaulted as a child by an adult cussed earlier and are being addressed the relatively high margin of error caretaker and/or as an adult by any in ongoing research funded by NIJ, associated with NVAW Survey type of perpetrator. BJS, and CDC that is comprehensively estimates of the average number of rape reviewing the two surveys’ estimates victimizations experienced by male For both women and men, the most fre- for physical violence. and female victims (see footnote in quently reported physical assault was exhibit 3). Comparisons of sampling slapping and hitting, followed by push- The risk of violence varies errors would help determine whether ing, grabbing, shoving, and hitting with among minority women the estimates are actually different or an object. Relatively few respondents whether apparent differences are not reported that an adult caretaker or The survey found that American statistically significant. other adult pulled their hair or threw Indian/Alaska Native women were something at them that could hurt. Still significantly more likely to disclose A study now under way and funded by fewer reported that an adult caretaker rape and physical assault victimization NIJ, BJS, and CDC will provide more or other adult choked or almost than women of other racial/ethnic comprehensive information about the drowned them, kicked or bit them, backgrounds, while Asian/Pacific Is- differences between the two surveys, beat them up, threatened them with a lander women were significantly less including error ranges for the esti- gun or knife, or used a gun or knife on likely to report rape and physical as- mates. Using data from the two sur- them (see exhibit 1). sault victimization (see exhibit 4). veys, the study is calculating estimates using the same counting rules and the The NVAW Survey also found that These findings must be viewed with same age populations. The results of 1.9 percent of surveyed women and caution, however, given the relatively this detailed comparison will be pub- 3.4 percent of surveyed men said they small numbers of American Indian/ lished in early 1999. were physically assaulted in the previ- Alaska Native and Asian/Pacific Is- ous 12 months. These estimates equate lander women included in the sample. Physical assault: Prevalence to approximately 1.9 million women Because information on violence and incidence and 3.2 million men who are physi- against American Indian/Alaska cally assaulted annually in the United Native women is limited, it is difficult The NVAW Survey used a modified States (see exhibit 2). Female victims to explain why they report more rape version of the Conflict Tactics Scale11 averaged 3.1 assaults and male victims and physical assault victimization. to query respondents about a variety averaged 2.5 assaults per year, which A previous study found that Native- of physical assaults they may have equate to approximately 5.9 million American couples were significantly suffered. Responses revealed that physical assaults perpetrated against more violent than their white counter- physical assault is widespread in women and 7.9 million physical as- parts.13 Another study found that the American society: 52 percent of saults perpetrated against men in the overall rates for Native surveyed women and 66 percent of 12 months preceding the survey (see Americans were approximately two

5 R e s e a r c h i n B r i e f

have produced contradictory conclu- Exhibit 5: Women Raped or Physically Assaulted in Lifetime by Hispanic/ sions;17 hence, findings from the Non-Hispanic Origin of Victim NVAW Survey neither confirm nor Women Raped or Physically Assaulted in Lifetime (%) contradict earlier findings. Total Hispanica Non-Hispanic Rape is primarily a crime (n=7,945) (n=628) (n=7,317) against youth Rapeb 18.1 14.6 18.4 Physical assault 51.9 53.2 51.8 It has been previously reported that rape in America is a “tragedy of youth” Rape and/or physical assault 55.1 54.9 55.1 because the majority of rapes occurs 18 a. Persons of Hispanic origin may be of any race. against children and adolescents. b. Differences between Hispanics and non-Hispanics are statistically significant: p-value ≤ .05. Results from the survey support this assertion: More than half (54 percent) of the female rape victims identified and demographic factors intersect with by the survey were under 18 years old Exhibit 6: Women Victims’ Age at race and ethnicity to produce differences when they experienced their first rape Time of First Rapea in rape and physical assault prevalence. (see exhibit 6). Of the women disclos- Finally, research is needed to determine ing rape, 22 percent were under 12 whether victimization rates vary years old when they experienced their significantly among women of diverse first rape, while 32 percent were 12 to 18–24 Years 29.4% American Indian tribes and Alaska 17 years old. Native communities. 12–17 Years Several authors have reported that 32.4% Because information is also lacking on women who are sexually assaulted as violence against Asian/Pacific Islander children and adolescents are at greater <12 Years women, it is unclear why they report risk of being sexually assaulted as 21.6% 19 > 25 Years less rape and physical assault victim- adults. Results from the NVAW 16.6% ization. Information from focus groups Survey provide further of and nonrandom samples of Asian- the link between sexual assault as a American women suggests that minor and subsequent sexual assault a. n=1,323 women victims. intimate partner violence is a major as an adult: 18 percent of the women concern among these women.16 How- who reported being raped before age ever, there are no accurate estimates 14 18 said they were also raped after the times greater than U.S. national rates. of intimate partner violence among age of 18, compared with 9 percent of Information from this study published Asian-American or Pacific Islander the women who did not report being elsewhere shows that Native-American women, and the differences between raped before the age of 18. women report significantly more Asian-American and Pacific Islander stalking victimization than do women women in rape and physical assault Women experience more of other racial/ethnic backgrounds.15 prevalence also may be significant. partner violence than men do Thus there is some evidence that Such differences, however, cannot Native Americans are at significantly be discerned from the NVAW Survey The NVAW survey found that women greater risk of violence than other because data on these two groups were are significantly more likely to be Americans. combined. Clearly, more research is assaulted by an intimate partner than men are. Twenty-five percent of sur- How much of the difference in rape needed on violence against Asian- American and Pacific Islander women. veyed women, compared with 8 per- and physical assault prevalence cent of surveyed men, said they were can be explained by differences in The survey found that Hispanic women raped and/or physically assaulted by willingness to report victimization to were less likely to report rape victim- a current or former spouse, cohabiting interviewers and how much can be ization than were non-Hispanic women partner, or date at some time in their explained by actual victimization expe- (see exhibit 5). Previous studies com- life; 1.5 percent of all surveyed women riences is unclear and requires further paring violence prevalence among compared with 0.9 percent of all study. Research is also needed to Hispanic and non-Hispanic women surveyed men said they were raped ascertain how social, environmental,

