Blavatnik School of Government, University of Oxford
On 5 and 6 November 2019, RISE joined education practitioners, researchers, and other members of the People’s Action for Learning (PAL) Network for the PAL Network Conference: Ensuring All Children Learn in Kathmandu, Nepal. The two-day conference explored how citizen-led assessments (CLAs) and the CLA community can evolve to meet existing challenges and further spur citizen action to raise learning levels.
In 2005, an NGO called Pratham first launched an ambitious project to measure how many children had achieved basic literacy and numeracy across more than 600 rural districts in India. This project was a citizen-led assessment, or CLA: a type of survey typically organized by a non-governmental organization or civil society organisation that relies on trained citizen volunteers to assess children’s literacy and numeracy in their homes. Like many CLAs that have followed, this survey was primarily motivated by concern about low learning levels in schools. "We could see kids having difficulties," Rukmini Banerji, CEO of Pratham, said in a speech at this year’s PAL Conference, and needed to understand at the core what was happening—"not just in the way you and I understood it, but the way children and their families understood it." With thousands of volunteers across a network of community organisations, half a million children were tested in the first Annual Status of Education Report (or ASER)—the first such CLA conducted. These were local community volunteers—parents, teachers, school administrators—hitting back, with data.
Figure 1. Findings from ASER over the years: The percentage of Grade 5 children in government schools in India who could read a Grade 2 sentence.
Since that first assessment, the ASER survey has been conducted nearly annually, and the findings have been even more shocking than expected. Not only does data suggest that less than half of Grade 5 children in government schools in India can read at a Grade 2 level, but the assessments also indicate the situation worsened between 2008 and 2016. ASER’s findings spurred a wave of debate in India around low learning levels and have proved a powerful tool for building local parent and wider community engagement around learning.
National, regional, and international assessments such as PISA, TIMSS, and PIRLS have important limitations when it comes to mapping children’s learning progress, especially in the early years. These kinds of assessments are done at only one (or at best, two) points in the schooling cycle, and tend to take place later on in schooling— providing little to no data around early and foundational stages of learning. They leave significant holes in the aggregation of learning data, which make it difficult to know if kids are mastering the most basic of skills, such as simple literacy and numeracy, when it matters most.
CLAs like ASER are designed to fill these important gaps in data and knowledge on learning outcomes. Most CLAs aim to collect data annually across a full cohort of children, allowing us to better understand kids’ learning trajectories through age and stages of development. In addition, CLAs use alternatives to reading and writing assessments, such as oral testing, which makes it possible to assess learning progress among younger children— a guiding feature of the PAL Network’s newly proposed citizen-led assessment of numeracy (CLAN) framework unveiled during the conference.
Since their beginnings more than a decade and a half ago, the use of CLAs has grown considerably, with CLAs conducted in Mexico, Kenya, Mali, Senegal, Tanzania, and Uganda. Uwezo, a large scale CLA conducted across Kenya, Tanzania, and Uganda, has sought to adapt assessment tools to meet the needs of these countries, providing valuable new information on learning outcomes in these countries that was not previously available. The most recent Uwezo report in Uganda, for example, showed that literacy and numeracy among Grade 3 and 6 students declined between 2015 and 2018. During the PAL Conference, Felipe Havia presented on a more recent CLA, Medición Independiente de Aprendizajes (MIA) in Mexico, whose important data has shown that 38 percent of Grade 5 kids were unable to resolve a simple maths computation (e.g., 3 + 4). Their discovery has been quickly evolving into interventions to remedy this.
Rukmini Banerji @Pratham_India @asercentre reminds us why and how citizen-led assessments began - different because working at grassroots collectively with communities to understand problem and want to something about it #PALconf2019— Pauline Rose (@PaulineMRose) November 5, 2019
CLAs are important, not just because they provide annual data to help local actors map children’s learning progress, but because they aim to put this information into the hands of parents and other local actors on the ground, enabling them to hold schools and local officials accountable for learning outcomes.
