Bangladesh has the earliest patterns of marriage in the current developing world, high adolescent fertility and high rates of spousal violence against women and girls. Women and girls are often denied the right to choice or consent with respect to marriage, sex, contraceptive use and childbearing. However, the denial of these rights and its consequences, have not been adequately recognized and addressed in the context of the existing discourse or interventions in Bangladesh on sexual and reproductive health or violence against women. It is envisaged that the focus on realizing the right to choice and consent will bring about change in attitudes and behaviors that could not be achieved through a single intervention.

A multi-sectoral action research-based intervention is proposed, involving access to information about rights, available remedies and related referrals. Information will be coupled with access to legal and health services. This action research project aims to create a body of evidence. This multi-sectoral intervention in urban Bangladesh will highlight the critical element of expressing or refusing consent and choice, through a strong network between legal services, reproductive and sexual health service providers, human rights advocates, and research organizations.

This project proposes to implement an integrated intervention with both primary preventions and curative components based on the findings of the formative research. This project will document the program implementation and impact through detailed quantitative and qualitative evidence gathering, and carrying out an advocacy program to disseminate the results and bring about change. It includes:

• designing culturally sensitive intervention activities with a joined up approach.
• a strong community mobilization campaign for creating an enabling environment for women to live violence free lives.
• individual access to reproductive and sexual health services provided through health clinics and legal services, with legal clinics providing information, advice and support access to judicial remedies for redress in cases of serious violence.
• broader advocacy activities involving key stakeholders to reflect upon the findings and understandings geneated by the study and their relevance for the administration of justice delivery mechanisms.

Community engagement in the project will occur at multiple levels. At the local level the project will engage through community mobilisation and service delivery with adolescent girls, women and men living in urban slums of Dhaka city. It will also undertake targeted advocacy programmes with the frontlines of the justice delivery system relevant to these areas. Finally, it will engage at a national level with policymakers, researchers and key stakeholders in the justice system and health through its advocacy related interventions, drawing directly upon its findings.

Background Bangladesh has high rates of physical and sexual violence against women and girls. Women and girls in Bangladesh are often denied the right to make decisions regarding marriage, sex, and childbearing, yet neither the denial of these rights, nor their consequences, have been adequately recognized and addressed in the context of ongoing interventions to address violence against women. While higher court judgments, and key legal and policy provisions recognize women's rights to exercise choice or the requirements of obtaining their consent with respect to marriage, these are rarely applied or invoked in practice.

In Bangladesh, more than 65 percent of girls are married before the age of 18. Child marriage is associated with demands for dowry, with amounts increasing with age. Additionally, women or girls' ability to exercise their right to consent or choice with regard to marriage, sex and sexuality, is routinely denied. Child marriage represents a set of increased reproductive health risks that are associated with limited knowledge and skills to negotiate adult roles and diminished status in the marital home for an adolescent girl. One major reason for the association with early marriage and negative reproductive health outcomes is that young age at marriage for girls is associated with larger age differences that reduce her power within the marital relationship. This may affect factors such as negotiating timing of births, choice of contraception or use of maternal and child health (MCH) services. Child marriage has highly detrimental sexual and reproductive health (SRH) consequences resulting in early childbirth, and adverse effects on child and maternal nutrition and also maternal morbidity and mortality. Child marriage and economic disempowerment may also be related to the high levels of intimate partner violence (IPV) observed in Bangladesh.

Gender norms and perceptions about male predominance, strength and superiority further exacerbate the problem. Evidence suggests that such presumptions of superiority and acceptance of male prerogative and dominance, and corollary acceptability of violence as a form of control, are associated with greater violence. The reason may be that in settings where there is high acceptance of gender based violence the social costs of violence to perpetrators is lower. There is also some research to show that in addition to such community level attitudes, individual attitudes among women and men independently contribute to higher risk of IPV. Thus, modifying individual attitudes and community norms related to gender based violence is an important basis for behavioural change.

