This week at World Bank EduTech
week of January 11, 2021
Happy New Year from the EdTech team!
We ended 2020 by looking back at what we had achieved during the year. The “black swan” event was evidently the COVID-19 Pandemic, the resulting disruption of the school system, and the successes and failures of EdTech to come to the rescue. Our last Medium post of 2020 was “EdTech is not an add-on anymore but an essential part of education systems going forward.”
We begin this year with an attempt to predict (or more aptly anticipate) what we think will be five of the most important trends in EdTech in 2021 in our work. Trying to predict the future is a human folly, after all, who could have predicted the COVID pandemic? Yet, perhaps, that’s not strictly true as some people were aware that a pandemic would happen although not aware of when. In a similar spirit, the trends we think will be front and center in 2021 seem obvious, many are underway and were becoming a priority even before the pandemic and some we will drive or accelerate as part of our work. Some may be off in terms of timing and some may not happen at all. But that is the fun in gazing into crystal balls!
We will be looking forward to seeing how our predictions pan out at the end of 2021.
Happy New Year to all the EdTech community, partners and friends!
EdTech used for remedial learning to stem learning losses: Adaptive Learning
The short-term focus will remain on continuity of (remote) learning and how to re-open schools safely and as soon as possible. However, it might be fair to say that the single biggest issue likely to be on the minds of education policy makers this year will be mitigating learning losses as a result of COVID-19 disruptions to the school system. The alarm bells on learning losses are already being sounded across the world. For Low and Middle Income Countries (LMICs), the COVID-19 pandemic is likely to exacerbate learning poverty. The solution appears to be remediation: “the ongoing effort to support students in catching up on lost learning through a variety of means, including additional teaching time, focused content, and specific pedagogy (e.g., compressed learning, accelerated learning)”- definition from this UNESCO Toolkit on Remediation worth a read and that includes strategies and case studies.
Remediation may take the form of going back to cover the previous year’s curriculum before covering the new year’s curriculum, tutoring and Teaching at the Right Level (TaRL) for groups of students — a form of adapting instruction to the children’s level. Brookings has modelled various remediation scenarios and concludes that going back over the syllabus and adapting instruction to the child’s level “both fully mitigates the long-term learning losses from the school closures and provides children with a full year’s more learning than if there had been no shock.” But given the pressure to get students to catch up on lost learning and to get back on track in the curriculum (especially for those countries that don’t end up writing off the last school year making every student repeat the year), technology is likely to be explored as a solution. One technology that is likely to be a focus for remedial actions is Adaptive Learning also called Personalized Adaptive Learning (PAL).
The World Bank has supported adaptive learning programs such as the one in Ecuador (at TVET level) and in Dominican Republic (learning mathematics at elementary level) with promising results.
The EdTech team is just published a blog describing a roadmap for policy makers considering an adaptive learning system.
EdTech used for remedial learning to stem learning losses: Tutoring
Another remediation strategy to stem learning losses is tutoring. Several studies (mostly from high income countries) show that tutoring is among the most effective interventions to support student accelerated learning and to stem learning losses.
One recent study from JPAL-North America summarizing a meta-analysis of 96 randomized evaluations of different tutoring programs found that “across all studies included in this analysis, tutoring programs consistently lead to large improvements in learning outcomes for students.” Most effective programs use teachers (instead of untrained volunteers), work best at early grades for math and reading and work best at school (less so after school or at home but mostly because of absence of qualified teachers outside school). The EdResearch for Recovery Project at Brown University draws on extensive research to recommend that “high-dosage tutoring that is directly tied to classroom content — helping students succeed in their coursework — can substantially accelerate learning in both math and reading for the most struggling students.” Some countries are taking note. The UK is launching a government funded national tutoring program. Some in the US are calling for a national tutoring program. A recent study from Italy using online tutoring for high school students also sees gains in academic performance, well-being, and socio-emotional skills of disadvantaged children and plans are underway to scale the program.
On the private sector side, the pandemic has led to growth of the private online tutoring industry with growth most prominent in Asia. We expect to see the conversation on tutoring and even government-led large-scale tutoring initiatives begin to take root in LMICs. For these countries, two remediation tutoring strategies could be worth exploring: First, the remote learning infrastructure put in place during COVID-19 could be used for remediation and tutoring when schools re-open. Secondly, some countries have experience using contract teachers and the use of contract teachers as a policy could be scaled up to provide focused tutoring for children in the short term. However, tutoring is expensive and in the context of expected economic difficulties, there may be limits to education systems investments. Already, there is evidence that contract teachers have not been paid during the pandemic. It will be worth watching how this plays out.
