This study evaluates whether a factory wide intervention, HERrespect, can reduce female garment workers' experiences of intimate partner violence, and experience and/or witnessing of violence in the workplace. This is a quasi-experimental study involving four intervention and four control factories.

Background Violence against women (VAW) is an important public health, human rights and development issue worldwide. Intimate partner violence (IPV) is the most common type of VAW ranging between 15% and 71% worldwide. IPV in South Asian countries is as high as 42%. Adverse consequences of spousal violence on women's health and well-being are well documented. This violence is also known to affect women's work and productivity. IPV affects not only the women, but also their children, families and the nation.

In Bangladesh, 53% of the ever-married reproductive aged women reported lifetime physical and/or sexual abuse and about a quarter of the currently married women reported exposure to this violence during the last 12 months. Common correlates of IPV in Bangladesh include young age, poverty, low education, history of wife abuse in the family, childhood experience of violence, dowry, and low spousal communication. The literature on correlates of IPV shows that, particularly in patriarchal contexts, working women are more vulnerable to IPV. Although research on IPV has gained great momentum over the last few decades, literature on IPV against working women in low income settings is still scanty. Unpublished data from icddr,b show that 70% of the garment workers aged 15 to 29 were physically and/or sexually abused by their spouses during the last 12 months, while according to the Bangladesh component of the WHO multi-country study on Women's Health and Domestic Violence against 53% of urban and 62% of rural peers of them reported such violence. Studies show economic IPV against female garment workers in Bangladesh. Thus, young female garment workers are highly vulnerable to violence demanding attention of the researchers, programme implementers and policy makers.

Approximately 80% of the workforce in the garment manufacturing industry in Bangladesh is female. According to a report, about 60% of female workers experience harassment at work. Qualitative research suggests that severe emotional and economic violence is universal in this sector, while physical and sexual violence is not uncommon. Most common perpetrators of violence are mid-and low-level factory management staff.

In general, evidence on what works in addressing IPV and workplace violence against women is thin worldwide. A few sporadic interventions implemented for addressing workplace violence against female garment workers over the last few years in Bangladesh have not been rigorously studied. Although there is evidence that a combination of economic empowerment and gender interventions reduces IPV effectively there has been no attempt at measuring effect of gender interventions among female garment workers, who are relatively much more economically empowered compared to their peers. The current research proposes to evaluate HERrespect, a pilot intervention for preventing and responding to spousal and workplace violence against female garment workers by providing interactive sessions to the management and the workers on gender and violence against women and girls (VAWG).

Objectives of the study The primary objective of this study is to measure whether HERrespect intervention reduces female garment workers' experiences of: (1) Spousal physical and sexual violence and (2) Workplace violence over a period of 24 months between baseline and end-line during which the intervention is delivered in the factories.

The secondary objectives are to assess whether HERrespect:

1. Increases gender equitable attitudes among female garment workers
2. Reduces acceptance of VAWG among female garment workers
3. Increases self-esteem of the female garment workers
4. Reduces rate of depression among female garment workers
5. Increases knowledge and uptake of service for addressing spousal violence
6. Increases gender equitable attitudes among garment management staff
7. Improves management skills of management staff
8. Increases knowledge of policies addressing gender discrimination and violence among management staff

The intervention HERrespect is a workplace programme developed by Business for Socail Responsibilities (BSR) and South African Medical Research Council (SAMRC), with inputs from Change Associates Limited. Design of HERrespect is guided by the Theory of Change (TOC) and the formative research conducted by icddr,b.

Theory of Change The base of the TOC begins with a core problem that female workers in the RMG industry in Bangladesh experience high levels of violence, both in the workplace and their intimate relationships. Such violence is not only a violation to their rights, but at the same time constrain them from contributing to their personal growth, household, community and the economy at large.

The barriers are based on BSR's experiences in implementing HERproject, a women empowerment initiative in the supply chain, in Bangladesh, as well as the formative research conducted by icddr,b. At an individual level, there is a lack of critical awareness about gender and rights among workers, both male and female, and VAW is normalized and generally accepted in the workplace. Outside the workplace, workers also do not have sufficient information on available resources and initiatives in the community targeting IPV victims. Management does not possess essential skills to manage the workforce, and perceives violence as the most accessible and effective way to achieve production targets. The dominant social norms of femininity and masculinity in Bangladesh and the disciplinary nature of operating a factory reinforce the submissive identity of female workers and the unequal relationship between managers and workers. Moreover, there are institutional barriers in the workplaces, particularly on the gender-blind policies and organizational structure to prevent and address violence; the lack or misuse of such may even incentivize VAW in workplace.