In Bangladesh adolescent girls and women are denied SRHR through patriarchal institutions and long standing practices such as child marriage and force marriage, segregation of the sexes and economic exclusion that relegate women to low status. Inadequate SRHR information and services make men vulnerable to diseases. Gender education is normally geared towards women and girls; thus men miss opportunities to become gender sensitive and responsible. While Bangladesh has traditionally afforded women reasonably good access to family planning services, the context of low status of women means that women are unable to exercise their right to choose in decisions regarding marriage, sex and childbearing to the full extent of their rights and ability. This project will seek to enhance people's ability to exercise their SRHR by strategically expanding their knowledge and understanding of the notions of full and informed consent and choice.

Child marriage and IPV are common in both urban and rural areas, with women in poor urban areas are most at risk, yet most programs and projects aimed at empowering women remain in rural areas. Data from Dhaka slums compared to non-slum populations as well as rural data shows higher prevalence and acceptance of gender-based violence among women living in urban slums. Migrant girls and women, particularly those living in urban slums, lack even the most rudimentary social protection - thus being particularly vulnerable to violence.

Bangladesh has clear obligations under international law and the Constitution to ensure equal treatment under the law to all individuals and to ensure freedom from violence as well as effective remedies. An overarching framework of fundamental rights -- to liberty, personal security equality and freedom of expression -- applies in the form of constitutionally guarantees to all persons - women and men - within Bangladesh. However, the content and application of specific laws often operate to deny or constrain the enjoyment of such rights in practice. This is due in large measure to the limits of the law, and inherent paradoxes, resulting from the continued prevalence of religion-based personal laws, which contradict and constrain the ambit of operation of constitutional rights. The Constitution clearly prohibits discrimination based on sex, and exhorts the state to ensure equal rights to men and women in the state and public sphere, as well as guaranteeing the fundamental rights to personal liberty, freedom of expression and freedom of religion. It also specifies that any laws which violate such fundamental rights will be void.

It is widely recognised that interventions should employ a multi-sectoral approach and work at different levels: individual, community, institutional, and laws and policies. Several reproductive health initiatives have shown that communication programs can be effective in improving knowledge and awareness. However, behavior change is more difficult to bring about. These operations research studies hypothesize the need to bring about attitudinal change in communities and among health care providers about the need to engage men in multiple ways to improve the communities' SRH.

Programs to delay marriage timing include community mobilization, behaviour change communication, educational incentives such as scholarships for girls and programs to promote lifeskills and livelihood skills. In Bangladesh a program entitled "Raising the Age of Marriage for Young Girls in Bangladesh" is one such project. A program in rural Bangladesh to delay marriage by providing adolescent life skills and livelihood skills showed increase in reproductive health knowledge, reduced school dropout and delayed marriage relative to girls in control groups.

The Government of Bangladesh's (GoB) has several initiatives on VAW aims to improve services for women survivors of violence and does so through targeting various different critical sets of actors, including health personnel, police. The Police Reform Programme promises to include a specific focus on addressing gender violence including training and collaboration with civil society actors. While several leading women's rights, human rights and legal service NGOs collaborate with the GoB programme, many others also run independent initiatives, either singly or in collaboration with others, aimed at ending VAW and at providing support and redress to survivors. This reluctance on the part of service providers to use available institutional processes reflects that of women survivors themselves. As research in Bangladesh has shown only 2% of the women physically assaulted by their husbands have ever sought help from any institutional source. Moreover, they approached these sources only when their situations became unendurable or the violence became life threatening or children were at risk. The findings show that while providing appropriate services is necessary, it is critical to make people aware of such services and to actively foster their use of such services by creating an enabling environment in the community. It is equally critical to make the services available within the justice sector more gender responsive.

Significance and Rationale Bangladesh is a signatory to international treaties to promote gender equality. Ending VAW is an integral part of that agenda. The primary focus of the current coalition is to reduce gender based violence beginning with the most vulnerable groups-poor and adolescent girls living in Dhaka city. The project is clearly framed within an understanding of the GoB's obligations to implement the provisions of key international human rights treaties such as the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW), in particular its understanding that the right to equality within the family (Article 16) includes the right to choice with regard to 'if, when and whom to marry', and its elaborate interpretation of the state's obligations to prevent VAW (Declaration on Violence against Women, as well as General Recommendation 19 of the CEDAW Committee).

