Tujiza Uwituze worked hard and ranked near the top of her class in her Rwandan secondary school, but her education was poor by international standards. She had instructors who made her memorize and regurgitate information, and the school she attended had no computers for her to use. As a result, Uwituze's English is imperfect, and her computer skills are weak. She lives with a great-uncle in Kigali and has $75 in savings. Despite hard work and an intense desire to succeed, her dreams are out of reach—or might be if not for an innovative project that could radically change her life.

The goal of the experiment, called Kepler and conducted by a small nonprofit called Generation Rwanda, is to use massive open online courses (MOOCs) to deliver a top-tier education to young Rwandans who were born around the time of the 1994 genocide. The first test began in March with a “prepilot” class called Critical Thinking in Global Challenges, an online offering from the University of Edinburgh in Scotland. A dozen students viewed video lectures downloaded from a MOOC platform and attended small seminars and coaching sessions in a Kigali classroom with an on-site teaching fellow, a form of education called blended learning.

For a student like Uwituze, who was an infant during the slaughter by Hutus of some 800,000 Tutsis and their Hutu sympathizers in 1994, this was an extraordinary opportunity. Her family fled during the genocide: first to Burundi, then to Tanzania, then to Kenya. “We lost our money, our house, everything,” she says. Schooling was all that Uwituze really had after years on the run. She returned to Rwanda at the age of 14 and graduated from secondary school last November. The annual tuition at Rwandan public universities is about $1,500 a year for a substandard education, which is more than Uwituze's family can afford. Her mother is jobless, and Uwituze has three younger siblings who look to her for support. When she was turned down by an organization that helps aspiring Rwandan students get scholarships at American universities, an official from that group suggested she join Kepler. She became one of 15 students invited to attend the prepilot course to test the MOOC format. She then applied to be accepted into a larger class, which will be enrolled in a full MOOC curriculum in the fall.

Kepler received 2,696 applications for just 50 slots in the fall program. Six hundred students were invited to take an exam in April, of which 200, including Uwituze, made it to a final round of cuts. Those 200 were interviewed in person and took part in group activities observed by Kepler staff to gauge personality traits such as leadership qualities, ability to work well with others and problem-solving skills. The aim was to put together a class that combined a range of personality types: outgoing and shy, funny and earnest, creative and conscientious. The stakes were high. Jean Aime Mutabazi did not make the first cut for the fall session and felt adrift. Most of his male relatives, including his father, were killed in the genocide. He lives with his mother, who has a mangled leg and sells charcoal from a cement hutch to earn a living. “Can you imagine what it's like when you have a problem, and there is no one to turn to for help?” Mutabazi asks. “Education is a kind of magic power that can open any door in the world. If you are educated, you can control the situation you are living in.”

Uwituze made the final cut. She originally wanted to be an airline pilot but now considers that beyond reach, so she has settled on banking as a possible career. She will be able to study business and finance through Kepler. “Education is the only way I can survive,” she says, “the only way I can take care of my sisters, who need me.” Those accepted to the fall session will take online classes from leading universities for free—with support and mentoring from American teaching fellows in Kigali—and will also have their living expenses paid for. Generation Rwanda's executive director Jamie Hodari estimates that after an initial outlay of $100,000 for curriculum design and evaluation, the annual per-student tuition costs to his organization, including laptops and teachers' salaries, will be around $2,000. He hopes to bring that down over time to $1,000. Initially students will be on one track: toward an associate of arts degree in general studies, with a concentration in business, from Southern New Hampshire University, which has a cutting-edge program that awards degrees based on proved competencies, not the number of hours spent in a classroom. After associate degrees are completed in the second year, Kepler plans to offer bachelor's degrees in business administration, computer science and perhaps engineering from a variety of institutions.

[break]

Uwituze has questions about the online approach, including concerns that the experiment will falter or that her credentials will not be widely accepted, but she is confident that she will learn more with Kepler than in a traditional Rwandan university. “Most of the students here are very poor,” she says. “You don't have a choice when you get a chance like this.”

Where MOOCs Matter Most

Bringing the world's best college courses to some of the world's most disadvantaged people is certainly the hope—and some would say the hype—of the MOOC movement. Leaders of big MOOC platforms such as Udacity and Coursera, for-profit companies co-founded by Stanford University professors, and edX, the nonprofit platform jointly run by the Massachusetts Institute of Technology and Harvard University, are explicit in their ambition to knock down the class and geographical barriers to an advanced education. Coursera co-founder Daphne Koller laid out her world-changing goals in a TED lecture in June 2012 that was viewed more than a million times. MOOCs would “establish education as a fundamental human right, where anyone around the world with the ability and the motivation could get the skills that they need to make a better life for themselves, their families and their communities,” she told an enthusiastic audience. “Maybe the next Albert Einstein or the next Steve Jobs is living somewhere in a remote village in Africa, and if we could offer that person an education, they would be able to come up with the next big idea and make the world a better place for all of us.”