6 R e s e a r c h i n B r i e f

and/or physically assaulted by such Exhibit 7: Persons Raped or Physically Assaulted by an Intimate Partner in a partner in the previous 12 months. Lifetime and in Previous 12 Months by Sex of Victim Based on U.S. Census estimates of the number of women and men in a In Lifetime the country, these findings equate to Percentage Numberb approximately 1.5 million women and 834,700 men who are raped and/or Women Men Women Men Type of Violence (n=8,000) (n=8,000) (100,697,000) (92,748,000) physically assaulted by an intimate partner annually in the United States Rapec 7.7 0.3 7,753,669 278,244 (see exhibit 7). Physical assaultc 22.1 7.4 22,254,037 6,863,352 Rape and/or physical assaultc 24.8 7.6 24,972,856 7,048,848 Results from the survey show that most physical assaults perpetrated against In Previous 12 Monthsa women by intimate partners consist of pushing, grabbing, shoving, slapping, Percentage Numberb and hitting. Fewer women reported that Women Men Women Men an intimate partner threw something at Type of Violence (n=8,000) (n=8,000) (100,697,000) (92,748,000) them that could hurt, pulled their hair, kicked or bit them, hit them with an e e Rape 0.2 – 201,394 – object, or beat them up. Relatively d Physical assault 1.3 0.9 1,309,061 834,732 few reported that an intimate partner d Rape and/or physical assault 1.5 0.9 1,510,455 834,732 threatened them with a knife or gun. a. Intimate partner includes current and former spouses, opposite-sex cohabiting partners, same-sex Only a negligible number reported that cohabiting partners, dates, and boyfriends/girlfriends. an intimate partner actually used a b. Based on estimates of men and women in the United States aged 18 years and older, knife or gun on them (see exhibit 8). U.S. Bureau of the Census, Current Population Survey, 1995. c. Differences between women and men are statistically significant: p-value ≤ .001. It is important to note that differences d. Differences between women and men are statistically significant: p-value ≤ .05. between women’s and men’s rates of e. The number of men rape victims was insufficient to reliably calculate prevalence estimates. physical assault by an intimate partner become greater as the seriousness of the assault increases. For example, women were two to three times more Exhibit 8: Percentage of Persons Physically Assaulted by an Intimate likely than men to report that an inti- Partner in Lifetime by Type of Assault and Sex of Victima mate partner threw something that Women Men could hurt or pushed, grabbed, or Type of Assault (n=8,000) (n=8,000) shoved them. However, they were 7 to 14 times more likely to report that an Total physical assault by intimate partnerb 22.1 7.4 intimate partner beat them up, choked Threw somethingb 8.1 4.4 or tried to drown them, threatened Pushed, grabbed, shovedb 18.1 5.4 them with a gun, or actually used a Pulled hairb 9.1 2.3 gun on them (see exhibit 8). Slapped, hitb 16.0 5.5 Kicked, bitb 5.5 2.6 The NVAW Survey finding that women Choked, tried to drownb 6.1 0.5 are significantly more likely than men Hit with objectb 5.0 3.2 to be victimized by intimate partners Beat upb 8.5 0.6 contradicts findings from some previ- Threatened with gunb 3.5 0.4 ous national surveys that found men Threatened with knifeb 2.8 1.6 and women to be at nearly equal risk Used gunb 0.7 0.1 of physical assault by an intimate part- 20 Used knife 0.9 0.8 ner. The finding, however, supports data from the NCVS, which have con- a. Intimate partner includes current or former spouses, opposite-sex cohabiting partners, same-sex sistently shown women to be at greater cohabiting partners, dates, and boyfriends/girlfriends. risk of intimate partner violence.21 b. Differences between women and men are statistically significant: p-value ≤ .001.

7 R e s e a r c h i n B r i e f

Violence against women is primarily partner violence. The Exhibit 9: Distribution of Adult Rape and Physical Assault Victims by survey confirms previous reports that Victim-Perpetrator Relationship and Sex of Victim

U.S. women are primarily raped and/or 80 physically assaulted by intimate 76 22 Women Victims (n=2,669) partners: 76 percent of the women 70 who were raped and/or physically Men Victims (n=3,606) assaulted since the age of 18 were 60 60 assaulted by a current or former husband, cohabiting partner, or date; 50 17 percent were victimized by an ac- quaintance, such as a friend, neighbor, or coworker; 14 percent were victimized 40 by a stranger; and 9 percent were 32.3 victimized by a relative other than a of Victims Percentage 30 husband (see exhibit 9). By comparison, 17.9 the survey found that U.S. men are pri- 20 16.8 14.1 marily raped and physically assaulted 8.6 by strangers and acquaintances, not 10 6.8 intimate partners. 0 Intimate Relative Other Acquaintanceb Strangerb Violence against women is Partnera, b Than Spousec primarily male violence. The survey also found that most violence perpe- a. Intimate partner includes current and former spouses, same-sex cohabitating partners, opposite-sex trated against adults is perpetrated by cohabitating partners, dates, and boyfriends/girlfriends. males: 93 percent of the women and b. Differences between women and men are statistically significant: p-value ≤ .001. 86 percent of the men who were raped c. Differences between women and men are statistically significant: p-value ≤ .01. and/or physically assaulted since the Note: Total percentages by sex of victim exceed 100 because some victims had multiple perpetrators. age of 18 were assaulted by a male. In comparison only 11 percent of these women and 23 percent of these men were assaulted by a female (see exhibit 10). Given these findings, adult vio- Exhibit 10: Distribution of Adult Rape and Physical Assault Victims by lence prevention strategies should Sex of Perpetrator and Sex of Victim focus primarily on the risks posed by 100 93.4 male perpetrators. 85.9 Rate of injury among rape 80 and physical assault victims 60 To generate information on violence- related injuries, respondents disclos- 40 ing rape and physical assault were 23.3 Percentage of Victims Percentage asked whether they were injured during 20 their most recent victimization. The 10.8 survey found that women who were 0 raped since age 18 were nearly twice Women Victims (n=2,494) Men Victims (n=3,606) as likely as their male counterparts to Female Perpetrator Male Perpetrator report an injury other than the rape itself (32 and 16 percent, respectively) Note: Total percentages by sex of victim exceed 100 because some victims had multiple perpetrators. (see exhibit 11). Similarly, women who were physically assaulted since age 18 were significantly more likely than

