Please fill out this short user survey of only 3 questions in order to help us improve the site. We appreciate your feedback!
Partnership to End Child Marriage: A Cluster Randomized Intervention to Delay Marriage in Bangladesh
Last registered on June 18, 2018


Trial Information
General Information
Partnership to End Child Marriage: A Cluster Randomized Intervention to Delay Marriage in Bangladesh
Initial registration date
June 14, 2018
Last updated
June 18, 2018 10:14 AM EDT
Primary Investigator
Population Council
Other Primary Investigator(s)
Additional Trial Information
On going
Start date
End date
Secondary IDs
Background: Despite significant improvements in health and development indicators in Bangladesh in recent years, child marriage remains stubbornly high, with about half of women 20-24 reporting being married by age 18, and social norms around early marriage persist.
Methods: To examine whether a social norms intervention can be an effective intervention to address child marriage, we designed a multi-arm randomized cluster trial. This project, called Accelerating Action to End Child Marriage in Bangladesh, is designed to reach adolescent girls ages 10-19 residing in two child marriage hot spot districts of Bangladesh. The program to be implemented in 72 randomly selected communities, while another 24 randomly-selected communities are serving as controls. Impact is assessed through baseline, midline and endline surveys and qualitative data collection is planned to provide data with which to target and improve intervention strategies.
Discussion: The project is expected to identify ways to scale up skill-building activities along with alternative community participation strategies to ensure the long-term sustainability of the program effect across the community. The project will evaluate a skill-building program that is girl-centered, cost effective, sustainable, and community owned that will contribute the country’s overall initiative to end child marriage. The research will edify the policy makers through evidence-based skill-building interventions by adding and evaluating a new community engagement component. The project will document intervention strategies in detail and provided cost effectiveness estimates of the three approaches applied in intervention. It will generate approaches as well as materials to gather programmatic evidence of relevance to future programming.
External Link(s)
Registration Citation
Amin, Sajeda. 2018. "Partnership to End Child Marriage: A Cluster Randomized Intervention to Delay Marriage in Bangladesh." AEA RCT Registry. June 18. https://doi.org/10.1257/rct.3077-1.0.
Former Citation
Amin, Sajeda. 2018. "Partnership to End Child Marriage: A Cluster Randomized Intervention to Delay Marriage in Bangladesh." AEA RCT Registry. June 18. http://www.socialscienceregistry.org/trials/3077/history/30912.
Experimental Details
The intervention aims to achieve the following:
• Empowering at risk and already married adolescents girls in particular, with life skills education and Comprehensive Sexuality Education
• Providing alternatives to child marriage and mitigate the impact on married girls, through enhancing access to education-, economic opportunities.
• Engage the community people within support groups and related activities to changing social norms

The intervention is designed to reach two cohorts of girls, for one-year programs over the course of two years, by offering skills development programs implemented in 72 randomly selected communities while 24 will serve as controls. A government partner will implement all activities in the Kishori Resource Centres, including training of field based Gender Promoter and behavior change communication (BCC) activities with technical assistance from Population Council.

The intervention design builds on findings from BALIKA (http://www.popcouncil.org/research/balika-bangladeshi-association-for-life-skills-income-and-knowledge-for-ado) , including using the learning center model that offers life skills, gender rights awareness training, and livelihood skills using community learning centers equipped with information and communication technology (ICT) resources. The program approach targets all adolescent girls living in high-child-marriage-prevalence districts and measures community-level impact. The engagement with girls focuses on skills development to enhance capabilities, while the engagement with communities focuses on changing gender norms and practices.

The intervention will test three strategies for engaging the community: one that engages only by reaching adolescents but does not otherwise engage community; one that reaches adult male and female support groups in addition to adolescent girls; and one that reaches adult female support groups in addition to adolescent girls. Community engagement involves frequent interactions (bi-monthly) with community members to foster an environment that values girls as assets rather than as liabilities to the household and community. All three intervention arms include the creation of safe spaces and provide training for basic life skills, gender rights, SRHR and livelihoods skills for adolescent girls. Communities assigned to the control group receive neither a skill-building intervention nor community engagement nor placebo exposures.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
Proportion ever married
Proportion married by age 18
Proportion ever pregnant
Proportion pregnant by age 18
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
The evaluation design is a four- arm cluster randomized controlled trial where 96 Unions will be randomly assigned to three intervention and one control arm. The intervention is implemented as follows: all three intervention arms will receive adolescent skill building interventions. The skill building approach differs from previously tested approach used in BALIKA in that it is offered by a local government partner with the objective of identifying a sustainable and scalable model implemented by the government. Arm 1 is skills only; Arm 2 includes skills plus outreach with women; Arm 3 includes skills plus outreach with men and women; and Arm 4 serves as a control.
Experimental Design Details
Randomization Method
We used the random number generation sampsi command in STATA 13.0 to randomly assigned the 96 clusters to one of four possible arms, and then randomly select 30 adolescents within 2 km of the the selected school of each union from the list of adolescents identified in household enumeration survey to participate in the adolescent survey.
Randomization Unit
Cluster randomization to identify geographic areas for intervention (clusters are unions). Participation in the survey then randomly selected from within assigned clusters.
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
96 unions
Sample size: planned number of observations
3000 adolescents
Sample size (or number of clusters) by treatment arms
24 control unions
72 intervention unions (3 arms-- 24 in each arm)
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Data from the 2011 Population and Housing Census were used to obtain estimates of the proportion of females aged 15-19 currently married in Nishindara union of Bogra district (33%) as the lowest possible value and the proportion in Khamarkandi union of Bogra district (61%) as the highest. On an average, 48% girls are currently married in our proposed areas and we expect a decrease of 20 percent (38%) child marriage. We calculated the optimal sample size that would allow adequate statistical power to detect an impact within a reasonable time-frame and would be logistically feasible to implement. We assumed an intra-cluster correlation of k=0.05 and an average cluster size of 30 adolescents per union. Based on these assumptions, calculation of the number of unions per arm was performed by using the following formula [37]. c=1+((〖Z_(1-α⁄2)〗_ +Z_(1-β) )^2 [(P_1 (1-P_1 ))/n+(P_2 (1-P_2))/n 〖+k〗^2 (P_1^2+P_2^2)])/〖(P_1-P_2)〗^2 Where, c is the required number of union per arm, Zx is the x’th percentage point of the standard normal distribution, P_1-P_2 represents expected magnitude of the effect of a program or difference between the treatment and control group, n the cluster size, P1 is the probability of an event in the control group, and P2 the probability of an event in the treatment group. We found that a minimum of 24 unions per arm and a sample of 30 adolescents in each union with 720 respondents per arm is sufficient to measure the impact of the program with 83% statistical power. We targeted a final sample size of at least 2880 girls (30 per union; but plan for 3000 girls) in order to detect a 20% reduction in child marriage prevalence by the end of intervention period.
IRB Name
Population Council Institutional Review Board
IRB Approval Date
IRB Approval Number
Post Trial Information
Study Withdrawal
Is the intervention completed?
Is data collection complete?
Data Publication
Data Publication
Is public data available?
Program Files
Program Files
Reports, Papers & Other Materials
Relevant Paper(s)