Nobody can argue with a goal like that. Yet educators who work in distance and online learning do argue that the MOOCs evangelists tend to oversell themselves and their product. They point out that online learning began well before MOOCs arrived and that MOOCs often do not incorporate the best, most up-to-date teaching methods. They also note that most of the developing world is not connected to the Internet and that MOOCs require skills and motivation possessed by only the very top students. “You have to find a solution that actually fits the reality of the Third World,” says Tony Bates, a Canadian consultant who specializes in online learning. “Yes, content will be free in the future, but what students really need is the kind of services instructors provide. How to study, where to find information, critical analysis, learning to have original ideas in what you do, discussion and high-level thinking: all have to be supported and developed by interaction with teachers.”

That is where an experiment like Kepler comes in: blending free content from the best professors in the world with low-cost instructors who can provide personalized help and prodding. The model particularly suits a country like Rwanda, where only a tiny fraction of the population has a college degree and the number of young graduates from secondary schools is soaring. “You could build 50 universities and not meet the rising demand for college here,” Hodari says. “There are people who don't go to college here, who, in an American context, would be going to Princeton.”

In the U.S., the fevered discussions about the potential of MOOCs mainly center on containing or reducing the soaring cost of a college degree. Koller noted in her TED lecture, for instance, that tuition at American colleges has risen at almost twice the rate of health care costs since 1985 and proposed MOOCs as a solution. Yet in the developing world, the much larger issue is quality. The facilities and level of instruction in many countries are pitiful, and college degrees often have negligible value to globally competitive employers. Rwandans who have taken computer-programming courses, for instance, often have little experience using computers. “It's like you have a degree in swimming but learned it in a book and never jumped in a pool,” says Michel Bézy, who is associate director at a small brick-and-mortar graduate program for Carnegie Mellon University in Kigali.

[break]

This is true not just in a tiny country like Rwanda but also in a giant, emerging power like India. The very best Indian universities produce excellent graduates, but the quality drops off precipitously from there. Much has been made, for instance, of the huge number of engineering graduates India produces. But of the 600,000 to 800,000 engineers who graduate in any given year, only “about 10 percent are getting a quality education,” says Ashok Jhunjhunwala, a professor of electrical engineering who is regarded as one of the country's high-level academics.

Testing has helped reveal specific shortcomings. “Only about 7 percent of computer-engineering graduates meet the industry standard for basic coding,” says Varun Aggarwal, co-founder and chief operating officer of Aspiring Minds, a company that conducts independent assessments of graduates on behalf of industries. A standardized “employability test” of 55,000 Indian engineering graduates in 2011 found that an alarming 42 percent could not multiply and divide numbers with decimals. More than a quarter did not have enough English to understand an engineering school curriculum. “It's a bit sad, but that's how it is,” Aggarwal says. “We're graduating a huge quantity, but the quality doesn't meet the mark.”

Part of the problem is poor instructors. “They are not well paid, and it's not seen as a very fashionable career to be in,” Aggarwal says. “Engineers who cannot get a job in industry go on to become teachers.” Another issue is the preparation of the students before they begin their higher education: many do not have proper English skills when they enter college or university, where English is the language of instruction.

If nothing changes, the situation is likely to worsen. As it is, India has one of the largest higher-education systems on the planet, with more than 600 universities and more than 33,000 colleges providing instruction to more than 20 million students. Yet the percentage of college-age Indians who pursue studies after secondary school is low compared with that of other countries. India's gross enrollment ratio is 17.9 percent, in contrast to China's 26.8 percent and the U.S.'s 94.8 percent. “To meet 50 percent gross enrollment, more than 30 million or 40 million new students will have to come into the higher-education system,” says Anand Sudarshan, former CEO of Manipal Global Education, which runs six universities and more than 40 other educational institutions. “That's not going to happen the way things are now. Technology-driven education is absolutely the only way India can hope to responsibly catch up, both in quality and in quantity.”

A Godsend

So are MOOCs the answer, in India and elsewhere?

For a small minority of exceptional students, MOOCs are a godsend. “We have a lot of people taking our courses as individual learners, and we get a lot of e-mails and other communications saying that the experience is changing their lives,” says Coursera's Koller. “There are many people in the developing world who would not otherwise have access to a top education. That can't be dismissed.”