To overcome the barriers, a combination of intervention strategies will be adopted to: (1) raise awareness on gender, rights and VAW among workers and management; (2) enhance the skills of workers to prevent and address sexual harassment and IPV; (3) build capacity of management to promote gender equitable environment; (4) create an enabling environment through development of gender-sensitive policy and organizational structure, and (5) bridge the information gap on community services and initiatives on IPV.

Together, the interventions will lead to five outputs. First of all, female workers will have enhanced awareness and knowledge on gender, rights and VAW; their skills will also be enhanced to respond to violence in workplace and family. In case they suffer from IPV, they would have more knowledge on the services available in the community and know where to avail them. For management and male workers, their awareness and skills will be enhanced to motivate and support colleagues in preventing and addressing VAW at workplace. Institutionally, the policies and structures of factory will be improved to address grievance in more equitable and effective manner, and management will be supported to design and implement policies to promote positive gender relations.

The outputs will lead to certain empowerment outcomes. Female workers will be empowered from within through improved self-esteem, confidence and assertiveness; communications and negotiations around IPV are also expected to improve through the training. The quality of interactions between management and workers would improve, contributing to a more supportive social network within the workplace. That will be further augmented by the improved implementation of policies and mechanism on VAW. Finally, with heightened awareness of community resources, it is expected that more women in need would avail services for VAW in the community.

The goal (or impact) of HERrespect is to cultivate more gender equitable attitudes and relationships among women and men in RMG industry in Bangladesh, which ultimately will contribute to preventing VAW at workplace and family.

Intervention components

The main intervention components in the workplace, will be implemented by Change Associates, are:

1. Separate gender transformative training for workers (single sex groups for female and male workers) and management staff of 18 hours (six 3-hour modules), held over 9 months and delivered to groups of 25. The session topics include communication skills (e.g. listening, body language, etc.); assertive responses; reflection and discussion of gender roles and norms, and relationships; power; violence in relationships (causes, consequences and support system); stress and conflict management; factory policy; and goal setting and being a change Maker. The curriculum is participatory, taking reference from Stepping Stones and Freirean reflective pedagogy.
2. Joint session between workers (15 female and male) and middle-management staff (10), held after third, fifth and sixth group sessions.
3. Factory-wide activities/campaigns using factories public address systems, skit play and other behaviour change communication materials
4. Awareness raising among top management (e.g., general manager, HR&A director, etc.; 2 hours)
5. Factory policy review and development. Its activities include at least one meeting in every two months with the Factory Well-being Committee. The topics include review existing and develop new gender policies and mechanisms to prevent and address sexual harassment; design and implement factory-wide promotional activities/campaigns; report progress and challenges to top management on a regular basis; and design and implement sustainability plan (initially support, and subsequently run, the joint sessions between managers and workers).

In addition to the workplace interventions, HERrespect will have an community engagement component, working with Awaj Foundation in their workers' café to develop a training module on IPV based on their existing programmatic framework, build capacity of cafés' staff to facilitate gender transformative curriculum, and develop training materials for use in the workers' café's network.

Research Design and Methods This is a quasi-experimental study involving four intervention and four matched control factories recruited by the buyers purposively. Factory matching criteria include size, location, and type of product. Care will be taken to select control factories at a distance from the intervention factories to prevent potential contamination. In the intervention factories a randomly selected cohort of married female workers and all management staff will receive group sessions. In addition, factory-wide campaigns will also be carried out. The control factories will be waitlisted for intervention once the evaluation is over.

All the workers receiving group intervention (n=400) and a cohort of randomly selected married female workers from the control factories (n=400) will be included in the baseline and endline worker surveys. The management survey will include 50 randomly selected management staff from each factory. The endline surveys will be conducted 24 months post-baseline.

Sample Size Calculation and Outcome The sample size for the worker survey was calculated considering the primary outcome, physical and/or sexual IPV against female garment workers. Considering 56% baseline prevalence, 20% effect size, 80% power, 5% level of significance and 20% lost to follow up the required sample size was 330 workers for both intervention and control groups. Accounting for a 20% lost to follow up and rounding up the sample size became 800 workers (400 in intervention and 400 in control), making 100 workers per factory.

Sample size for the management survey was calculated based on high or moderate gender inequitable attitudes of the management staff. Considering 50% baseline prevalence, 30% effect size, 80% power and 5% level of significance a sample size of 183 was derived both for intervention and control factories. Accounting for a 10% loss to follow up the final sample size for each group of factories was 200, giving a total of 400 (50 per factory).

Translation of data collection tools The worker and management survey questionnaires will be developed in English. Once the English version is finalized the questionnaire will be translated into Bengali. The Bengali questionnaire will be finalized based on feedback from pre-testing and piloting. Blinded back translation will be carried out by a third party once the Bengali questionnaire is finalized. Where needed, revisions will be made in the Bengali version based after comparison between the Bengali version and the back translation.