The project is timed to seize a window of opportunity provided in part by recent national democratic elections - which have resulted in clear government commitments to: 1) address discriminatory laws and practices, including through taking up long-standing demands for legal reform (e.g., the recent Cabinet approval of the domestic violence prevention and protection bill); 2) strengthen existing institutional protections on VAW (victims support centres, legal aid programme); and 3) increased space and a more enabling environment for the operation of women's rights organisations and related service providers. The Project plans to engage directly with key policy makers and justice sector institutions to press for legal and policy reforms as necessary, and also to simultaneously reach out at the frontlines to press for more effective enforcement of existing laws and rules safeguarding women's rights to consent and choice. It will focus in part on enabling an informed understanding within key justice sector institutions at a local level of the issues relating to investigation and prosecution and prevention of VAW, as part of the roll out of the domestic violence law. The evidence yielded by the project will be critical for engaging with ongoing legislative advocacy initiatives on law and policy reform on women's rights.

Aims

This project aims to:

• Increase awareness among adolescent girls and women in urban slums regarding SRHR, and the right to freedom from violence, in particular the right to consent and choice

• Develop a community environment by engaging and working with men to address denial of the right to choice and consent of adolescent girls and women and violence against them are less condoned

• Enable access to and more effective use of legal and health services by adolescent girls and women in dealing with violence

• Reduce incidence of violence against adolescent girls and women due to the deterrent effect of increased awareness among them of their rights and remedies in this connection.

• Ensure law, policy and procedural reforms regarding access to justice, including with respect to developing protocols and practices for investigation and prosecution of VAW, procedures for registration of marriages which involve identification of the age of the woman and of whether her consent has been obtained, and recognition of age and the right to consent in court proceedings Research Design and Methods The proposed action research project is a mixture of research and programmatic intervention activities with different kinds of interventions targeted at different levels and research findings being fed back into the program.

  1. Awareness Raising: The main interventions will be awareness-raising in three domains, on the rights of women and girls: i) to live a violence-free life; ii) regarding SRH; and iii) legal provisions. In each area, a key focus would be on women's rights to exercise choice and consent. The awareness raising activities will be based on curriculum and materials developed jointly by project partners. These will draw heavily upon existing local and global resources, and be tailored to the needs of the given context.

     The awareness raising activities will seek to engage groups such as women (aged 10 to 29); men (aged 18-35) and community leaders from Dhaka slums. Initial activities will involve establishing adolescent and women's groups. Upon entry in the community, Marie Stopes will organize separate groups of women and men (about fifteen members each). About 450 women's groups and 150 men's groups will be formed. Regular sessions will be held with each group with varying frequency. Women will receive about five sessions per quarter, while men will receive one session per quarter. These sessions will focus on discussions on SRHR issues and life skills; right of a young woman to live a violence free life; and legal provisions for addressing violence including rights within the family, and the rights to consent and choice with regard to marriage, sex and childbirth.

     Advocacy in these groups on the question of consent will be an important area for enhancing women's and girls' capacity to combat and to seek redress for sexual abuse, underage marriage and forced marriage, and child bearing. Skilled trainers and paralegals will impart this training. As appropriate they will also provide referrals to health care providers, and lawyers for legal assistance and assist women who need to make complaints of violence to the police or to the courts and accompany them to support them through court hearings.

  2. One-Stop Service Centre (OSC): Awareness raising activities will be backed up by provision of health, legal and other support services to women survivors of violence from OSC. The services will be provided by Marie Stopes, BLAST, and Nari Maitree. The existing Marie Stopes clinics near the slums will be focal points for the dissemination of information and advice, service provision and referrals. Marie Stopes will provide SRH services while BLAST will provide legal services to the clients. Nari Maitree will make services available so that the community can take part in the campaign activities from these service points.

  3. Community Mobilisation and Community-wide Campaign: The above interventions will be supplemented by community mobilisation efforts by Nari Maitree, which will commence with the formation of community groups (e.g., local ward commissioners, leaders of youth clubs, teachers etc) as well as training and orientation targeted to them. Information and communication using relatively innovative means, such as rickshaw/wall paintings, mobile phone messages/ring tones, and mobile van campaign, etc will also be utilised as appropriate.The community-wide campaigns will be led by Nari Maitree, applying the model already developed by the We Can Campaign.

