Early Findings from Longitudinal Study: AMPLIFY Partners Demonstrate Agency Gains Among Girls Across Dosage Categories
Our work in girls’ agency is innovative as it not only addresses a critical outcome for girls but also centers the voices of community leaders across Eastern Africa. We have been able to do this work and create a powerful measurement tool because of their commitment to creating a measure that demonstrates powerful, and contextually relevant outcomes for girls from a community-driven perspective.
AMPLIFY Girls began collecting longitudinal data on girls’ agency in 2020. The longitudinal agency study seeks to establish our partners’ collective impact in increasing girls' agency. This is an anticipated culmination of a 3-year collaborative effort to develop a validated measure of agency - the AMPLIFY-Sidle Agency survey- which is currently in use by all AMPLIFY Girls partner organizations to evaluate the impact of life skills programs on girls' agency.
Our goal was to get baseline (a measure of agency at the beginning of the program) and endline data (results at the end of partner programs) at the 1-year follow-up mark from a representative sample from various life skills programs targeting adolescent girls beneficiaries.
In 2020 we were able to produce some early cross-sectional findings which suggested that AMPLIFY partners were substantially increasing girls’ agency across the board. Shortly after this cross-sectional study was completed programs and schools closed in every country. These disruptions made collecting endline data nearly impossible for most of our partners and so data collection restarted in 2021. As of February 2022, we have endline data from 9 out of 17 partner programs participating in the longitudinal study.
Our sample included 479 girls and young women, aged 10 -24 attending one of 9 AMPLIFY Girls’ Member programs in Kenya, Tanzania, and Rwanda. The sample does not yet include any data from Uganda given the long school closures and related program closures experienced in that country, we anticipate limited representation of our Uganda partners in our final sample. The sample is also missing several high and medium dosage programs, whose results we expect to be included in the final report.
AMPLIFY Partners Demonstrate Agency Gains Across Dosage Categories
From some early insights from our longitudinal analysis, we are seeing substantial increases in the average agency scores across our partners’ cohorts. The size of these results varies by dosage category.
Dosage categories were determined based on the number of hours per year girls participate in programming.
Low dosage programs typically engage girls a few weekends a year (<100hrs).
Medium dosage programs engage girls in regular weekly after-school programming (100hrs-300hrs).
Medium-high dosage programs engage girls 350 - 1,000hrs.
High dosage programs are boarding schools with 1,000+hrs of programming in a year.
Figure: The graph represents standardized agency scores, the 0 (zero) line is identical to the mean score, any scores below the line indicate below-average, and those above indicate higher than average. The differences between baseline and endline represent gains in agency scores following participation in programming.
Here we see that there are gains in agency across all dosage categories—with the largest increases in the medium and low-dosage programs. Medium and low dosage programs are typically offered to any girl wishing to join in the context of a government school, whereas for most other programs girls have to “qualify” to join through varying types of selection processes. In the results, medium dosage programs appear to engage girls who start with the lowest average agency scores.
These early findings confirm findings from our cross-sectional results which also showed the largest gains for medium dosage programs - that engage girls in regular weekly after-school programming.
Gendered Norms
There are four categories of skills or beliefs that together make up the AMPLIFY Agency measure: Self-Beliefs, Environmental Beliefs, Self-Governance Skills, and Leadership Skills. This will constitute the difference in baseline to endline for each of the four factors as well as for Agency overall.
As of now, our results show a steady and substantial increase for Agency overall as well as three of its component factors, Self-Beliefs, Self-Governance Skills, and Leadership Skills, with relative stagnation in the Environmental Beliefs category. Our research team has flagged environmental beliefs as an important area for further study on best practices amongst our partner cohort and beyond.
Learning on the Effects of Age, Program Duration, and Parental Education on Agency
In order to investigate the influence of different factors on girls’ Agency scores, we modeled the effects of program dosage, program duration, and girls’ individual characteristics such as age and parental education, on increases in Agency.
Our current results show that girls in younger age groups (10-14 and 15-18) have larger gains in agency compared to girls who were 19-25. This is an interesting finding for practice suggesting that there may indeed be a ‘sweet spot’ in age for improving girls’ agency.
The diversity of AMPLIFY Girls’ program models, is also an interesting opportunity to study the effects of diverse program structures on agency outcomes. As mentioned, AMPLIFY Girls’ partner programs fell into one of four dosage categories based on the number of hours/year girls were engaged in programming.
Significant positive effects were found for low, medium, and medium high-dosage programs when compared to high dosage, with the largest increase in medium dosage programs.
‘With each increase in program duration, agency scores also increased.’
When the effects of dosage and program duration are taken together, our model points to medium dosage, year-long programs as having the greatest impact on girls’ agency outcomes above and beyond the effects of the other covariates in the model.
We are excited to continue the process of innovation and design of participatory evaluation and research strategies for our shared outcomes, our collective impact, and our girls.
Next Steps on Longitudinal Study (2022):
Finish data collection and conduct final analysis on collective impact.
Investigate the relationship between predictor variables and girls’ gendered beliefs.
Conduct qualitative comparative analysis—mapping longitudinal results to Group Concept Mapping findings.
Write-up results in AMPLIFY Research Brief
Stay tuned for the final report of our Agency longitudinal Study.