Early warning systems, learning analytics and reforming EMIS
Even before the pandemic, high income countries (led by the US) were looking at early warning systems to “ identify students that are at risk allowing educators to provide timely interventions and supports”. These systems use data on student performance through regular assessment and student attendance and modern data analytical systems to predict students most at risk of dropping out or falling behind. Some researchers are beginning to look at how early warning systems could support students during remote learning. Early warning systems rely on learning analytics - “the collection and analysis of data about learners and their environments for the purpose of understanding and improving learning outcomes.” Learning analytics are widely used in higher education but there is also increasing interest in basic education. As interest has grown, standards are being developed to support inter-operability and exchange of data. Learning analytics are foundational for adaptive learning. We expect to see a growing interest in learning analytics and early warning systems linked to PAL in LMICs this year. However, this will require that these LMICs upgrade their Education Management Information Systems (EMIS). At their heart, learning analytics require that every student have a unique student ID, which few education systems in LICs countries have. Thus, part of the modernization of EMIS systems will center on moving from what we call EMIS 1.0 focusing on school census to EMIS 2.0 focusing on learning data.
The World Bank is supporting unique student IDs: for example in a project under development in Sindh (Pakistan) intends to “utilize an automated system to create unique student identification numbers and then monitor and track student attendance — and in the long-term student performance.” We expect to see more of these projects.
The uptake of these data intensive systems brings up several challenges related to data privacy and security explored in this EdTech blog.
Closing the digital divides: access to digital infrastructure
If there was one glaring gap revealed by the pandemic, it was the inequality or gap in access to digital technology between countries, within countries, between communities and between genders — with the most disadvantaged communities and girls particularly facing the biggest digital gap.
The issue of the digital divide has been recognized since the 1990s as the internet permeated society and initially it was focused on access to the internet. The internet access digital divide is still a big challenge in many countries and within countries even in the most developed countries. But there is now understanding that the digital divide is actually composed of a number of divides — access to the internet, access to devices, access to software and digital content along with the skills and organizational capabilities to utilize this technology.
We expect to see renewed efforts to close the digital divides focusing on digital infrastructure. Expect to see many countries focus on connecting schools and perhaps even a renewed interest in community Wi-Fi networks to connect the poorest households to the internet. Expect to also see efforts to get digital devices and software into the hands of students and teachers and a rally to try and get lower prices for digital devices. Could we see any rallying call around a $100 laptop again? We don’t know but we would not be surprised if renewed calls for low-cost special education devices emerge. Finally, another important access gap will be digital content. The World Bank and the development community will have an important role.
The Education GP is working on a comprehensive strategy to operationalize the EdTech Approach that focuses on addressing the digital divides (stay tuned for more on this!).
Our Digital Development GP colleagues are leading the connectivity drive through initiatives such as the Digital Economy for Africa (DE4A) which could be a vehicle to connect schools and education institutions.
At the corporate level, IDA 19 commitments include a target to support 25 IDA countries to double their broadband penetration (16 on the African continent), including eight in landlocked countries, by 2023 leading to enhanced access to broadband internet for 50 to 60 million people.
The World Bank is partnering with other players such as UNICEF and ITU on the GIGA initiative to connect every school around the world to the internet.
Closing the digital divides: digital skills
As the world shut down due to the COVID-19 Pandemic, many routine activities moved online or relied on technology for continuity: from learning, working, shopping, healthcare, entertainment, access to news to keeping in touch with friends and relatives. Zoom even became a verb. While the digital divide relating to digital infrastructure quickly became apparent, it soon dawned that digital skills were also a key barrier to effectively transition to a life online. Even before the pandemic and the ensuing lockdowns, it was already evident that digital skills constitute a key barrier to adoption of digital technology in LMICs even when (mobile) broadband exists (GSMA). WEF opines in a recent report that “Digital literacy is an additional barrier to adoption by certain communities, both in developed and developing markets” and that “Digital literacy is also closely intertwined with the availability of relevant use cases, localized content, products and services attractive for user consumption, or lack thereof in some markets”. Further, digital skills are essential for the future of work. A recent World Bank-IFC report across five countries in Sub Saharan Africa estimates that by 2030, ~50–55% of all jobs in Kenya, ~35–45% of all jobs in Rwanda, Nigeria and Côte d’Ivoire and ~20–25% of all workforce in Mozambique are expected to require digital skills. We expect to see an increasing demand for digital skills program in our portfolio.
Already, the Education GP in collaboration with the Digital Development GP has undertaken a significant amount of work and more work on this topic to support our clients is underway (stay tuned for more during the year).
