Mobile and Immersive Learning for Literacy in Emerging Economies (MILLEE)

Poor literacy remains a decisive barrier to the economic empowerment of many people in the developing world. Although technology cannot replace learning through interactions with proficient language users, it can nevertheless provide learners – especially those who are disadvantaged and lack access to other educational avenues – with the foundation for further growth in listening, reading, writing and conversational skills. After more than two years of educational experiments in developing regions, a clear opportunity emerged: while the desktop computer is currently the platform of choice in practically any grassroots literacy program, mobile devices offer significantly more convenience due to inadequate building infrastructure and irregular electricity in rural areas. There is a more fundamental social argument: since a substantial fraction of children in developing regions have limited time to attend school regularly when they need to work for the family in the home or agricultural fields, learning in out-of-school settings made possible by mobile technologies can potentially increase access to literacy by at least an order of magnitude.

Informed by educational theories on language acquisition, we are designing a suite of mobile learning applications that target conversational skills, listening comprehension, phonetic decoding and sight reading. These applications will run on cellphones, the fastest growing technology platform in emerging economies. Learners in these regions are usually skeptical of formal education, and as such, some applications make use of the built-in camera so that students can use and articulate language anchored in local scenes. The suite will include subtitled videos based on vector graphics (XML-based SMIL) and interactive games, some of which will leverage Bluetooth or infrared to promote the group-play experiences prevalent in everyday village life. Other applications will employ speech output and artificial intelligence techniques for conversational practice.

To address local learning needs, it is essential that our applications are easy to adapt for a new learner community by reusing existing modules as much as possible. Worse, in developing regions, the learner’s educational baseline can deviate sharply from the official syllabus, and learners in the same classroom or baseline may belong to a relatively wide age bracket. In a peer-reviewed paper, we presented a framework called PACE (Pattern-Activity-Curriculum-Exercise) that supports the rapid, scalable development of language learning software localized for a particular community. The key insight is to represent pedagogy, software and curricula at separate levels so that an application can be modified at one level without affecting its overall correctness or behavior. In an initial pilot, designing our applications according to PACE provided the flexibility to adapt our prototypes for more advanced learning material, alternative teaching philosophies and local cultural conventions without incurring significant localization costs. The same applications were also enthusiastically received by children and demonstrated some positive learning outcomes.

Project mission:

To enhance access to literacy among children of school-going age in the developing world. In particular, we aim to complement the formal schooling system by applying mobile learning technology to augment educational opportunities in out-of-school settings.

In order to make the literacy challenge more tractable, we are currently focusing on English language literacy in rural India. But we expect our lessons to be applicable to other languages and to other developing regions throughout the world.

English is widely seen as a key to socioeconomic success in India. English is taught in almost all schools: as a second language in public schools, and as a first language and the medium of instruction in most private schools. Fluency in English can almost be equated with membership in the middle and upper classes. A recent article states that mastery of English is the “single most influential factor that determines access to elite educational institutions, and hence to important avenues of economic and social advancement.” Language competency is also the biggest barrier to technology empowerment, e.g. 90% of the indigenous web content in India is in English.

Potential impact:

Generalizing beyond India, English is a global language used by at least 1.5 billion people in more than 170 countries. Fluency in English opens the door to continuing education where English is the medium of instruction, “new economy” outsourced jobs that involve call centers, basic medical transcription, data entry, etc. – even in business process outsourcing centers located in rural areas – and better-paying occupations with multinational corporations or the civil service, and often improves access to government, health and legal services. These benefits also foster a stronger sense of self-esteem and social status. Thus our current focus is on English as a Second Language (ESL), even though we believe that many of the lessons will transfer to other languages.

Unfortunately, the school systems in underdeveloped regions face many difficulties, especially with ESL. Two significant obstacles stand out: irregular school attendance due to the need for students to work in the fields, homes, etc., and disinterest in schoolwork due to the perceived opportunity costs or lack of benefits of formal schooling. Another factor is the preparation of local ESL teachers, who often cannot communicate with us without the help of interpreters. We believe that ESL learning games on cellphones can address the above challenges. In particular, we hypothesize that learners can improve their ESL skills by using mobile devices in out-of-school settings. We also believe that game-like design can improve enjoyment of the learning experience and encourage spontaneous adoption.

Technology-assisted learning in developing regions is not far-fetched. At least two non-government organizations, Pratham and the Azim Premji Foundation, have used computer games in their initiatives for children in the urban slums and rural areas of India respectively. Most importantly, a large-scale evaluation by Pratham showed significant gains on math test scores from playing computer games that target math learning. It is plausible that similar learning outcomes can be replicated using mobile games for ESL.

Required resources:

1. Total project needs: $400,000
2. In-kind contribution needs: airline miles, laptops with built-in DVD burner drives and Firewire ports, video cameras, MiniDV tapes for video cameras, DVD-Rs, educational materials for English learners, office supplies
3. Other:

   The five field studies that we have conducted so far have involved usability studies and learning assessments that were restricted to two weeks for each feasibility study. These limitations were acceptable when we were going through the “due diligence” process, but they made it impossible for us to evaluate long-term retention and other complex learning metrics, nor observe the emergent social practices that evolve around the mobile learning technology when learners have prolonged access to the technology. Hence, we are seeking funding for application development, localized content creation, international travel, local staff salaries, etc. that will sustain a longitudinal deployment lasting one year. We hope to commence this deployment in the summer of 2007.