8 R e s e a r c h i n B r i e f

their male counterparts to report that Exhibit 11: Distribution of Adult Rape and Physical Assault Victims by they were injured during their most Injury, Type of Medical Care Received, and Sex of Victima recent physical assault (39 and 25 per- Rape Physical Assault cent, respectively). When only physi- Victims (%) Victims (%) cal assaults by intimates are considered, Injury/Medical Care Women Men Women Men the difference between injury rates for women and men is even greater (41 and Was victim injured? (n=734) (n=62) (n=1,862) (n=2,972) 19 percent, respectively). Yes 31.5c 16.1 39.0c 24.8 No 68.5 83.9 61.0 75.2 Most of the adult rape and physical Did injured victim receive assault victims (both men and women) medical care?d (n=236) (n=722) (n=736) Yes 35.6 —b 30.2 37.1 who reported being injured sustained No 64.4 —b 69.8 62.9 relatively minor injuries, such as Type of medical care receivede (n=84) (n=218) (n=306) scratches, , and welts. Rela- Hospital 81.9 —b 76.1 85.7 tively few sustained more serious types Physician 54.8 —b 52.8 42.1 of injuries, such as broken bones, dislo- Dental 16.9 —b 9.6 10.6 cated joints, concussions, lacerations, Ambulance/paramedic 19.0 —b 17.5 23.5 or bullet (see exhibit 12). Physical 16.7 —b 9.2 12.8 Type of hospital care receivedf (n=68) (n=166) (n=234) Victims’ use of medical services. Emergency room 50.0 —b 61.4 66.7 The survey found that 36 percent of Outpatient 36.8 —b 22.3 22.6 the women injured during their most Overnight 13.2 —b 15.1 10.3 recent rape since age 18 and 30 per- a. Estimates are based on the most recent victimization since the age of 18. cent of the women injured during their b. The number of men rape victims was insufficient to reliably calculate medical utilization estimates. most recent physical assault since c. Differences between women and men are statistically significant: p-value ≤ .001. age 18 received some type of medical d. Estimates are based on responses from victims who were injured. treatment (e.g., ambulance/paramedic e. Estimates are based on responses from victims who received medical care. services, treatment in a hospital emer- f. Estimates are based on responses from victims who received hospital care. gency department, physical therapy) Note: Total percentages for type of medical and hospital care received exceed 100 because some (see exhibit 11). Some medically victims had multiple forms of medical/hospital care. treated victims received more than one type of medical treatment (e.g., hospi- Exhibit 12: Percentage of Injured Adult Rape and Physical Assault Victims Who talization as well as outpatient physi- Sustained Specific Types of Injuries: Men and Women Combineda cal therapy). Others received one type of medical treatment more than once—for example, three nights in the Scratch, , welt 72.6 66.6 hospital or 10 physical therapy ses- 6.2 Laceration, knife 16.9 sions. Hence, the annual number of Broken bone, dislocated joint 14.1 medical treatments provided to rape 12.2 and physical assault victims exceeds Head, 6.6 Rape victims (n=241) 10.1 the annual number of rapes and physi- Physical assault victims (n=1,463) Sore muscle, sprain, strain 5.8 5.5 cal assaults that resulted in treatment. Internal injury 5.8 0.8 Exhibit 13 provides estimates of the 3.3 Chipped or broken tooth 1.8 average number of nights spent in the Burnb hospital or the average number of vis- 0.7 its made to specific medical providers Knocked unconsciousb 0.8 by rape and physical assault victims. b Bullet wound 1.8 These estimates are based on responses 0 1020304050607080 from victims who received the specific type of medical care considered. For Percentage of Victims example, the estimate of the average a. Estimates are based on the most recent victimization since the age of 18. number of nights spent in the hospital b. The number of rape victims was insufficient to reliably calculate injury estimates. by female rape victims (3.6) is based