  4. Broader Advocacy: The intervention is partly linked to service delivery and community mobilisation, and partly free-standing, will involve more broadly targeted advocacy initiatives at a local and national level, which will engage all project partners on the recognition and realisation of women's right to consent and choice. Such advocacy will engage law makers and policy makers, and focus in particular on key actors within the justice system, i.e., police, prosecutors, social workers, health professionals and the judiciary.

Evidence gathering through collaborative action research The project has been designed to enable a more joined up approach to service delivery targeted to the most vulnerable for ensuring effective access to information and health services and justice to ensure that those needing advice and assistance are able to access appropriate services. The research component enable capacity building of the service delivery organizations to fine tune the existing services to better meet the needs of the people. The quantitative and qualitative evidence generated by the program will help to (i) identify vulnerable populations and the nature of their vulnerabilities to inform the intervention approach; (ii) enable designing a culturally sensitive appropriate set of initervention activities; (iii) monitor program inputs and document the activities in the interventions; (iv) give constant feedback into the program; (v) assess impact to understand what works and what does not work; and (vi) inform advocacy messages.

Thus, the project aims to ensure that the findings from on the ground interventions are fed into learning, which can be disseminated through professional and ongoing education for the relevant policy makers including judges, prosecutors and lawyers and police. This will ensure that the health education, services as well as justice systems are sensitised to the real needs of the poorest and most vulnerable users, and at the same time to use such information to develop focused and specific recommendations for policy and programmatic or procedural reforms.

Baseline and Scoping Study, and Impact Evaluation The research will be carried out through a scoping study and baseline survey in the first year, and an impact evaluation at the end of the third year. In particular, the baseline and scoping study will focus on the influence of prevalent norms and the potential for diffusion dynamics in changing normative structures and breaking rules. The impact evaluation will allow comparison between intervention and control areas in order to measure changes due to intervention.The study will be carried out using mixed methods of qualitative and quantitative data collection, using separate sample pre-test post-test control group design.

Qualitative data collection: Qualitative data will be collected for the scoping study and for evaluation of the intervention. Qualitative data will inform the design and questionnaire wording for the quantitative study, to interpret quantitative results. Tools for qualitative data collection will invovle 12 Key Informant Interviews (KII) (5 males and 7 females), 15 Focus Group Discussions (FGD) (3 with unmarried young women, 3 with married young women, 3 with unmarried young men, 3 married young men and 3 with community leaders) and 61 Indepth-Interviews (IDI) (20 with unmarried young women, 25 with married young women, 7 with unmarried young men and 9 with married young men). in the baseline study. Potential key informants are female and male NGO workers from the study site and women and men from the slums who have first-hand knowledge regarding the study population and are knowledgeable in the area this study is interested in exploring. The selection of informants for each kind of data collection will be opportunistic. In the endline study, there will be 16 KIIs (8 with females and 8 for males), 10 FGDs (2 with unmarried young women, 2 with married young women, 2 with unmarried young men, 2 married young men and 2 with both females and males irrespective of their marital status), 116 in-depth interviews (34 with unmarried young women, 34 with married young women, 24 with unmarried young men and 24 with married young men).

Quantitative data collection: The baseline and endline survey will be cross-sectional and will be conducted among 4,212 adolescent girls and women between the ages of 15 and 29; and 1,458 young and adult men ages between 18 and 35 living in the sample area at the time of the survey. Survey sample will include 15-19 years old unmarried and married girls; 20-29 married and unmarried females; and 18-35 years old married and unmarried males. Samples will be drawn separately for these three groups of respondents. The study will have 3-arms: (a) Strategy 1: group level awareness raising activities with females and males, plus a community wide campaign and one-stop service centres; (b) Strategy 2: group level awareness raising activities only with females plus a community wide campaign and one-stop service centres; and (c) Strategy 3: community wide campaigning activities coupled with one-stop service centres. Thus, there will be no "blank" or "no intervention" control condition. The design will allow comparison across male involvement versus no male involvement in strategy 1 versus strategy 2; and added advantage of reaching females on top of community campaigning and onestop service centres in comparison between strategy 2 and strategy 3.

