1. Introduction

Bangladesh is a developing country with accelerated population growth. Presently, in the world, the control of population explosion is one of the most burning issues particularly in developing countries. The population of Bangladesh has increased from about 42 million in 1941 (BBS, 1998) to 149.5 million in 2011 (BBS, 2011). There are serious consequences of this rapid population growth. The land-man ratio continues to decline considerably. Surprisingly, it is true that the above population in Bangladesh is very high compared to her land and resources. The above population age structure increases the dependency ratio and creates pressure on Maternal and Child Health (MCH) and nutrition services, pre-school care and different levels of educational facilities (Mittra & Kumar, 2004).

Previous studies in Bangladesh on the magnitude of physical violence against women were either focused on a particular socioeconomic status or the prevalence estimates did not relate exclusively to intimate partner violence. Thus, the study by Schuler et al. (1996) covered economically disadvantaged women in rural Bangladesh and addressed the prevalence of abuse of wives, whereas the study by Koenig et al. (2003) covered women from all the socioeconomic strata, but it explored factors associated with abuse of women by husband or in-laws, and the reference period for abuse was not explicit. The abuser tortures and controls the victim by calculated threats, intimidation and physical violence. Actual physical violence is often the end result of months or years of intimidation and control. Perhaps a better definition of domestic violence is emotional abuse, physical abuse or sexual abuse between people who have at some time had an intimate or family relationship (Newton, 2001).

Physical violence is one form of domestic abuse. Physical abuse is when an intimate partner has the following characteristics: pushed or shoved at her, slapped or bitten at her, kicked or choked at her, hit or punched at her, thrown objects at her, locked at her out of the house, abandoned at her in dangerous places, refused to help at her when she were sick, injured or pregnant, forced at her off the road or driven recklessly, threatened to hurt at her with a weapon (Newton, 2001). Physical violence is directly related to the duration of marriage. For example, 30% of women who have been married for less than five years report over having experienced physical violence, compared with 47% of women married 5-9 years and 54% of women married for more than 10 years (BDHS, 2007). In contrast, women are less likely to report sexual violence as the duration of marriage increases. Women are more likely to report physical violence if they have been married more than once or are currently separated, divorced or widowed. This is not surprising since violence is often a root cause of marital dissolution (Kishor and Johnson, 2004).

Violence against women is a worldwide problem, crossing cultural, geographic, religious, social and economic boundaries. It has become recognized as an important issue on the international arena and a lot of research work has been undertaken on this issue over the last couple of decades (Counts, Brown, & Campbell, 1999; Levinson, 1989; Straus & Gelles, 1999; Islam et al., 2012; Islam et al., 2013). None of these studies used questions that defined physical violence, thus leaving it up to the women to define. Less is known about the disclosure of violence and the help seeking of abused women in Bangladesh. Consequently, this paper focuses on the prevalence of physical violence against women and their disclosure to others or helps seeking.

2. Data Source

The data was collected from Bangladesh Demographic and Health Survey (BDHS) 2007. The BDHS-2007 survey was conducted under the authority of the National Institute for Population Research and Training (NIPORT) of the Ministry of Health and Family Welfare. The survey used the sampling frame provided by the list of census enumeration areas (EAs) with population and household information from the 2001 population census. From the data 10,146 respondents were interview and that only 4,190 respondent (yes=820 and no=3370) women are response to physical violence.

3. Methodology

In this study χ²- test and logistic regression are used to fulfill the aforementioned objectives. Cox (1958) is the pioneer of logistic regression model. This model was developed by Walker and Duncan (1967) and Cox (1970). It can be used not only to identify risk factors but also to predict the probability of success. This model expresses a qualitative dependent variable as a function of several independent variables, both qualitative and quantitative (Fox, 1984). In this analysis, physical violence is considered as dependent variable that is classified in the following way:

4. Results and Discussion

Background Characteristics and χ² Analysis:

The table shows that about 28.1% of the women are in the age group of 15-24 are more likely to report ever having experienced physical violence. Again, about 19.3% and 11.9% of the women are in age group 25-34 and 35+ respectively report violence. From the results it is clear that the prevalence of physical violence decreases due to increases of the respondent’s age. Education is one of the most important determinants of physical violence. Literacy rate, especially, among the females in Bangladesh is very low. In this study, it is observed that the majority of Bangladeshi women have never received any formal education and they reported that they are the mostly violated, which is for respondents 22%, compared with those who have some educational qualification. From the Table 1, it is also observed that the respondents who have completed secondary and higher education also have 16.2% physically violated.

Parity is also closely linked to the experience of physical violence. As the number of children increases, there is a notable chance of increase of physical violence because the increasing number of children needs more resources, more foods, more educational facilities, and more health facilities etc. which clearly contribute to increase the physical violence. Table 1 showed that the respondents who have 2 children they are the mostly violated, which are 22.2%, compared with those who have no children (18.5%). It is found that about 22.3% of the respondents who are involved to manual occupation (day labor, farmer, poultry firming, fisherman, agricultural worker, domestic servant, etc.) are mostly violated. On the other hand, the respondents who are involved to non-manual occupation (land owner, businessman, small business, teachers, doctors, engineers, nurses, etc.) and who are unemployed are 14.5% and 18.7% respectively. Wealth index is the important determinants of physical violence. The respondents who are poor, basically they are the mostly violated over the middle class and the rich. It is observed that about 25.4% of the respondents who are poor have experienced to physical violence compared with the respondents who are middle class, are 21.2%, and the respondents who are rich, are 14.0%.