9 R e s e a r c h i n B r i e f only on responses by female rape vic- According to estimates generated victims. Because these groups were tims who were treated in a hospital on by the NVAW Survey, hospital emer- excluded from the NVAW Survey esti- an inpatient basis. It should be noted gency department personnel treated mate, it is not surprising that it is lower that some of these average frequency approximately 1.3 million adults for than the SIVV estimate of 1.4 million. estimates have a relatively high mar- injuries related to rapes and physical gin of error (see footnotes c through e assaults in the 12 months preceding Stalking is a significant in exhibit 13) and should be viewed the survey (128,700 female rape vic- social problem with caution. tims, 546,900 female physical assault victims, and 588,300 male physical Using a definition of stalking that Exhibit 14 presents estimates of the assault victims). This figure is some- requires victims to feel a high level of number of female rapes and physical what lower than an estimate generated fear, the survey found that stalking is assaults resulting in injuries annually, from the Study of Injured Victims of much more prevalent than previously as well as estimates of the number of Violence (SIVV), a hospital record- thought: 8 percent of surveyed women specific types of medical treatment to extraction study conducted for the and 2 percent of surveyed men said these victims annually. It is important Bureau of Justice Statistics by the U.S. they were stalked at some time in their to note that because women are prima- Consumer Product Safety Commission. life; 1 percent of surveyed women rily raped and physically assaulted by The SIVV found that during 1994, and 0.4 percent of surveyed men said intimate partners, the injury and medi- hospital per- they were stalked in the previous 12 cal utilization estimates for women sonnel treated an estimated 1.4 million months. Based on U.S. Census esti- presented in exhibits 13 and 14 pertain people for injuries from confirmed or mates of the number of women and men mostly to rapes and physical assaults suspected interpersonal violence.23 in the country, these findings equate to perpetrated by current and former hus- approximately 1 million women and bands, cohabiting partners, dates, and Included in the SIVV estimate (but 371,000 men who are stalked annually boyfriends. Because men are primarily excluded from the NVAW Survey esti- in the United States (see exhibit 15). raped and physically assaulted by mate) are hospital emergency depart- male strangers and acquaintances, the ment treatments to victims of all ages If a less stringent definition of stalking injury and medical utilization estimates (including children and adolescents), is used, one requiring victims to feel for men pertain primarily to violence victims of suspected interpersonal vio- only somewhat frightened or a little perpetrated by men other than intimate lence, male rape victims, and male frightened of their assailant’s behavior, partners. and female sexual assault and the lifetime stalking prevalence rate increases from 8 to 12 percent for women and from 2 to 4 percent for Exhibit 13: Average Number of Visits/Overnights in Hospital for Rape men; the annual stalking prevalence and Physical Assault Victims by Type of Medical Care and Sex of Victima rate increases from 1 to 6 percent for women and from 0.4 to 1.5 percent Rape Victimsb Physical Assault Victims for men. Based on these higher preva- Type of Medical Care Womenc Womend Mene lence rates, approximately 12.1 mil- Emergency room 3.2 1.7 1.4 lion women and 3.7 million men are Outpatient 2.2 5.0 2.8 stalked at some time in their life, and Overnight in hospital 3.6 8.5 13.3 an estimated 6 million women and Physician visit 4.8 3.3 7.2 1.4 million men are stalked annually. Dental visit 5.0 3.8 5.7 Stalking prevalence estimates gener- Ambulance/paramedic visit 1.2 1.1 1.0 ated by the NVAW Survey are higher Physical therapy visit 13.0 18.5 10.8 than previous nonscientific stalking a. Estimates are based on the most recent victimization since the age of 18. estimates. In 1992, when asked by a b. The number of men rape victims was insufficient to reliably calculate medical utilization estimates. reporter to estimate the extent of stalk- c. The standard error of the mean for each estimate in this column is 1.539, 0.433, 1.049, 0.889, ing in America, forensic psychiatrist 1.995, 0.145, and 4.273, respectively. Park Dietz guessed that 5 percent of d. The standard error of the mean for each estimate in this column is 0.2, 2.011, 3.365, 0.454, 1.016, U.S. women are stalked at some time 0.07, and 7.61, respectively. in their lifetime and 200,000 U.S. e. The standard error of the mean for each estimate in this column is 0.132, 0.789, 3.0, 1.252, 1.6, 24 0.023, and 2.475, respectively. women are stalked annually. The NVAW Survey estimate that 8 percent Note: Estimates are based on responses from victims who received the specific type of care considered.