Consider the case of Amol Bhave, a 17-year-old from Jabalpur, India, who was only 16 when he took a MOOC from M.I.T. called Circuits and Electronics. Bhave had been nosing around his father's engineering books since he was a child and taught himself BASIC. He got a Microsoft Certification in programming when he was still in secondary school. As a hobby, he also enjoyed electronics. He successfully completed Circuits and Electronics as a high school senior, and when edX failed to offer the follow-up course, Signals and Systems, Bhave was crestfallen. So he teamed up with two other students he met in online forums to create his own MOOC version of the course—based on M.I.T.'s videotaped lectures and online quizzes, along with other interactive elements created by Bhave. “It was my own code and everything from scratch,” Bhave says. Roughly 1,100 students took part in the course. Bhave cited this effort when he applied to M.I.T. as a full-time student this year. “On March 14 the admissions decisions were out. And guess what? I got in!” Bhave says. “My family and I were so excited. This was the first time anyone from my city was going to M.I.T. for undergraduate studies.”

[break]

The anecdote can be read two ways. One, an M.I.T. MOOC opened a fantastic window of opportunity for a young man living in central India, and two, MOOCs cannot fully replace traditional courses, as Bhave's highest aspiration was to fly to the U.S. and attend M.I.T. in person. The reasons Bhave craves the M.I.T. classroom experience are clear: for starters, it is difficult to study hard sciences, as Bhave would like to do, without being able to do hands-on research in a lab. More important, Bhave cannot get an M.I.T. degree from online studies, and a degree is key to pursuing a career.

The vast majority of Indians, in any case, may not have the drive or the smarts to do what Bhave did. He also had the benefit of coming from a family that was able to provide him with a computer and reliable Internet access. His father is an engineer and could afford to send Bhave to a private secondary school. Internet penetration is getting better in India, but it is still dismal: about 10 percent of Indians used the Internet in 2011. Reliable electricity remains a problem in much of the country, where the per capita annual income is less than $1,500. For hundreds of millions of Indians, a computer is an unimaginable luxury.

Yet technology is spreading and getting cheaper. Even a landlocked country like Rwanda is crisscrossed with fiber-optic cables and getting more wired by the year. Computing devices are becoming more affordable: at the behest of India, U.K.-based firm Datawind has produced a basic Android tablet for $40.11, and the Indian government has provided it to students for about half that price. The Aakash 2 cannot compete with top-of-the-line tablets. But for the price, it could be revolutionary. Datawind aims to ship a million of the tablets to India this year. “When I held that thing in my hand, I thought that this is going to change the world, and the world doesn't know it yet,” says Andrew Ng, who co-founded Coursera with Koller.

MOOCs are evolving, too. Michael Horn, co-founder of the Clayton Christensen Institute for Disruptive Innovation, a think tank that focuses on education and health care innovation, likens today's MOOCs to the earliest moving pictures. “The first movies were films of stage plays, and they look silly and absurd,” he says. “MOOCs today are essentially filming the stage plays. They are filming the lectures and splicing them up.” It is partly for this reason that less than 10 percent of people who sign up for MOOCs actually complete them. Horn expects online lectures to steadily become more engaging. The aim is to produce interactive courses that will not only teach students but also learn from them, so the courses will adapt and tailor themselves to individual skills and needs.

Some educators even envisage a broad decoupling of learning and assessment: a student who takes MOOCs and then gets tested and earns a competency-based degree at a high level could become more competitive in the job market than a graduate with a traditional degree from a brick-and-mortar university.

That is still in a hypothetical future, however. For now, students do not often see a tangible benefit to taking MOOCs: in the developing world, probably more so than elsewhere, youngsters need to have confidence that a particular education will lead to a job and a paycheck. “Students need to see a clear continuum between taking a course, getting a certificate for the course and having employers recognize that certificate,” Aggarwal says. “The courses need to conform to what the industry is looking for. If that virtuous cycle is built and is clear to the student, then MOOCs can scale.”

MOOC providers are aiming to award credentials that will be acceptable to colleges and employers, but that is still in the early stages. Part of the challenge is to develop safeguards against cheating that are seen as reliable. One method requires students to go to testing centers; other methods are technological. “We have something called Signature Track, where we ask someone to show a photo ID from the start, and when you do homework, you're asked to take a photo and provide a typing sample,” Ng says. “Your typing rhythm is hardened to your being. It is very difficult for you to type like I type or for me to type as you do.” The process is called keystroke biometrics, and it can be used to ensure that someone completing an assignment is the same person who signed up for the course.

