2004, Number 2
Validez factorial de una escala de violencia hacia la pareja en una muestra nacional mexicana
Natera RG, Juárez GF, Tiburcio SM
Full text
Language: Spanish
References: 18
Page: 31-38
PDF size: 185.15 Kb.
ABSTRACT
Studies carried out all over the world have pointed marital violence as a main public health issue. Most of such studies also show that women have several disadvantages when compared to males, for instance, in the United States the risk of partner homicide is 8 times higher for females than for males, and the risk for being injured by an armed partner is 7 times higher.In Mexico, epidemiological research of family violence is still incipient. One of the first epidemiological studies was developed in the Southern area of Mexico City, the findings showed that around 38% of the women interviewed had been subjected to both physical and emotional violence at least once in their lifetime.
On the other hand, 37% of the women attending public health care institutions had suffered some form of violence at least once during their lives while 21.6% had been abused by their partner during the previous year. The group more likely to suffer violence was between 31 and 45 years and had low schooling. According to the Health Ministry the State with the highest prevalence of violence was Quintana Roo, and Aguascalientes had the lowest.
The study of violence and its different manifestations often implies a significant rate of under registration, a fact due to its hidden nature. Thus, it is necessary to develop valid instruments to periodically assess this phenomenon in a reliable way. Therefore, the objective of the present paper is to present the factor validity of a scale aimed at assessing violence towards women in general population.
Data were gathered in the Third National Survey on Addictions, which included an adapted version of the violence assessment scale. It was administered to urban women aged between 18 and 65 years, currently living with a partner, or who have once lived in couple. Different violent acts were explored as whether they occurred once in a lifetime or during the last year. The sample design was probabilistic, stratified, clustered and multi-stage, it was based upon the Basic Geostatistical Areas and the National Population Counting 1995. Interviewers were instructed regarding the specific requirements to hold an interview and to warrant the confidentiality of information. An exploratory factor analysis was performed with the Statistical Package for Social Sciences (SPSS) for Windows V.10 using principal components method and varimax rotation. The resulting factors were analyzed according to sociodemographic variables; differences were verified through X2 tests. The prevalence of both resulting factors was also obtained.
The twelve items of the scale were grouped into two factors that altogether explain 52% of the total variance. The first factor was called physical and verbal violence because it includes acts that involve beating, shouting at, and threatening women. The second factor was named threats of suicide and death. Finally, a confirmatory factor analyses was performed (EQS for Windows V.6). The internal consistency coefficient was 0.86 for the physical and verbal violence factor and 0.68 for threats of death and suicide.
The presence of physical and verbal violence was significantly higher among women with less schooling and lower income, separated or divorced, non-religious and who also reported domestic violence in their family of origin. Separated and divorced women, as well as women with family history of violence also had more threats of suicide and death.
The main contribution of the present paper is the scale used to assess situations where emotional and physical violence towards women is involved. It is important that these data contribute to policy making to protect children and facilitate special support to women and their aggressors.
REFERENCES