10 R e s e a r c h i n B r i e f of women have been stalked at some 2. Rape should be viewed as a at some time in their lives. Of these, time in their life is nearly two times crime committed primarily against more than half (54 percent) were raped greater than Dietz’s earlier guess, and youth. The survey confirms previous before the age of 18. The survey also the survey’s estimate that 1 million reports that most rape victims are chil- found that women who are raped women are stalked annually is five dren or adolescents. Eighteen percent before the age of 18 are significantly times greater than Dietz’s earlier guess. of surveyed women reported being the more likely to be raped as adults. Given Given the high rate of stalking victim- victim of a completed or attempted rape these findings, rape prevention strategies ization uncovered by the NVAW Sur- vey, it is imperative that stalking be Exhibit 14: Average Annual Injury and Care Utilization Estimates for treated as a legitimate criminal justice Adult Rape and Physical Assault Victims by Sex of Victim problem and public concern. Estimated Number of Victimizations per Year Policy implications Rape Victimsa Physical Assault Victims The NVAW Survey provides compel- Women Women Men ling data on the prevalence, incidence, Victimization 876,064 5,931,053 7,883,580 and physical consequences of violence Victimization with injury 275,960 2,313,111 1,970,895 against women, as well as victims’ Victimization with injury receiving utilization of medical services. Infor- medical treatment 98,242 693,933 729,231 mation presented in this report can help inform policy and intervention Victimization treated medically with: directed at violence against women. Hospital care 80,460 527,389 627,139 Based on findings from the survey, Physician care 53,837 367,784 306,277 the authors conclude the following: Dental care 16,603 69,393 80,215 Ambulance/paramedic care 18,666 124,908 175,015 1. Violence against women should Physical therapy 16,406 62,454 94,800 be treated as a significant social Victimization treated in the hospital with: problem. The survey data validate Emergency room care 40,230 321,707 420,183 opinions held by many professionals Outpatient care 29,609 116,026 144,242 about the pervasiveness of violence Inpatient care 10,621 79,108 62,714 against women. More than half the surveyed women reported being physi- Visit/Overnight cally assaulted as a child by an adult Emergency room visit 128,736 546,902 588,256 caretaker and/or as an adult by another Outpatient visit 65,140 580,130 403,878 adult, and nearly one-fifth reported Overnight in hospital 144,828 672,418 834,096 being raped at some time in their lives. Physician visit 258,418 1,213,687 2,205,194 Further, 2 percent of the surveyed Dental visit 83,015 263,694 457,226 women reported being raped, physically Ambulance/paramedic visit 22,399 137,399 175,015 assaulted, or both in the 12 months Physical therapy visit 213,278 1,155,399 1,023,840 preceding the survey. This equates to a. The number of men rape victims was insufficient to reliably calculate medical utilization estimates. an estimated 2.1 million U.S. women who are raped and/or physically as- saulted annually. Because some rape Exhibit 15: Persons Stalked in Lifetime and in Previous 12 Months by and physical assault victims experience Sex of Victim multiple victimizations per year, an es- timated 876,000 rapes and 5.9 million Percentage Numbera physical assaults are perpetrated against Women Men Women Men U.S. women annually (see exhibit 14). (n=8,000) (n=8,000) (100,697,000) (92,748,000) Given the pervasiveness of rape and Stalked in lifetimeb 8.1 2.2 8,156,460 2,040,460 physical assault among American women, Stalked in previous 12 monthsb 1.0 0.4 1,006,970 370,990 it is imperative that violence against women be treated as a major criminal a. Based on estimates of men and women in the United States aged 18 years and older, justice and public health concern. U.S. Bureau of the Census, Current Population Survey, 1995. b. Differences between women and men are statistically significant: p-value ≤ .001.

11 R e s e a r c h i n B r i e f should focus on rapes perpetrated 4. Women are at greater risk of sustains an injury. Further, in about against minors, and rape research partner violence than men. The one-third of all such injury victimiza- should focus on the long-term effects survey found that women were signifi- tions, the victim receives some type of rape occurring at an early age. cantly more likely than men to report of medical care (e.g., paramedic care, being raped and physically assaulted emergency room treatment, dental 3. Studies are needed to deter- by a current or former intimate partner, care, or physical therapy). Thus, of mine why the prevalence of rape whether the timeframe considered was the estimated 6.8 million rapes and and physical assault varies signifi- the person’s lifetime or the 12 months physical assaults perpetrated against cantly among women of different preceding the survey. Moreover, women U.S. women annually, 2.6 million will racial and ethnic backgrounds. who were raped or physically assaulted result in an injury to the victim and The survey found that American by a current or former intimate partner 792,200 will result in the victim Indian/Alaska Native women were were significantly more likely to sus- receiving some type of medical care. most likely to report rape and physical tain injuries than men who were raped Because many rape and physical assault victimization, while Asian/Pa- or physically assaulted by a current or assault victims receive multiple forms cific Islander women were least likely former intimate partner. Given these of care for the same injury victimiza- to report rape and physical assault vic- findings, intimate partner violence tion, medical personnel in the United timization. The survey also found that should be considered first and foremost States treat literally millions of rape Hispanic women were less likely to a crime against women. and physical assault victimizations report rape victimization than non- annually. Hispanic women. 5. Violence against women is predominantly partner violence. Given the high number of injury vic- It is unclear from the survey data Data from the survey confirm previous timizations perpetrated against women whether the differences in prevalence reports that violence against women annually and the extensiveness of rates among women of different is predominantly intimate partner medical treatment to female victims of racial and ethnic groups are caused by violence. Of the women who reported violence, it is imperative that medical differences in reporting practices or being raped and/or physically assaulted professionals receive information about differences in actual victimization since the age of 18, three-quarters (76 the physical consequences of violence experiences. It is also unclear how percent) were victimized by a current against women and the medical needs social, environmental, and demographic or former husband, cohabiting partner, of female victims. factors intersect with race and ethnicity date, or boyfriend. Given these find- to produce differences in rape and ings, violence-against-women interven- 7. Stalking should be treated as a physical assault prevalence among tion strategies should focus on the risks significant social problem. Using women of different racial and ethnic posed to women by current and former a definition of stalking that requires backgrounds. Thus more research is husbands, cohabiting partners, dates, victims to feel a high level of fear, needed to establish the degree of vari- and boyfriends. the survey found that stalking is more ance in prevalence among women of widespread than previously thought. different racial and ethnic groups and 6. America’s medical community Eight percent of women and 2 percent to determine how much of the variance should receive comprehensive of men reported being stalked at some may be explained by differences in training about the medical needs time in their lives, and 1 percent of willingness to disclose information and of female victims of crime. The women and 0.4 percent of men re- how much by social, environmental, injury and medical utilization data ported being stalked in the 12 months and demographic factors. Research is generated by the NVAW Survey pro- preceding the survey. Thus an esti- also needed to determine whether vide compelling evidence of the physi- mated 1.4 million people are stalked differences exist in rape and physical cal and social costs associated with annually in the United States. Given assault prevalence among women of violence against women. the high number of stalking victims, diverse Asian/Pacific Islander groups, it is important that stalking be treated American Indian tribes, and Alaska The survey found that in about one- as a legitimate criminal justice and Native communities. third of all rapes and physical assaults public health concern. perpetrated against women, the victim

