Combining High-Tech and High-Touch to Personalize Learning for Every Child

8th January 2019

The Education Commission Learning Generation report delivered a stark warning: if nothing changes, by 2030, 825 million young people – roughly half of the young generation today – would reach adulthood without the skills they need to thrive. The World Bank’s 2018 World Development Report echoed a similar warning and described the global learning crisis as “a moral crisis.” Its 2019 WDR on The Changing Nature of Work called on countries “to invest in education and health with a fierce sense of urgency to harness the benefits of technology and to blunt its worst disruptions.” The crisis is exacerbated by the rapidly shifting global economic, social, and geophysical environment. The World Economic Forum predicts that 65 percent of children entering primary school today will eventually work in jobs that do not yet exist. Education systems developed in the age of the first industrial revolution are far outpaced by the current fast changing environment for which they should be preparing young people. There has been rising evidence that this traditional way of teaching is failing to provide adequate learning opportunities and breadth of skills for the 21st century, which might be leaving millions of kids and young adults behind.

Root Causes of the Global Learning Crisis

We have to look back almost two centuries to find the root causes of today’s global learning crisis. The current classroom and schooling model, where teachers deliver uniform and standardized content through lectures for students with little space to adjust to students potential and needs, appears not much different from the mass production systems of the first Industrial Revolution era in the late 18th century.

In the meantime, however, economic systems have been continuously transformed from the 1st to 2nd, 3rd and now 4th Industrial Revolution. They have moved from mass production toward mass customization. In addition to digital technologies, artificial intelligence (AI) technologies are now creating disruptive change in markets for products and services by calculating optimal solutions for everyone in the digital space, combining big data and cloud computing, and providing them cheaply in the physical space with 3D printers or digital devices. This approach is being applied to address some of the world’s most intractable problems, such as the preservation of natural capital. The Global Forest Watch, for example, is using billions of satellite data points and cloud computing to monitor and manage forests around the world, call out unsustainable activities, defend the land and resources, and fight climate change.

A critical question before us is how we could take advantage of new technologies to develop human capital with a similar personalized approach? Education experts and policy makers have long recognized the importance of personalized learning – the education term for customization. But it has never been realized in practice for all students, only for the lucky few. Addressing the global learning crisis and providing equal opportunities for all children and young adults hinges on our ability to provide personalized learning at scale for every student.

But how can this be done? It could be a daunting challenge. The insurmountable burdens of meeting the demands of each student with vastly different backgrounds and abilities within today’s models of education have hindered and discouraged even the most enthusiastic teachers from providing the personalized learning students need. A new approach is needed.

High-Touch High-Tech Learning

Technologies that have so far been used as solutions to address the learning crisis have had mixed results at best. New tools are constantly emerging, however, and we believe that one of the most promising breakthroughs towards truly making education available for all in a personalized way could be through the application of artificial intelligence (AI). Why? Because new types of technologies like AI are able to cater to the different abilities and backgrounds of every student, something that earlier and more standardized technological applications were not able to do. In this way AI addresses one of the key binding constraints of learning, which is the mismatch between classroom instruction (delivered in a standard way) and student learning levels.

AI adaptive learning algorithms have advanced with remarkable speed in recent years. They can provide targeted support at scale by adapting teaching materials to students’ prior knowledge. Students can use AI tools to understand and remember content in different subjects at their own pace. Teachers can use AI algorithms to provide different learning paths for each student because AI can analyze learning progress of each individual student.

One example of successful integration of a personalized, technology-aided instruction called Mindspark can be found in India. According to a recent evaluation finds that “a key feature of the software is that it uses its extensive item-level database of test questions and student responses to benchmark the initial learning level of every student and dynamically personalizes the material being delivered to match the level and rate of progress made by each student.” A study of the effects of the instruction program with middle school students in India shows significant effects on test scores. While the study finds similar absolute gains in test scores for all students, relative gains for academically weaker students were greater. This underlines that well-designed technology-aided instruction can sharply improve productivity in delivering education.

Another positive example can be found at the Arizona State University (ASU), where more than 65,000 students have already benefited from an adaptive learning system in 12 basic courses. These basic courses are often a predictor of academic success because they serve as a gateway to numerous academic majors and career paths. Introducing the adaptive algorithm called Assessment and Learning in Knowledge Spaces (ALEKS) in a college algebra course resulted in the drastic increase in completion rate by 20.5 percent on average and by 28.5 percent for those with math placement below algebra. Students who received adaptive tutoring rather than traditional algebra were also more likely to attempt a subsequent, more advanced mathematics course and major in a STEM field.

Adaptive learning platforms provide a cost-effective way of learning at the right pace at the right level for students in resource constrained contexts. It highlights the potential for new technologies to enable low-income countries to leapfrog current constraints to development as suggested in earlier research. Above all, the availability of such platforms could radically reduce the burden of teachers to provide basic content across a large number of diverse students and free up time for teachers to focus on “high-touch” learning of students through projects, discussions, hands-on experiences, and nurturing higher order skills like complex problem-solving and socio-behavioral skills such as teamwork. A combination of “High-Touch” and “High-Tech” learning models may be the best way forward. The diagram below, based on Bloom’s Taxonomy, shows how High-Touch High-Tech Learning enables students to create: they first remember and understand content through interacting with AI and mobile technology (high-tech), and then learn to apply, analyze, evaluate what they understand, and finally learn to create, through engaging with teachers (high-touch).