Sample selection: The respondents will be selected from slums in the vicinity of three Marie Stopes clinics in Dhaka. Each of the three clinic site slums will be divided into 51, 27, and 27 non-contagious clusters of 15-19 year old females; 20-29 year old females and 18-35 years males, respectively. These clusters, then, will be randomly assigned to the three strategies. Male respondents will not be selected from the same household because their exposure to the survey may make the females vulnerable. There are different sizes of slums according to population and area, which may impact on violence towards women at different levels. Therefore, the study will incorporate both large and small slum areas in drawing sample according to size of the population.

Appropriately trained same sex interviewers will recruit respondents only after receiving consent or assent when appropriate. Before entering into the slums and households, permission will be taken from the gatekeepers, and household heads. The survey questionnaire draws heavily upon widely used standard research tools from sources such as the WHO multi-country study on Women's Health and Domestic Violence.

Study Area The study will be conducted in areas within urban slums of Dhaka city. The intervention will be carried out in communities surrounding three Marie Stopes clinics in Dhaka located in Mohammadpur, Mohakhali, and Jatrabari. According to Marie Stopes records there are 4 slums in Mohammadpur; 5 slums in Mohakhali and 10 slums in Jatrabari in the vicinity of these clinics. Marie Stopes estimated a total population of 51,514 in these slums. Applying the proportion of women aged 10 to 24 found in Dhaka slums by the Urban Health Survey we estimated about 8,336 women and girls to be residing in these slums.

Project Beneficiaries: The project will primarily affect some 6,750 slum adolescent girls and women, while the secondary group affected would be all the women (slum and non-slum) in the intervention area during the intervention period in Dhaka. For some of the interventions such as nation-wide media campaign or sensitization of key actors in the judiciary system a much broader section of Bangladeshi women will be the beneficiaries. In order to create an enabling environment for girls and women to practice their rights to freedom from violence, this project will also engage young men. Thus, in Dhaka 2,250 young men will be directly involved. These men will be encouraged to sensitize other members of the community, including the community leaders. Thus, the ultimate target becomes the entire community consisting of a total (slum and non-slum) population of around 128,785.

Sample Size Calculation and Outcome (Primary and Secondary) Variable(s). Sample sizes for this study are calculated assuming that treatment effects will be observed by measuring change in key outcomes between baseline and endline surveys and by comparing treatment and control groups. The sample also takes into account an extensive baseline survey that will allow the statistical analysis to control for known factors associated with the outcomes. Sample for the quantitative survey has been calculated using the three-level multi-site cluster randomized trial design. In this design, individuals are nested within clusters and the clusters are randomly assigned to the three strategies 1 or 2 or 3. Using the Optimal Design (OD) software, assuming 5% significance level, 80% power, intra-class correlation of 0.01, and cluster size of 15 respondents, 153 female groups of 15-19 years old (51 groups per site); 81 female groups of 20-29 years old (27 groups per site); and 81 male groups of 18-35 (27 groups per site) are needed to detect a minimum detectable effect size (MDES) of 45% to 55% (Table 3). Allowing for 20% oversampling to group's size in order to address 5% non-response and 15% migration at the endline, the cluster size increases to 18 for which the sample size increases from 4,725 to 5,670 (4,212 females and 1,458 males).

Ethical considerations There are number of ethical considerations that need to be made when conducting research on VAW in families. The CIOMS International Guidelines for Ethical Review of Epidemiological Studies (1991) and the WHO recommendations (2001) are being adhered to in the development of this Protocol.

Do no harm and respect adolescent girls' and women's decisions and choices Particular care will be taken to ensure that all questions about violence and its consequences are asked sensitively, in a supportive and non-judgmental manner. Interviewers and program staff will be trained to be aware of the effects that the questions may have on the informant and, if necessary, will terminate the interview. Care has been taken when designing the questionnaires and interview guides to try to carefully and sensitively introduce and enquire about adolescent girls' and women's experiences of violence. Each interview or session will aim to end in a positive manner to provide the participant with a positive outlook and reinforces her coping strategies.