12 R e s e a r c h i n B r i e f

Survey Screening Questions R ape was defined as an event that they experienced as an adult at the hands of — Sent you unsolicited letters or occurred without the victim’s , another adult: written correspondence? that involved the use or threat of force — Made unsolicited phone calls to to penetrate the victim’s or anus ● [Physical assault as a child] When you you? by penis, tongue, fingers, or object, or were a child did any parent, step-parent, or the victim’s mouth by penis. The defini- guardian ever . . . — Stood outside your home, school, tion included both attempted and com- or workplace? ● [Physical assault as an adult] Not counting pleted rape. The following questions any incidents you have already mentioned, — Showed up at places you were were used to screen respondents for after you became an adult did any other even though he or she had no rape victimization:1 adult, male or female, ever . . . business being there? ● [Female respondents only] Has a — Throw something at you that could — Left unwanted items for you to or boy ever made you have sex by using hurt? find? force or threatening to harm you or someone close to you? Just so there is — Push, grab, or shove you? — Tried to communicate in other no mistake, by sex we mean putting a ways against your will? — Pull your hair? penis in your vagina. — Vandalized your property or — Slap or hit you? destroyed something you loved? ● Has anyone, male or female, ever — Kick or bite you? made you have by using force or Respondents who answered yes to one threat of force? Just so there is no mis- — Choke or to drown you? or more of these questions were asked take, by oral sex we mean that a man or whether anyone had ever done any of — Hit you with some object? boy put his penis in your mouth or some- these things to them on more than one one, male or female, penetrated your — Beat you up? occasion and whether they felt fright- vagina or anus with their mouth. ened or feared bodily harm as a result of — Threaten you with a gun? these behaviors. Only respondents who ● Has anyone ever made you have anal — Threaten you with a knife or other reported being victimized on more than sex by using force or threat of harm? weapon? one occasion and who were very fright- Just so there is no mistake, by anal sex ened or feared bodily harm were we mean that a man or boy put his penis — Use a gun on you? counted as stalking victims. in your anus. — Use a knife or other weapon on you? Notes ● Has anyone, male or female, ever put Stalking was defined as a course of conduct 1. Rape screening questions were fingers or objects in your vagina or anus directed at a specific person that involves adapted from those used in The against your will or by using force or repeated visual or physical proximity; non- National Women’s Study, see Rape threats? consensual communication; verbal, written, in America: A Report to the Nation, or implied threats; or a combination thereof ● Has anyone, male or female, ever National Victim Center and the Crime that would cause fear in a attempted to make you have vaginal, Victims Research and Treatment Cen- (with repeated meaning on two or more oral, or anal sex against your will, but in- ter, 211 Wilson Boulevard, Suite 300, occasions). The survey used the following tercourse or penetration did not occur? Arlington, Virginia 22201, April 23, questions to screen for stalking 1992, p.15. Physical assault was defined as behav- victimization: iors that threaten, attempt, or actually 2. Straus, Murray, “Measuring ● Not including bill collectors, telephone inflict physical harm. A modified version Intrafamily Conflict and Violence: The solicitors, or other salespeople, has anyone, of the Conflict Tactics Scale2 was used to Conflict Tactics (CT) Scale,” Journal of male or female, ever . . . screen respondents for physical assault and the Family 41 (February 1979): 75–88. they experienced as a child at the hands — Followed or spied on you? of an adult caretaker and physical assault

13 R e s e a r c h i n B r i e f

Survey Methodology T he National Violence Against interviews were conducted by female inter- weighting was considered to correct for Women (NVAW) Survey was conducted viewers and half by male interviewers. possible introduced by the fact that from November 1995 to May 1996 by some households had multiple phone lines interviewers at Schulman, Ronca, and A Spanish-language translation was adminis- and multiple eligibles, and for over- and Bucuvalas, Inc. (SRBI), under the direction tered by bilingual interviewers for Spanish- under-representation of selected sub- of John Boyle.1 The authors of this report speaking respondents. groups. Although there were some in- designed the survey and conducted the stances of over- and under-representation, To determine the representativeness of the analysis. the overall unweighted prevalence rates sample, select demographic characteristics of for rape, physical assault, and stalking the NVAW Survey sample were compared The national sample was drawn by were not significantly different from their with demographic characteristics of the gen- random-digit dialing from households with respective weighted rates. As a result, eral population as measured by the U.S. a telephone in all 50 States and the District sample weighting was not used in the Census Bureau’s 1995 Current Population of Columbia. The sample was adminis- analysis of the NVAW Survey data. tered by U.S. Census region. Within each Survey of adult men and women. Sample region, a simple random sample of work- ing residential “hundreds banks” of phone Exhibit 16: Estimated Standard Errors Multiplied by the z-Score (1.96) for numbers was drawn. (A hundreds bank is a 95-Percent Confidence Level by Sample or Subsample Size the first eight digits of any 10-digit tele- Percentage of the Sample or Subsample Giving a Certain phone number; e.g., 301–608–38xx.) A Response or Displaying a Certain Characteristic for randomly generated two-digit number Percentages Exactly or Approximately Equal to: was appended to each randomly sampled hundreds bank to produce the full 10- Size of Sample digit, random-digit number. Separate or Subsample 10 or 90 20 or 80 30 or 70 40 or 60 50/50 banks of numbers were generated for 16,000 0.5 0.6 0.7 0.8 0.8 male and female respondents. These 12,000 0.6 0.7 0.8 0.9 0.9 random-digit numbers were called by SRBI 8,000 0.7 0.9 1.0 1.1 1.1 interviewers from their central telephone 4,000 0.9 1.2 1.4 1.5 1.5 facility, where nonworking and nonresi- 3,000 1.1 1.4 1.6 1.8 1.8 dential numbers were screened out. Once 2,000 1.3 1.8 2.0 2.1 2.2 a residential household was reached, 1,500 1.5 2.0 2.3 2.5 2.5 eligible adults (i.e., women and men 18 1,300 1.6 2.2 2.5 2.7 2.7 years of age and older) in each household 1,200 1.7 2.3 2.6 2.8 2.8 were identified. In households with more 1,100 1.8 2.4 2.7 2.9 3.0 than one eligible adult, the adult with the 1,000 1.9 2.5 2.8 3.0 3.1 most recent birthday was selected as the 900 2.0 2.6 3.0 3.2 3.3 designated respondent. 800 2.1 2.8 3.2 3.4 3.5 700 2.2 3.0 3.4 3.6 3.7 A total of 8,000 women and 8,005 men 600 2.4 3.2 3.7 3.9 4.0 18 years and older were interviewed 500 2.6 3.5 4.0 4.3 4.4 using a computer-assisted telephone 400 2.9 3.9 4.5 4.8 4.9 interviewing system. (Five completed 300 3.4 4.5 5.2 5.6 5.7 interviews with men were subsequently 200 4.2 5.6 6.4 6.8 6.9 eliminated from the sample during data 150 4.8 6.4 7.4 7.9 8.0 editing due to an excessive amount of 100 5.9 7.9 9.0 9.7 9.8 incongruous data.) Only female interview- 75 6.8 9.1 10.4 11.2 11.4 ers surveyed female respondents. For male 50 8.4 11.2 12.8 13.7 14.0 respondents, approximately half of the