Source: Johnson, Dale P., Adoptive + Active Model: A New Approach to General Education, Arizona State University, 2018.

Taking Innovations to Scale

Despite the very impressive demonstration, High-Touch High-Tech learning is not always an easy sell. Innovation theorists stress that only one or two out of 1,000 good ideas are commercialized. Despite the odds, entrepreneurs are taking risks because they have strong incentives and an enabling environment to do so. Unlike an economic ecosystem where innovations have often been circulated and diffused rapidly, most innovations in education remain unable to be diffused and then disappear.

We can think about three possible channels for taking innovations in learning into scale.

1. The Education Workforce
Teacher colleges could help future teachers learn how to design and provide High-Touch High-Tech personalized learning for every student in the classroom, rather than passively deliver a set curriculum. Teachers could be trained to focus on students’ higher-order cognitive skills and socio-behavioral skills, skills that are much less easy to be supported by technology. This could help redefine and enhance the role of teachers and possibly the whole education workforce. At ASU, the key to the successful application of AI has been professors who are enabled to embrace adaptive learning technologies for High-Tech learning and at the same time to change their roles from delivering lectures to High-Touch learning. At the same time, they are supported by teaching assistants and instruction designers to help deliver High-Touch High-Tech learning.

2. Education Finance
Given the challenges for individuals to pay for adaptive learning tools, publicly financed initiatives will need to play an important role in delivering on their promise. Contrary to innovators in industry who have access to funds for their new initiatives from venture capitalists, accelerators, and angel investors, innovations in education often suffer from lack of sustained financial support. For example, in the Daegu school district – one of the most innovative school districts in Korea – the budget for innovation in learning including support for collaborative teaching, smart learning, and contests for innovative teaching as well as digital devices and school computer networks, amounts to only 0.3 percent of the total school district budget. This is strikingly small compared to the R&D expenditure of Korea, which is equivalent to 4.3 percent of GDP. This clearly shows that education finance is failing to support innovations in learning.

3. Innovation Ecosystem for Learning
Successful implementation requires a judicious mix between High-Tech and High-Touch learning, but the work of technology experts and those engaged in the teaching profession do not often overlap. An innovation ecosystem of collaboration should enable teachers, principals, professors, technology experts, business and community leaders, and policy makers to collaborate to design and implement the massively personalized learning model best suited for the context of a school, a district, or a country. This may require education systems to be much more decentralized.

Education’s Moment? Time for Global Action

At country level, taking innovations in learning to scale has often clashed with political constraints: changes in teacher education are much less visible than building new schools and, therefore, much less bankable during electoral campaigns. Nurturing the innovation ecosystem for learning requires a long-term commitment of national and community leaders.

The international community has an important role to play to fill the void and promote a longer term vision. Rather than advising developing countries to follow the traditional learning model, we should encourage countries to explore a new path for future learning. There is a need for global action to develop stronger evidence for new tools and practices to show the way forward. It is also key to connect people working at the cutting edge of technological innovation with key education actors around the world in order to bring education into the 4th Industrial Revolution of mass personalization. Governments, international organizations, and philanthropic foundations should join together to support innovators and key actors in education systems to co-design prototypes, test cases, and best practices and rigorously evaluate the outcomes before taking them to scale.

As an important first step, the Education Workforce Initiative (EWI) of the Education Commission is exploring the impact of technology on the roles of teachers and other members of the workforce. In the process, EWI is analyzing how High-Tech High-Touch applications are being introduced around the world. Building on this evidence, the Commission is testing the application of such approaches in different country contexts. For example, a collaborative project between ASU and the Ministry of Education and Training of Vietnam is developing a prototype for High-Touch High-Tech Learning for 7th grade mathematics. By demonstrating the impact of prototypes of High-Touch High-Tech Learning, the Commission hopes to stimulate and inspire actions for future learning in other countries – including high-income countries.

Financing challenges also need to be addressed. The costs of new technology to upgrade teacher training, school infrastructure, or learning are often outside the reach of public school systems. Inspired by the work of the GAVI Vaccine Alliance which helped reduce the costs of life-saving immunizations, diagnosis, and HIV medication, organizations such as the Global Business Coalition for Education (GBC- Ed) are mobilizing their network to negotiate similar agreements between technology companies and national education systems across the world. Through this work, a menu of reduced-rate, at-cost, and pro-bono technology resources for governments could be developed and supported by international organizations.

The Education Commission has highlighted the growing need for new ways to effectively implement educational change, including by building evidence on delivery approaches and reinforcing innovative ecosystems. One of its key motivations is to provide evidence that is useful to policy makers and leaders both at country and global levels as they formulate, adopt, and implement policies to improve education outcomes.

This is just a starting point and the Commission is looking forward to building a network with other organizations to turn research into action and create a Learning Generation.


Jo Ho Lee is a Commissioner with the International Commission on Financing Global Education Opportunity, Professor at the KDI School of Public Policy and Management and former Minister of Education, Korea.

Liesbet Steer is the Director of the International Commission on Financing Global Education Opportunity.

Back to listing