Recruitment of survey participants

In order to achieve the required sample size of 100 per factory, assuming that 40% of the female workers will be married and that 80% of them will finally join the groups a list of 315 workers containing information on ID and name will be obtained from each factory. An enumeration of these selected 315 workers will be carried out within the factory to collect some socio-demographic information mainly for the purpose of screening eligibility and obtaining residential addresses and other contact information for conducting the interviews and for tracking the cohort. The eligibility criteria include:

• Minimum one year of work experience in the current factory
• Currently married and living with husband. We will randomly select 100 female garment workers from the list of eligible workers. The selected workers from the intervention factories will be assigned for group sessions and will be interviewed both at baseline and endline. The selected workers from the control factories will be interviewed both at baseline and endline. If more than one worker belongs to the same household, one will be chosen randomly for participating in the study.

The management staff from mid-level factory management will be eligible for participating in the study. A list of 50 management staff will be obtained from the factories for inclusion in the study.

Pre-testing and piloting The Bengali version of the worker and management survey questionnaires will be pre-tested on 20 female workers and 10 management staff respectively using paper questionnaire. Cognitive pre-testing will be carried for sections that have not been previously used in Bangladesh. The pre-test will help us identify sections and questions that women find difficult to understand and answer and the cognitive pre-testing will help us find solutions. Training will be conducted using the revised questionnaire. About 50 interviews will be conducted by the worker and 30 by the management survey team members during piloting. The questionnaire will be finalized based on feedback from pre-testing and piloting.

Cohort tracking This will be a closed cohort study. Therefore, it is important to carefully track the study participants for successfully interviewing them both at baseline and endline. In case of loss to follow up such tracking will help us understand the reasons for dropout. One staff will be assigned to follow up the study participants over phone every two months for ensuring retention. The tracking form will include the following information (the last three are applicable only after the baseline): allocated ID, name; name of factory; name and cell phone number of husband; name and cell phone number of a guardian other than the husband; residential address and cell phone number, name, relationship, address and phone numbers of other relatives or friends or co-workers or neighbours in close to her; whether she is still working at the same factory; if not, what is the reason for leaving the factory; current employment; current residential address

Tracking loss to follow up If a woman is unavailable or untraceable for follow up interview we need to find out the reason. If she is traced and refuses to be interviewed the reason will be recorded. If she cannot be found in her previous address, we will try to track her through additional contact information.

Data collection and management Three teams, each will consisting of four female data collectors, one female supervisor, and one male field assistant will be employed for the worker survey. One quality control officer (QCO) will be responsible for rechecking all questionnaires and providing feedback to the interviewers and the supervisors. A survey coordinator will be responsible for coordinating survey implementation in the field. In order to recruit the necessary number of survey team members we will train a greater number of the candidates shortlisted based on written and verbal tests. The team will receive a 15-day participatory training on gender, violence against women, ethics, survey methods, the questionnaire, and use of tablets. Final selection of the survey team members will be based on performance during the training and pilot. The data will be collected using Personalized Digital Assistants (PDAs) to address ethical (privacy) issues and to maximize disclosure. An offline-based survey software will be developed. The interviews of the workers will be conducted in private in a location convenient for the participants outside the factory. During working days most interviews are likely to take place in the evening after work. Female workers are likely to give interviews at home commonly located in slums. If the study participant is unavailable a maximum of three attempts will be made to complete the interview. We anticipate two months of fieldwork for workers survey per round of interviews.

Two teams, each consisting of four male data collectors and one male supervisor will conduct the management survey. The QCO will be responsible for rechecking questionnaires and providing feedback. In order to recruit the required number of management survey team members we will invite a greater number of shortlisted candidates to a 8-day training on gender, violence against women, ethics in research VAWG and questionnaire modules. The final selection will be done in the way similar to the workers survey. The interviews will be conducted in private within the factory. An offline-based software will be developed for collecting data using PDA.

One programmer will be available during the survey period for necessary trouble shooting and data uploading. Confidentiality will be maintained by keeping the identification information in separate files.

Data quality monitoring Worker survey Fieldwork for the worker survey will be distributed in such a way that each interviewer is accompanied to the interview venue either by a supervisor or by a field assistant. Before moving ahead with the next assignment the supervisor and the field assistant will make sure the interview can be started in private. Once all the interviews are started the supervisor and the field assistant will make rounds to check whether the interview is going on uninterrupted. They will help, if intervention is needed, to handle gatekeepers. The supervisor will also observe the quality of the interviews, keep notes and discuss problems at review sessions. The survey coordinator will randomly choose survey teams for spot checks.