14 R e s e a r c h i n B r i e f

A technical report describing the survey sample and subsample sizes (n’s) vary time and annual prevalence rates are methods in more detail and recording from table to table. based) is 1.1 percentage points if the re- sample characteristics and prevalence rates sponse distribution is a 50/50 split, while using weighted and unweighted data is The estimates from this survey, as from any the standard error at the 95-percent confi- available from the authors.2 sample survey, are subject to random sam- dence level for a sample or subsample of pling error. Exhibit 16 presents the estimated 50 is 14 percentage points if the response Data were analyzed using SPSS Base 7.0 standard errors multiplied by the z-score distribution is a 50/50 split. for Windows software. Measures of asso- (1.96) for specified sample and subsample ciation (e.g., Lambda) were calculated sizes of 16,000 or less at different response Notes between nominal-level independent and distributions of dichotomous variables (e.g., 1. John Boyle, Ph.D., is senior vice president dependent variables, and the chi-square raped/not raped, injured/not injured). These and director of the Government and Social statistic was used to test for statistically estimated standard errors can be used to de- Research Division at SRBI. Dr. Boyle, who significant differences between men and termine the extent to which sample estimates specializes in public policy research in the ≤ women (p-value .05). Any estimates will be distributed around the population pa- area of health and violence, also manages based on five or fewer responses were rameter (i.e., the true population distribution). the firm’s Washington, D.C., area office. deemed unreliable and therefore were not As exhibit 16 shows, larger sample and tested for statistically significant differences subsample sizes produce smaller estimated 2. To obtain copies of the National Violence between groups and not presented in the standard errors. Thus the estimated standard Against Women Survey Methodology Re- tables. Because estimates presented in this error at the 95-percent confidence level for a port, call or write to the Center for Policy report generally exclude “don’t know,” sample or subsample of 8,000 (the number of Research, 1570 Emerson Street, Denver, “refused,” and other invalid responses, women and men, respectively, on which life- Colorado, 80218; 303–837–1555.

Notes 7. Craven, Diane, Sex Differences in Violent 13. Bachman, Ronet, Death and Violence on the Victimization, 1994, Special Report, Washing- Reservation: Homicide, Family Violence, and 1. Kennedy, Leslie W., in Foreword to Danger- ton, D.C.: U.S. Department of Justice, Bureau in American Indian Populations, ous Domains: Violence Against Women in of Justice Statistics, September 1997, NCJ Westport, CT.: Auburn , 1992. by Holly Johnson, Scarborough, 164508. Ontario: International Thomas Publishing, 1996. 14. Wallace, L.J.D., A.D. Calhoun, K.E. Powell, 8. The two rape/sexual assault screening ques- J. O’Neill, and S.P. James, Homicide and Suicide 2. Wilson, Carolyn F., Violence Against Women: tions used in the NCVS are as follows: (1) (Other Among Native Americans 1979–1992, Violence An Annotated Bibliography, Boston: G.K. Hall & than any incidents already mentioned,) has any- Surveillance Summary Series, No. 2, Atlanta, Co., 1981. one attacked or threatened you in any of these GA: Centers for Disease Control and Prevention, ways: . . . (e) Any rape, attempted rape, or other National Center for Injury Prevention and 3. National Research Council, Understanding type of sexual attack? (2) (Other than any inci- Control, 1996. Violence Against Women, Washington, D.C.: dents already mentioned,) have you been forced National Academy Press, 1996: 40–44. or coerced to engage in unwanted sexual activity 15. Tjaden, Patricia, and Nancy Thoennes, by (a) someone you didn’t know before, (b) a ca- Stalking in America: Findings From the 4. Ibid. sual acquaintance, or, (c) someone you know well? National Violence Against Women Survey, Research in Brief, Washington, D.C.: U.S. 5. Michael, Robert T., John H. Gagnon, Edward 9. See, for example, Helton, A.M., “The Preg- Department of Justice, National Institute of O. Lauman, and Gina Kolata, Sex in America: nant Battered Women,” Responses to Victimiza- Justice, April 1998, NCJ 169592. A Definitive Survey, New York: Warner Books, tion of Women and Children 9. (1) 1986: 22–23; 1994: 223. Koss, Mary P., “Detecting the Scope of Rape: A 16. Ho, C.K., “An Analysis of Domestic Review of Prevalence Research Methods,” Jour- Violence in Asian American Communities: 6. The NVAW Survey was conducted from nal of Interpersonal Violence 8 (2) (June 1993): A Multicultural Approach to Counseling,” in November 1995 to May 1996 (see “Survey 198–222. and Complexity in , Methodology,” page 14). Respondents reported ed. L.S. Brown and M. Root, New York: on events that spanned the 12 months prior to 10. See “Survey Screening Questions,” page 13. Harrington Park, 1996: 129–150; Huisman, their interview. For example, a person who was K.A., “Wife Battering in Asian American interviewed in November 1995 reported on 11. Straus, Murray, “Measuring Intrafamily Con- Communities,” Violence Against Women 2 events that occurred between November 1994 flict and Violence: The Conflict Tactics (CT) (1996): 260–283. and November 1995; a person who was inter- Scale,” Journal of Marriage and the Family 41 viewed in May 1996 reported on events that (February 1979): 75–88. 17. Sorenson, S.B., J.A. Stein, J.M. Siegel, J.M. occurred between May 1995 and May 1996. Golding, and M.A. Burnam, “The Prevalence of 12. Craven, Sex Differences in Violent Adult Sexual Assault: The Los Angeles Epide- Victimization, 1994. miologic Catchment Area Project,” American

