|Violence against Women (World Bank, 1994, 84 pages)|
1. The Chilean study, for example, asked women whether they had experienced specific forms of behavior and then asked them to code the frequency of the behavior on a rivepoint ordinal scale ranging from I (all the time) to 5 (never). A woman was considered abused if her partner had ever used physical force against her beyond pushing, slapping, or throwing things-for example, if he had hit her with a fist (Larrain 1993).
2. These observations, published in an article in Lancet on the abuse of New Zealand women, are typical of the experience of most researchers: "In those reporting abuse during childhood, all recounted experiences involving genital contact, with most describing incidents of gross abuse. The great majority of those sexually assaulted in adult life had been the victims of rape or attempted rape. The physically abused group had all been subjected to unequivocal battery, with many suffering repeated injurious assaults. These women showed no tendencies to inflate minor incidents, but on the contrary, demonstrated a conservative approach in which only obvious assaults were labeled as abuse" (Mullen and others 1988).
3. These studies are all based on random probability samples of adult women interviewed in person or by telephone. All studies used legally grounded definitions of rape; thus, forms of penetration other than penile-vaginal are included, and women were not instructed to exclude rape by their husbands. Questions were typically framed to explicitly define the behaviors that should be included in the definition-for example: "Has a man made you have sex by using force or threatening to harm you? When we use the word sex we mean a man putting his penis in your vagina even if he didn't ejaculate (come)." This is followed by: "If he did not try to put his penis in your vagina, has a man made you do other sexual things like oral sex or anal sex, or put fingers or objects inside you by using force or threatening to harm you?"
4. In evaluating the sources of variability in prevalence of sexual abuse, Peters, Wyatt, and Fmicelhor(1986) suggest that differences in definitions and methods used in these studies probably account for most of the variations reported. The operational definitions of abuse varied according to whether they include contact or noncontact violations, or both, the nature of the violations, the age of the girl at the time of abuse, and the difference in age between the perpetrator and the victim. Most definitions include contact and noncontact abuse, include respondents 17 and under (or sometimes 15 and under), and use an age difference of five years (Peters, Wyatt, and Finkelhor 1986). When Wyatt's (1985) study, which reported a 62 percent rate of prevalence, was recomputed to exclude peer experiences (and thus to conform to the definition used by Russell and Finkelhor), the prevalence rate decreased to 53 percent, almost matching Russell's 54 percent finding (1983). Finkelhor's (1979) study, which found 19 percent prevalence, used a self-administered questionnaire, while the others used face-to-face interviews. The definition used in these studies includes all types of intrafamilial contact abuse. For abuse that occurs outside the family, it includes all types of contact abuse up to age 13, but only completed or attempted forcible rapes for ages 14 to 17.
5. The survey measured sexual abuse using a series of questions that asked about sexual contact between a child and his or her parent, stepparent, or parental partner, including breast or genital touching, or sexual intercourse that occurred before age 16 with an individual more than five years his or her senior. Also included was any forced sexual contact regardless of age. Respondents were asked to complete an anonymous survey questionnaire that was placed in a sealed envelope and dropped into a bag containing at least six other questionnaires.
6. It may be that rapes of young girls are reported more often, but in the United States at least, this is not the case: girls under 18 are less likely than older women to report their assault (U.S. Congress 1990).
7. A recent analyst of national demographic from China contends that underreporting of female births and an increase in prenatal sex selection through ultrasound and selective abortion can explain almost all the increase in the reported sex ratio at birth in China during the late 1980s, ruling out the possibility of widespread female infanticide (Zeng, Tu, and Gu 1993).
8. China manufactured its first ultrasound machine in 1979, and by the end of the 1980s it was making 10,000 and importing more than 2,000 each year. A Chinese demographer estimates that 100,000 ultrasound scanner were in place around the country by 1990. One study reported that all of the peasants interviewed-even those with no schooling and living in remote regions-knew that there was a device that could determine the sex of a fetus (Kristof 1993).