Table 1 also revealed that about 21.2% of the respondents report they are violated by their intimate partners because they wants more children within 2 years, and about 18.1% of the women report that they have experienced physical violence because they wants no more children and about 21.9% of the respondents are experiencing violence by other causes. Physical violence is directly related to the duration of marriage. From the same table, it is seen that about 20.9% of the respondents married before the age of 18 years reported ever experiencing physical violence. These figures are about 10.2% in case of women who get married 18+ years. From the results, it is clear that prevalence of physical violence decreases due to increases of the age at first marriage.

Table 1. Background Characteristics of the Respondents Related to Physical Violence and their Association

As most of the people live in rural areas and early marriage is more common in rural areas than that of urban areas, so, physical violence is high in rural areas. It is seen that majority of the rural respondents are violated (20.1%), compared to urban respondents, which is about 18.6%. For administrative purpose, Bangladesh is divided into six divisions (Barisal, Chittagong, Dhaka, Khulna, Rajshahi and Sylhet). From the Table 1, it is showed that 24% of the respondents from Barisal division are violated domestically which is highest, followed by Chittagong, Khulna, Rajshahi, Dhaka and Sylhet division.

The results of association (χ²-value) of physical violence among the respondent’s age, respondent’s education, CEB, respondent’s occupation, wealth index, desire for more children, age at first marriage, religion, type of place of residence and divisions are being calculated and presented in Table 1 and this table showed that respondent’s age, respondent’s education, CEB, respondent’s occupation, wealth index, desire for more children, age at first marriage, religion, and divisions are significantly associated with the physical violence and only type of place of residence is insignificantly associated with the physical violence.

Factors Associated with Physical Violence: A Logistic Regression Analysis

According to the fitted model, there are six variables, out of eight variables, appear as the significant predictors of physical violence. In accordance with their importance, respondent’s age, respondent’s education, CEB, wealth index, religion and age at first marriage are statistically significant with the physical violence and respondent’s occupation and desire for more children are statistically insignificant.

The regression co-efficient of the respondent’s current age are being calculated. From the results, it is found that the respondent’s current age is the most significant factor affecting the physical violence. From Table 2, the odds ratio corresponding to the age group 25-34 and 35+ are 0.541 and 0.300 respectively. It clearly indicates that the respondent’s having current ages 25-34 and 35+ years have 0.541 (45.9%) and 0.300 (70%) times lower probability in affecting from the physical violence as compared with the respondent who are aged 15-24 years (reference category).

Respondent’s educational level has also come out as an important factor affecting the physical violence. The prevalence of physical violence is higher among the women who have no educational qualification. A small and statistically significant difference in physical violence by educational level is also observed after controlling the other variables. There is no large variation in physical violence between primary and secondary or higher level educated respondents. The number of living children has also been appeared as an influential factor for physical violence. Women who have 3+ and 2 children’s have experienced higher domestic violence than those who have no children. From Table 2, it is observed that the odds ratio corresponding to the women who have 3+ and 2 children are 1.526 and 1.492 respectively. It indicates that the respondent’s having 3+ and 2 children have 1.526 and 1.492 times higher experiencing to physical violence than those women who have no children.

Furthermore, it is found that respondent’s wealth index is the most significant factor affecting the physical violence. Respondent’s, who are poor, have higher experienced to physical violence as compared with the middle class and the rich. It has already been examined in bivariate analysis. From Table 2, it is seen that the respondents who are in the middle class and the rich have 0.899 (10.1%) and 0.699 (30.1%) times lower experience to physical violence as compared with that of the reference category. Religion is another important factor affecting the physical violence. From Table 2, it is observed that respondents who are Non-Muslim have 0.614 (38.6%) times lower experience to physical violence as compared with that of the reference category.

Physical violence is directly related to the age at first marriage. From Table 2, it is also seen that respondents who married within 18 years and above have 0.622 (37.8%) times lower experience to physical violence than that of the respondents who married before 18 years of age. The prevalence of physical violence decreases due to increases of the duration of marriage. As the age at marriage increases, the respondents become more matured and have higher responsibility to the family members. The respondents also have better understanding to her intimate partners. So, they have lower experienced to physical violence than that of the respondents who get married early.

Table 2. Results of Logistic Regression Analysis for Physical Violence

5. Conclusions

Respondent’s age, education, age at first marriage, wealth index and religion are the key determinants of physical violence. From the study, it is showed that 28.1% of the women are in the age group 15-24 are more likely to report physically violated. Again only 11.9% of the women are in the age group 35+ years have lower experiencing to physical violence, about 22% of the respondents are physically violated because of their illiteracy. This number is lower in those women who have completed secondary and higher education (16.2%), about 20.9% of the respondents married before the age of 18 years and reported they are the mostly physically violated. These figures are about 10.2% in case of women who get married 18 years and above and about 25.4% of the respondents who are poor have experienced to physical violence compared with the respondents who are middle class, are 21.2% and the rich, are 14%. From the logistic regression analysis, it is clear that the respondents having current ages 25-34 and 35+ years have about 45.9% and about 70% lower probability in affecting from the physical violence as compared with those respondents who are aged 15-24 years. The respondents who have primary and secondary+ education have 0.793 and 0.684 times lower experience to physical violence than that of no educational qualification, respondents who married after 18 years and above have 0.622 (37.8%) times lower experience to physical violence than that of the respondents who get married before 18 years also, the middle class respondents and the rich are about 10.1% and about 30.1% lower experience to physical violence as compared with poor class respondents. Religion is another important factor that affecting the physical violence.

Respondent’s education, children ever born, mass media and age at first marriage are statistically significant with the physical violence. In this situation, improve education, children ever born, increase age at first marriage and mass media are probably the most powerful tools through which these changes can be achieved. Finally, it is concluded that designed legal literacy and education programs will be undertaken to create public awareness about women legal rights.