Emotional support and skills Given the prevalence of VAW, many interviewers and program staff would either have experience of IPV themselves or have been affected by it in some way. For this reason, the training will be explicitly designed to provide an opportunity for them to reflect on their own experiences with abuse. At all times interviewers will be informed that they have the option of withdrawing from the project. Trainings will also include how to be empathetic and supportive, allowing the participant time to take a break and giving her the opportunity to reschedule or terminate the interview or discussion. At all interviews, participants will be informed orally of the purpose and nature of the study, why and how s/he has been selected, what is expected from the participant, privacy, anonymity and confidentiality, the expected benefits of the study, future use of the data, the principles of compensation and the right of the participants not to participate and to withdraw from the study if they so wish. Details for some of these areas of informed consent are given below Oral consent and assent Verbal consent will be taken on the following ground: 1) Request for signature on the consent form may arouse the respondent's suspicion; 2) The issue of maintaining confidentiality is further challenged when the respondent's signature is placed on the form; 3) It is expected that a major proportion of the respondents will be non-literate. In this situation asking for a thumbprint would further raise suspicion. In order to interview minors (aged below 18) consent from guardians would normally be sought. However, in a sensitive study such as this taking consent from potential perpetrators (fathers, brothers, husbands, in-laws) of abuse would further jeopardize the girls' and women's safety as well as the principle of confidentiality employed in the study. This is why oral consent from a guardian (if available) will be sought first introducing the study as a study of health and life experiences of young women. Then the full consent procedure will be followed with the young women in question. Assent form will be used to interview minors.

Voluntary participation Participation in the study will be on a voluntary basis. No inducements will be made. The participant will be free not to participate in the study; to terminate the interview at any point, and to skip any questions that s/he does not wish to respond to. As a token sign of appreciation SAFE one-stop services (legal and health) will be offered to all study participants; they will receive free treatment, free medicine that is available in the Marie Stopes clinics and travel allowance of TK. 60. This is applicable only for the first visit and at any time throughout the project period.

Confidentiality Much of the information provided by the participants will be extremely personal. The dynamic of a violent relationship is such that the act of revealing the painful details of abuse to someone outside the family nucleus could provoke another violent episode. As part of the consent procedure, the participant will be informed that the data collected will be held in strict confidence.

Mechanisms to attend to researchers' and field workers' needs The high prevalence of VAW worldwide means that, almost without exception, one or more project staff will have been a direct target, or have familial experiences of violence. While this may improve the interviewers' skills and empathy, the process of being involved in the study may awaken images, emotions, internal confusion and conflict. These reactions may affect their ability to work, may have a negative impact on their health, and may create tension in the home. During the fieldwork, regular meetings will be held with interviewers for emotional debriefing. In contrast to more technical meetings that focus on evaluating progress with data collection or program and other aspects of project logistics, these debriefing sessions will aim to provide staff with an opportunity to discuss their own feelings about the work.

Physical safety of informants and researchers

The physical safety of study participants and project staff from potential retaliatory violence by the abuser is of prime importance. If the focus of the project becomes widely known, this may place the participant or the staff member at risk of violence, either before, during or after the project. For this reason, the following measures will be adopted to ensure that the research topic does not become widely known:

• The survey will be framed as a survey of young women's health and life experiences, and will be introduced at the local and household level in this manner. The interviewers will carry a separate dummy questionnaire for use in situations.
• Interviews and discussion sessions will only be conducted in a private setting. Only children under 2 will be permitted to be present. If necessary, locations outside the household where the interview can be conducted in private will be identified.
• The participant will be free to reschedule (or relocate) the interview or session to a time (or place) that may be more safe or convenient for her.
• Interviewers will be trained to terminate or change the subject of discussion if an interview or discussion is interrupted by anyone. During the interview, the interviewer will forewarn the respondent that she will terminate or change the topic of conversation if the interview is interrupted, and will be able to skip to these questions at any point if needed.

Harmful publicity During dissemination of study findings care will be taken to highlight the extent to which violence against women is cross cutting, existing in all communities and socioeconomic groups. Particular attention will be paid to ensuring that the findings are not used as a means to describe one setting or group as being worse than another.

Inclusion Criteria:

• Female aged 10-29 years from the studied slums
• Male aged 18-35 years from the studied slums

Exclusion Criteria:

• Female aged below 10 and above 29 years
• Male aged below 18 and above 35 years
• Female and males who are living outside the studied slums