The QCO will recheck all the questionnaires and will provide feedback. The data will be uploaded to the server every day after the data collection has been completed for that day, and uploaded data will be checked by the Research Officer using a computer -based data checking routine. Due to the efficiency of this system, inconsistencies in the data can be identified within a short period of time. Problems identified in the data will be communicated to the survey team. The supervisor will resolve the problems through discussion with the interviewer if possible. If necessary, the interviewer would revisit the respondent and solve the issues consulting her. If the problems cannot be resolved using this strategy the researchers will be informed. They will suggest ways of resolving the issue depending on the nature of the problems. Five percent of the study participants will be revisited by the supervisors and the coordinator for administering a short questionnaire mainly focused on identifying problems in adhering to ethical guidelines and administering questions on particular topics.

Management survey Supervisors will assign interviews to the team members, observe the quality of the interviews, keep notes and discuss problems at review sessions. Supervisors both in worker and management survey will report to the coordinator and will communicate directly to the researchers if needed. The role of the QCO will be in line with the worker survey.

Data Analysis Intention to treat (ITT) analysis will be used for assessing the impact of HERrespect. Thus, all the workers selected for the study will be included in the analysis. The primary analysis will be a comparison between intervention and control factories will enable us to determine the impact of full HERrespect intervention (group sessions and factory-wide mobilization over control). Chi-square (for categorical variables) and t-tests (for continuous variables) will be performed to test whether there are differences in background characteristics between intervention and control groups. If differences between intervention and control factories are evident, the pre-existing differences will be controlled in subsequent analyses. The impact of HERrespect intervention on main outcomes of interest will be assessed using risk ratios derived from binary regression analyses adjusting for baseline rates. All analyses will be adjusted for the baseline prevalence and age. The same approach will be followed for assessing the secondary outcomes.

Ethical considerations This study will be guided by the WHO recommendations for ethical considerations in researching violence against women. The participation of the intervention and control factories will be based on the factories' consent to be a part of the study. Attention will be paid when designing the questionnaire to carefully and sensitively introduce and enquire about workers' experiences of violence and to ensure that questions are framed in a manner that is non-judgmental. There is evidence in the literature that the factory management is opposed to disclosure of workplace violence and workers disclosing it or attempting to address it are victimized. They may be harassed, threatened and dismissed. The questions on workplace violence will be included in a style which was as far as possible non-threatening to the factory management. Individual verbal consent will be sought prior to the interview with each female garment workers and management staff. The participants will be informed orally of the purpose and nature of the study, its expected benefits, and voluntary nature of participation. As part of the consent procedure, the participant will be informed that the data collected will be held in strict confidence. To ensure that the participant is aware that the survey includes questions on highly personal and sensitive topics, the interviewer will forewarn the participant that some of the topics are difficult to talk about. The respondent will be free to terminate the interview at any point, and to skip any questions that she does not wish to respond to. Because of the low levels of literacy and concerns regarding confidentiality the interviewer will request verbal consent of the participant to conduct the interview.

Participation in the study will be entirely voluntary. Interviews will only be conducted in a private setting outside the factory. The participant will be free to reschedule (or relocate) the interview to a time (or place) that may be more safe or convenient for her. The physical safety of interviewees and interviewers from potential retaliatory violence by the perpetrator will be of prime importance. If the focus of the research becomes widely known either within the factory or household or among the wider community it may risk the safety and security of both. Thus, at the factory level the intervention and the study will be thus introduced as activities focused on addressing IPV and improving management and in the family and community it will be framed as a survey of factory work management and female workers life experiences. Even where a researcher or field worker has not herself experienced violence, listening to stories of violence and abuse may be draining and even overwhelming for her. During the research regular debriefing meetings will be scheduled to enable the research team to discuss what they are hearing, their feelings about the situation, and how it is affecting them. These meetings will aim to reduce the stress of the field work, and avert any negative consequences. This strategy proved effective in the Bangladesh component of the WHO multi-country study conducted by icddr,b. Despite these measures, some field workers may need to be given less emotionally taxing tasks, be given a break from the study or to withdraw from the research altogether. To account for these possibilities, sufficient numbers of field workers will be recruited to allow for a 10% attrition rate of interviewers over the study. All the study participants will be given a unique a code and all the identifying information will be kept in a separate file exclusively accessed by the research team and will be used for tracking the individuals over the intervention period and to contact them during the endline survey. Care will be taken to present the research findings in sufficiently aggregated form to ensure that no participating factories and workers can be identified.

Inclusion Criteria:

• Currently married woman
• Working in the factory for at least one year

Exclusion Criteria:

• Not currently married woman
• Working in the factory for less than one year