15 R e s e a r c h i n B r i e f

Journal of 126: 154–164, 1987; 20. See, for example, Morse, Barbara, “Beyond ington, D.C.: U.S. Department of Justice, Bureau Sorenson, S.B., and C.A. Tells, “Self-Reports of the : Assessing of Justice Statistics, April 1984, NCJ 093449. Spousal Violence in a Mexican American and a Differences in Partner Violence,” Violence and Non-Hispanic White Population,” 10 (Winter 1995): 251–272; Straus, 22. See, for example, Langan, Patrick A., and Victims 6 (1991): 3–16. Murray, and Richard Gelles, “Societal Change Christopher A. Innes, Preventing and Change in Family Violence From 1975 to Against Women, Special Report, Washington, D.C.: 18. See National Victim Center and the Crime 1985 as Revealed by Two National Studies,” U.S. Department of Justice, Bureau of Justice Sta- Victims Research and Treatment Center, Rape Journal of Marriage and the Family 48 tistics, 1986, NCJ 102037; Timrots, Anita D., and in America: A Report to the Nation, 211 Wilson (August 1986): 465–479. Michael R. Rand, by Strangers and Boulevard, Suite 300, Arlington, Virginia Nonstrangers, Special Report, Washington, D.C.: 22201, April 23, 1992. 21. See, for example, Bachman, Ronet, Violence U.S. Department of Justice, Bureau of Justice Against Women: A National Crime Victimization Statistics, January 1987, NCJ 103702. 19. See, for example, Browne, Angela, and Survey Report, Washington, D.C.: U.S. Depart- David Finkelhor, “Initial and Long-term ment of Justice, Bureau of Justice Statistics, NCJ 23. Rand, Michael, Violence-Related Injuries Effects: A Review of the Research,” in 145325; Bachman, Ronet, and , Treated in Hospital Emergency Departments, Sourcebook on Child , ed. David Violence Against Women: Estimates From the Special Report, Washington, D.C.: U.S. Depart- Finkelhor, Beverly Hills: Sage Publications, Redesigned Survey, Special Report, Washington, ment of Justice, Bureau of Justice Statistics, 1986: 158; Miller, J., D. Moeller, A. Kaufman, D.C.: U.S. Department of Justice, Bureau of August 1997, NCJ 156921. P. Divasto, P. Fitzsimmons, D. Pather, and Justice Statistics, August 1995, NCJ 154348; J. Christy, “Recidivism among Sexual Assault Gaquin, Diedre, “Spouse Abuse: Data From the 24. See Puente, Maria, “Legislators Tackling Victims,” American Journal of 135 National Crime Survey,” 2 (1977– the of Stalking: But Some Experts Say (1978): 1103–1104; and Russel, D.E.H., The 1978): 634–643; and Klaus, Patsy, and Michael Measures Are Vague,” USA Today, July 21, Secret Trauma: in the Lives of and Rand, Family Violence, Special Report, Wash- 1992: A9. Women, New York: Basic Books, 1986.

Patricia Tjaden, Ph.D., and Nancy The authors thank Lois Mock at NIJ The National Institute of Justice is a component Thoennes, Ph.D., are with the Denver- and Linda Saltzman at the CDC for of the Office of Justice Programs, which also includes the Bureau of Justice Assistance, the based Center for Policy Research. their advice and support in completing Bureau of Justice Statistics, the Office of this project. The authors also thank This research was supported by grant Juvenile Justice and Delinquency Prevention, anonymous reviewers and Steven and the Office for Victims of Crime. number 93–IJ–CX–0012, awarded to Leadbetter and Marcie-jo Kresnow, the Center for Policy Research by the mathematical statisticians at the NCJ 172837 National Institute of Justice (NIJ) and CDC, who provided helpful comments sponsored jointly by NIJ and the on drafts of this report. Findings and conclusions of the research re- Centers for Disease Control and ported here are those of the authors and do not necessarily reflect the official position or poli- Prevention (CDC). cies of the U.S. Department of Justice or the U.S. Department of Health and Human Services.

U.S. Department of Justice Office of Justice Programs BULK RATE U.S. POSTAGE PAID National Institute of Justice DOJ/NIJ Permit No. G–91 Washington, DC 20531

Official Business Penalty for Private Use $300

16