   In addition to making a credible longitudinal evaluation possible, the donations will enhance the educational experience of our undergraduate researchers. We have a track record of undergraduate mentoring and have worked with approx. 30 undergraduate researchers so far. On average, there are 5-8 active undergraduate researchers in the MILLEE team at any time. We would like to support the personal development of the most promising undergraduate researchers, such as by giving them the opportunity to conduct fieldwork in India. We believe that practical experience in the field combined with exposure to a different cultural and socioeconomic environment is an invaluable way of nurturing the next generation of outstanding graduate students and industry leaders.

Our sponsors:

Our project is primarily funded by the National Science Foundation (USA) under Grant No. 0326582. We have received initial support to commence a longitudinal pilot evaluation through a BREW Wireless Reach prize award from Qualcomm Inc. The speech recognition component is supported by a Microsoft Research Digital Inclusion Award. Other sources of financial support include the Intel Undergraduate Research Program, Seeqpod Inc. and a prize award from the Bears Breaking Boundaries 2006 competition (serious games category). We thank Sony Creative Software for their sponsorship-in-kind.

Project page URL:

• http://www.cs.berkeley.edu/~mattkam/millee
• http://tier.cs.berkeley.edu

Additional info:

Literacy levels are difficult to gauge accurately in many developing countries. Official reports for India put it at 60% of children, but English literacy is much lower. Learning English in schools in India is difficult in spite of a national program to teach it for reasons outlined above. About half of the children in rural areas and the slums in India have difficulty attending school regularly. We are primarily targeting that fraction of the population. Our applications are usable by a large fraction of school-age children, which is a potential market of about 160 million. However, it is likely that a game-like ESL solution would be attractive to the other 50% who wish to improve their English skills. Children are not direct consumers of cell phones (or application programs), but parents, no less in developing countries, are passionate about providing opportunities for their children. The cell phone market in India is growing very fast (about 20% annually), and phones will be accessible to a significant fraction of poor children. We expect that it will help drive handset sales to a small extent, and encourage families to purchase additional handsets for their children while sharing airtime minutes.

Over the next few years, we will concentrate on ESL in India. However, the effects of world-language literacy appear to be very similar in other parts of the world. Huge nations like India and China are really federations of culturally diverse provinces. Many cultural groups are not fluent in the official national language, and indigenous people are often not literate even in the regional dialect. These groups are inevitably the most economically disadvantaged. And we have seen very similar “illiteracy cycles” in many parts of the world, just as in India. We call it a “cycle” because it is reproduced by the school system – the schools are poor and there are no fluent world-language teachers. Students graduating from these schools are grossly under-represented in the higher education system and are particularly poor in world-language competence. They rarely if ever receive formal teaching credentials in world language teaching. So the pipeline of teachers never grows, and language learning never improves.

There are significant opportunities for the adoption of our solution in both the private and public sectors. In the public sector, English language learning is already mandated in most Indian states and significant resources are spent on it. If we are able to show some success for our approach in the field, especially compared to in-school English learning (which is widely known to be quite ineffective) there would be a very good case for uptake as part of the education budget. Success of an ESL learning pilot may also encourage states to experiment with migrating other parts of the school curriculum to cellphones. This could range from a handful of assignments to large parts of the school curriculum for the many students who for practical purposes are unable to attend school. States may choose to subsidize phones or the applications (including airtime charges) or both.

Private sector adoption is harder to predict – the technology marketplace in developing regions is eclectic and quite different from wealthier countries. But there are some promising possibilities we already know about. Case in point: an innovative pilot project to impart skills in English language and computing at a Rural Community Learning Center. A local entrepreneur was “loaned” five computers. He has invested in the required infrastructure to set up a computer-equipped training center in his village, near Mysore (Karnataka, India). Local residents attend classes everyday on computing skills, and on ESL literacy during weekends. Furthermore, this center assumes the role of a community library that loans books to local residents to improve their reading skills. The latter is especially significant because the dearth of reading material is a major factor for low levels of literacy in underdeveloped regions, even as free voluntary reading (i.e. whole language instruction) is argued to be the most effective means of improving vocabulary and reading comprehension skills.

This model can be extended easily to cellphones and phone-based software. A regional center may loan cellphones, and provide technical support and language tutoring on the content. Software might be accessed either over-the-air, or through a flash card or cable download for those users able to reach a community center. These centers often provide an easier “on-ramp” to users who are unsure about owning a phone themselves, or paying for the software.

Cellphones are increasingly being adopted in developing countries. In India, the number of people using cell phones has risen from 10 million in 2002 to more than 140 million at the end of November 2006. By 2010, the government of India expects half a billion citizens to be using mobile phones. The world's fourth-largest cellphone manufacturer, Sony Ericsson, will be having some of its mobile telephones produced in India in order to profit from the booming market on the subcontinent. The World Bank notes that the infrastructure is already in place in most parts of the world, as 77% of the world’s population lives within range of a mobile network. Manufacturers are also driving down the cost of cellphones. For instance, as part of the GSM Association (GSMA) initiative to make mobile telephony affordable in developing countries, Motorola devised a stripped-down design that drove wholesale costs of cellphones down to under US$30. In Africa, there has been a sharp increase in cellphone adoption – there were 51.8 million mobile subscribers at the end of 2003, which is an increase of 1,000% since five years ago. Furthermore, this figure is an understatement because many mobile handsets and service subscriptions are shared between several users. The citizens of low- and middle-income countries now make up more than 20% of the world’s cellphone users. Gartner Research projects that there will be 2.6 billion cellphones in use worldwide by the end of 2009.