9. Sen originally estimated that more than 100 million women were missing. In a detailed recalculation, however, Princeton demographer Ansley Coale estimated that the true number of missing women was closer to 60 million in the 1980s (in countries including Bangladesh, China, Egypt, India, Nepal, and Pakistan, and in West Asia). According to Coale, Sen's estimate used an inappropriate comparison that failed to take into account the loss of male lives in the west due to war and an age composition that reflected past low fertility. Coale s revised estimate adjusts for these factors (Coale 1991).
10. ID the United States in 1980, 20 million people were an a "serious" diet at any one moment (Harvard Medical School Health Letter 1980). Dieters spent more than $10 billion that year on efforts to lose weight (Chernin 1981).
11. Post-traumatic stress disorder (PTSD) is a complex of symptoms sees in the aftermath of extreme trauma. It is characterized by intrusive memories, hyperarousal alternating with psychological numbing and withdrawal, and attempts to avoid any reminders of the trauma.
12. A recent study by the Addiction Research Foundation in Toronto found that battered women use sedatives 74 percent more and sleeping pills 40.5 percent more than women who are not battered (Groenveld and Shain 1989).
13. The World Health Organization defines maternal mortality as a death during pregnancy or within 42 days afterward, from causes related to or aggravated by the pregnancy or its management.
14. Although male approval is by no means always the greatest determinant of contraceptive use, see Gallen 1986 and Kincaid and others 1991 for examples of eases where it is.
15. Variables controlled for include years in legal or common-law union during previous five years, raised in lower-class home, education of mother, education of father, raised in stable nuclear family, raised solely by mother, raised with a stepfather, degree of affection mother's partner showed her, degree of physical and emotional abuse to mother, degree of affection mother showed son, degree of affection mother's partner showed son, degree to which mother's partner physically and emotionally abused son, man's educational status, and man's occupational status.
16. The Davis and Carlson (1987) study found that 90 percent of abused boys and 75 percent of those witnessing violence (compared with 13 percent for controls) had behavior problem scores greater than one standard deviation above the norm.
17. The women's police station in Argentina was eliminated in 1993, apparently due to lack of political support. But by the end of 1992 Colombia had So special stations? and Peru had added four new women's units in Lima (ISIS International 1993).
18. A study in Punta Arenas, Chile, that cross-checked hospital records with local police records found that of 736 cases of violence against women registered between 1988 and 1989 at the Regional Hospital and Clinica Magallanes, only 15.4 percent were also registered at the local police court These findings underscore the importance of using health facilities to identify and refer victims of violence (Ojeda and Perez 1992).
19. In addition, few violence prevention programs have been systematically evaluated. A survey sponsored by the Carnegie Council on Adolescent Development of 83 teen violence prevention programs in the United States found that, although they were geared toward teen violence, many focused on elementary school-age children or early adolescents. Of the 51 programs reviewed, only 21 percent evaluated outcomes, generally through simple measures of the pre- and post-test attitudes and knowledge of participants. Only a handful have been evaluated at a level approaching rigorous experimental design. But less formal evaluations indicate that some of the programs hold great promise (Wilson-Brewer and others 1991).
20. Between 1984 and 1989 arrests for minor assaults in the United States increased 70 percent, largely because of new "presumptive policies and mandatory arrest laws for domestic violence (Schmidt and Sherman 1993).
21. One potentially important finding is positive effect that issuing arrest warrants appears to have on batterers who left the scene before the police arrived. In a study in Omaha absent offenders who were randomly assigned to receive a warrant were less violent at follow-up than those who did not receive a warrant (Dunford, Huizinga, and Elliot 1990). Because more than half of all offenders leave the scene before the police arrive, this intervention may have more impact on rates of recidivism than any strategy targeting perpetrators who arc present at the scene (Schmidt and Sherman 1993).
22. Studies show that about a third of violent men can be expected to remain violence-free one to three years after a violent incident even without treatment or legal intervention (Rosenfeld 1992).