Today, CLAs have become an indispensable part of the ecosystem for change: both a means to raise awareness around low learning levels and a force for bottom-up accountability and action to improve children’s learning in schools. "It’s disturbing at a personal level [as a parent] when you think you’re doing the best you can for your children, and it’s simply not good enough," Rukmini Banerji expressed during her plenary session at the PAL Conference. By doing CLAs "in a way people could feel the problem and want to do something about it," local actors can create real change in communities – from the chain of libraries started by a district partner in Bihar, to the letter from a community leader to his district education officer demanding action in Kenya in response to Uwezo results. Data in the hands of parents, teachers, and schools is a powerful tool, empowering locals to be disruptors of the learning crisis and giving them a stake in its resolution.
They’ve also played a fundamental role in shifting national and global conversations to improving learning in school—for instance, by contributing to an increased focus on learning outcomes thanks to their evidence showing the seriousness of the learning crisis. There is also some evidence that CLAs have contributed to the prioritisation of learning in national policy documents.
At the same time, CLAs also help to inform teachers and local actors on the ground in their quest to support learning improvements. As Pauline Rose suggested during the conference, CLAs are crucial because they enable teachers to better understand where kids are and design interventions at the right stage to keep kids on track.
In the closing plenary @PaulineMRose reflects on the power of citizen-led #assessment tools for #teachers in the classroom to better understand the progress of their students - an approach effectively used by @TaRL_Africa #PALconf2019 #SDG4 pic.twitter.com/E0GA855frT— Hannah-May Wilson (@hmaywilson) November 6, 2019
During the conference, both Lant Pritchett and Rukmini Banerji spoke on the need for CLAs to move beyond providing data alone to spearheading action, both challenging and re-organising the way that education systems work in order to ensure all children can learn. To that end, Rukmini Banerji reminded us that "all of this has to be translated, whether on large-scale or small-scale into ways of showing the world that a solution is possible."
ASER from Pratham in India is itself a great example of how assessment data can, in turn, rally support for action towards improvements in learning. Pratham have since collaborated with government to design a series of interventions to raise learning levels in schools. One example of this is the Teaching at the Right Level (TaRL) Programme—an intervention that groups children by level instead of age to ensure that curriculum is aligned with children’s understanding. Cases like this underscore how data framed for action can have exactly that result.
Meanwhile, in East Africa, Uwezo has provided inspiration for new programs within Twaweza—Uwezo’s parent organisation that works on enabling children to learn and citizens to hold governments accountable for learning—to design teacher-driven strategies which promote learning, for example, in Ugandan schools. Representatives from Young 1ove, another NGO, also spoke at the conference to share how CLAs like Uwezo are being used to inform approaches to scale TARL across Africa in countries like Botswana
With these successes in hand, CLAs have the potential to go the distance from data to action—and with the support of a global community, they are likely to have even greater impact. Rukmini Banerji’s suggestion that "this is not an unsolvable problem," but one that ordinary people can solve by learning from other people’s experiences, is an encouraging one. Friendships being forged at the intersection of CLA theory and practice through partnerships being facilitated by the PAL Network, in the coming together of research and practitioner communities at the PAL Network Conference, and through the South-South sharing of lessons, have a critical support role. Lessons from others in the Network provide an important basis, not for doing more of the same, but for learning from one another, and giving other civil society organisations and NGOs the support they need to build solutions with local people at the centre.
The Kathmandu Declaration, a product of the two-day conference, articulates the commitment of the PAL Network and a growing community to these values. "We are different and we need to continue to be different," Rukmini Banerji said, acknowledging that the path to change is hard, not easy, and "as a group of people who have come together voluntarily from across the world, it is our responsibility to share easy-to-do solutions to solve these problems at scale."
The PAL Network Conference, "Ensuring All Children Learn: Lessons on Inclusion and Equity from the South" took place in Kathmandu, Nepal from the 5-6 November 2019 and was attended by members of various RISE teams and participants in RISE’s Community of Practice activities.
Videos are available from the Conference (provided by Pratham):
RISE blog posts and podcasts reflect the views of the authors and do not necessarily represent the views of the organisation or our funders.