[break]

Coursera is also working with an outfit called ProctorU, which monitors exams by Web cam. ProctorU asks test takers to show one or more forms of ID and to use their computer to scan the room to be sure no test aids are posted. Students may also have to fill out a multiple-choice quiz about their personal background—based on information gathered from public databases—to verify their identity. A ProctorU employee then watches students by webcam during the test. The process is more difficult to conduct overseas, but it is doable.

MOOC providers also see credentialing as a way to monetize their services. Coursera courses are free on an individual basis, for instance, but credentials are not. At the moment, if a student opts to take a Signature Track class from Duke University on Coursera, he or she will pay a fee of less than $100. After completing the course and passing exams, the student gets a “verified certificate” of completion with a Duke logo on it. For a handful of Coursera classes currently accredited by the American Council on Education, credit recommendations from that organization—which are accepted at many traditional institutions—are available. The cost is $100 to $190. Coursera also offers financial aid to participants who cannot afford those fees.

Yet such high-tech solutions on a broad scale still seem far-fetched for parts of the world where clean water and proper sanitation are in short supply. For now, the primary focus in India is to explore the use of MOOC technology to enhance the quality of teaching within existing institutions. Microsoft Research is working on a pilot project to develop online classes in the style of MOOCs, taught by leading Indian professors, which would fit the existing curriculum at Indian engineering schools. The program is called Massively Empowered Classrooms, or MEC. “There is no single answer for everybody,” says Jhunjhunwala, who believes that most Indian students, because of language and culture issues, would have trouble comprehending online courses offered by American universities. He recalls being utterly perplexed while taking a chemistry course in college because he could not understand the accent of his American professor. “Just taking something from outside and importing it doesn't work,” he says. “It has never worked.”

MOOC proponents counter that top-notch textbooks are used all over the world and that online classes can be seen as a kind of digital textbook. Courses can be designed for a wide variety of audiences. Europeans are developing their own MOOC platforms, and the big American MOOC providers are signing up foreign universities to provide courses in languages other than English. “Online education is still in its infancy now,” says Bhave, who brims with teenage idealism. “But it certainly has the potential to change the face of the developing world.” Based on his own experience, he observes, “there is an education revolution waiting to happen in the coming years.”

A High-Stakes Experiment

The educators at Kepler are not waiting. To them, the only workable model is clear: delivering the best instruction from the best teachers online but with significant hands-on support and classroom interaction. “The idea of just dropping MOOCs on Africans or others without facilitation and without assistance is a nonstarter,” Hodari says. “A lot of students haven't been taught how to use a computer. Really simple stuff is complicated: launching a program, closing a program, even typing.”

The first teaching fellow to join Kepler was Christine Yarng, a former teacher at a Knowledge Is Power Program (KIPP) charter school in Austin, Tex. Emma Stellman, the co-founder of a premier charter school in Cambridge, Mass., is designing the curriculum. Both are working for nonprofit wages in a remote corner of the globe because they believe that quality education can dramatically transform lives and because they have a hunger for challenge and adventure. Stellman aims to use chunks of various MOOCs and combine them in ways that are most suitable to her Rwandan students. The emphasis, at first, will be on learning how to learn—particularly in a digital context—and on quantitative analysis and critical thinking. “Employers are dying because they can't find people who can think for themselves,” Stellman says. “And the students become very proud when they realize they can have their own ideas. You see these lightbulbs go on: it's a beautiful phenomenon and very empowering.”

[break]

The Kepler prepilot class was conducted to identify problems and come up with solutions before the full program in the fall. By the fifth week Yarng and Stellman had learned several important lessons. They needed better Internet access, for starters, and so plan to move into new offices connected to Rwanda's fiber-optic network this summer. More important, they realized that many students needed significant English-language training to be able to follow online lectures and analyze complex subject matter. (In recent years Rwanda has switched from French to English as the primary language of instruction in schools.) Kepler now plans to conduct intensive English classes during an orientation period before the fall session begins and to assign a lot of writing in English during the regular term. “And when we're done writing,” Stellman says, “we'll write some more.”

Ideally, Hodari would like to expand Kepler in Rwanda and then export the model to other countries. That depends, however, on how the next two years go. “This is a pilot,” he says. “Our focus is on experimentation. There's been a lot of heavy breathing about changing the world but not a lot of experience in doing it. We want to spend two years testing the model—to see which pedagogical methods get the best student results.” Above all, Hodari wants to be sure the Kepler experiment succeeds, if only because 50 very hopeful Rwandans have no plan B.