Sub-Saharan African women continue to face socio-economic challenges and limited reproductive freedoms, which diminishes their ability to exercise agency and choice in their lives and their environment. The goal of this study is to generate rigorous scientific evidence on empowering women individually and in households through a cluster randomized control trial. The trial will test the efficacy of a multi-sectoral program targeting three critical domains of female empowerment through gender socialization education, counseling and improved access to family planning, financial literacy among couples in Ibadan, Nigeria. The innovation in this approach is the focus on creating a supportive intra-familial environment to accelerate progress towards female empowerment, not just with the multi-sectoral intervention, but also by targeting both partners of couples, individually and together. It is hoped that there will be a shift of broader community norms by building the capacity of study couples to transfer their newly acquired knowledge and skills to other couples in their community, thereby creating a ripple of change.

Background: Despite the growing influences of modernization and westernization, most Nigerian subcultures remain largely traditional. Household division of labor continues to follow traditional gendered roles with the male partner being the breadwinner and decision-maker, while women are ascribed tasks such as childcare and homemaking. These traditional roles often disempower the woman such that she has no say regarding how household funds will be spent and inhibit her from making important household decisions such as whether or when she or her children will receive preventive, curative or rehabilitative healthcare. Moreover, in Nigeria, both the customary law and the major religions stipulate that a woman should be submissive to her husband. The sociocultural and religious contexts appear to be in a conspiracy against women, and favor men who are encouraged to "be in control" of their homes. In some cases, this translates to the controlling husband preventing his wife from working outside the home, making it difficult for her to become economically independent and own assets. Such women, and those who lack the education and/or skills to join the labor force are further disempowered in the sense that they are unable to contribute to the family income, a quality known to improve women's status and decision-making autonomy. Women who do not work for cash or kind often find themselves fully dependent on their husbands for both economic and social identity, predisposing them to intimate partner violence, IPV. This lack of decision-making autonomy penetrates all spheres of women's lives, including reproduction, such that many women are unable, on their own, to choose to use modern contraception to space or limit births, without their husbands' permission.

Study goal and design: The goal of this study is to generate rigorous evidence on how to empower women individually and in households, by targeting both partners in a couple, with a multi-sectoral approach that affects three critical domains of female empowerment - creating more equitable spousal relationships, expanding reproductive choice, and increasing economic empowerment, and is potentially sustainable and can be scaled-up. By influencing these three domains of female empowerment, while simultaneously involving male partners, the intervention aims to set into motion a set of processes that will generate equitable and supportive gender relationships at the household level.

To rigorously evaluate the intervention, a cluster randomized control design was chosen. The primary goal of the evaluation is to test the impact of the gender socialization arm, the key intervention, and the added improvements in female empowerment and spousal supportiveness measures that may be seen with the incorporation of financial literacy and family planning interventions. The trial consists of 4 arms: Arm 1 - couples participate in a structured group education program on gender socialization; Arm 2 - couples participate in a structured financial literacy education program, in addition; Arm 3 - couples receive the complete package of interventions (gender socialization education, financial literacy and family planning counseling and services, with vouchers for the poorest women); Arm 4 - control group.

Study hypotheses: Couples receiving the complete package of interventions will have stronger female empowerment outcomes as reflected through measures of spousal supportiveness and equitable gender relationships in households as compared to couples receiving fewer components of the intervention, or no intervention. In addition, couples that receive only the gender socialization intervention or a combination of interventions will have stronger female empowerment outcomes as compared to couples in the control group.

Study area: The study is taking place in Ibadan, the capital of Oyo State, Nigeria. As part of the 2006 population census, the Nigerian National Population Commission prepared a list of 665,000 EAs that has served as a sampling frame for many national and subnational surveys, where a census enumeration area is defined as a cluster and is the primary sampling unit. Enumeration areas consist of approximately 211 inhabitants or 48 households. Enumeration areas form localities, localities form Local Government Areas (LGAs), and LGAs form states. The LGA is a large enough area, where multiple intervention arms can be implemented while avoiding contamination, but at the same time couples residing in an LGA are more likely to be similar as compared to couples living elsewhere. The research team worked in two urban (Ibadan North and Ibadan Southwest) and two peri-urban (Akinyele and Oluyole) LGAs.

Sample size calculations: These were based on the use of selective empowerment measures for Oyo state from raw data available from the 2013 Nigeria Demographic and Health Survey dataset (www.measuredhs.com). Assuming the intervention would improve each of the three main outcome measures by 15%, 80% power, a 2-sided type 1 error of 5%, and the proportion of married/cohabiting women ages 18-35 years who were involved in decisions regarding large household purchases to be 0.275, 225 participants in each arm would be required. An intra-class correlation coefficient of 0.015 was assumed in the calculations, as well as a fixed number of clusters of 12 per arm (48 in all). Adjustments for 20% loss to follow up rates gave a total of 282 couples per study arm. Calculations based on the above assumptions produced a minimum sample size of 1125 couples. Sample size calculations for other outcomes were smaller.

Sampling strategy: Study participants were selected through a 3-stage randomization process. First, geographical landmarks/boundaries were used to split the 11 LGAs (5 urban, 6 peri-urban) in Ibadan into two roughly equal halves. Then one urban and one periurban LGA was selected randomly from each half, using the random number generator application, after assigning numbers to LGAs in each half. Second, using a map of each of the four selected LGAs displaying localities, alternate distinct localities were selected in a serpentine fashion, following a random start. This was done to ensure that 12 geographically distinct clusters were selected from each LGA. Using a sampling frame of enumeration areas by locality in each of the 4 LGAs, one index enumeration area, and its adjoining enumeration area with a higher code was randomly selected to form study clusters. Thereafter, household-listing surveys were conducted in the selected EA clusters to enable identification and recruitment of 282 eligible couples, ages 18-35 years, per study arm to participate in the study. Where there were not enough listed couples, additional adjacent EAs were listed. Couple selection only took place once, at the beginning of the study.

Training: Training was conducted in multiple phases. Initially, a 3-day training was conducted to introduce team members to human subjects' research, the study protocol, interviewing techniques and the study instruments. Thereafter, once the team was ready to start, a refresher training was held to describe and practice the protocol for the listing and recruitment exercise. Another refresher training was held, and pre-testing done once the final instruments were field-ready. A subset of the field workers, all with at least high school education, were retained to serve as facilitators for the intervention. A 3-day training of trainers' workshop was organized to prepare the facilitators to train the couples that had been recruited into the study. Thereafter, weekly refresher trainings were held to ensure facilitators were fully ready for the roll out of the weekly intervention sessions.

Data Collection: Formative research was conducted to assess the feasibility of the proposed program design. In all, five focus group discussions were held among men, women and couples. Co-investigators facilitated the FGDs which held in neutral places that were acceptable to the respective groups. All sessions were audio-recorded and later transcribed. Where the session took place in Yoruba, the local language, both transcription and translation were done. A quick analysis was performed so that the information obtained could be used to refine the implementation of the intervention.

Randomization: Following selection of clusters, recruitment of couples, and baseline interviews, a statistician conducted block randomization with stratification (block size of 4) whereby each LGA had all 4 study arms represented. Clusters were randomly allocated into one of four study arms (Arm 1: gender socialization training; Arm 2: gender socialization + financial literacy training; Arm 3: gender socialization + financial literacy training + FP counseling/service; Arm 4: control).

Intervention: A total of 3 gender socialization (GS) and 2 financial literacy (FL) workshops were held weekly by cluster, separately for male and female participants to avoid any potential gender conflicts. To minimize loss to follow up, those who received both of these interventions had the sessions back to back, on the same day. Such sessions lasted a total of 3 hours, at most. Additionally, one couple session was conducted for both the husband and wife together for each of these interventions, as well as one joint workshop of family planning counseling for the arm that was randomized to the full intervention. The workshops were interactive and organized around lectures, exercises, role-play and group activities. Peer learning formed an important part of the skill acquisition process. Gender socialization workshops raised awareness on gender, its social construction and inequality, notions of masculinity and femininity, division of labor, access and control over resources. The workshop also included skill-building sessions on effective communication and negotiation and relationship building. Financial literacy workshops promoted knowledge and skills in budgeting, financial planning and accessing and using financial services and income generating activities. Couples were provided with study materials for their own review and to facilitate their subsequent teaching activities. Immediate post-intervention surveys were conducted immediately after the final intervention session for each group, or within a week for couples who did not attend the last session.

The FP component of the intervention package was delivered by FP providers in NURHI-supported health facilities. These facilities were well stocked and had training and re-training activities for the providers. Couples were referred to these facilities, but had the option of choosing any facility they would like to attend.

In the FP arm, the poorest participants were invited to be part of a voucher system to secure a contraceptive method of their choice from a health facility. Most contraceptive methods are available free of cost in public sector health facilities in Nigeria. However, users have to pay a registration fee, pay for a pregnancy test, and in some facilities, there is an additional fee associated with consumables, and the insertion of IUDs and implants, as well as sterilization. A team member, the voucher officer, was responsible for accompanying women to the health facility of their choice to obtain FP counseling and their method of choice. The voucher officer paid all the bills at the health facility and refunded the woman her transportation expenses. She was also responsible for reminding the woman of an upcoming appointment prior to the date, usually 3-5 days in advance.

Post-intervention assessments: Surveys were conducted immediately following the last intervention session (Immediate post-intervention) for all three intervention arms, and 6 months post-intervention for all 4 study arms (Endline). At endline, six focus group discussions were also conducted in order to better understand the impact the intervention had on participants.

Inclusion Criteria:

• Women ages 18-35 years who live with their husbands or partners within the pre-selected study clusters
• Consent from husband/partner as well as eligible woman
• Women in polygynous unions, if the wife co-resides with her husband

Exclusion Criteria:

• Either woman or partner does not give consent

The funders require that data be made publicly available, and we will do this. However, we are yet to discuss detailed preparations for public release with the funders.

Data will become available six months following completion of analysis of baseline data. Duration of availability will depend on the host server from the funders. The current funding stream does not elaborate on this.

Completion of a form indicating demographic data and affiliation of the person making the request, as well as the intended use of the data, and variables of interest.

The intervention involved attendance at training workshops, so once randomization was completed by the biostatistician, random assignments had to be revealed.