Barefoot In the head

If you think so....

My Photo
Location: Newcastle upon Tyne, Tyne and Wear, United Kingdom

I am around if you are around.

Monday, February 26, 2007

Self organised learning

Girls at a hole-in-the-wall computer in village D.Salhundi, Karnatak, India

Sunday, February 25, 2007

SOS for UPE - Self Organising Systems for Universal Education

Schools, education, or the lack thereof

A world with Universal Primary Education (UPE) would be one where every child is educated. In order for that to happen it is assumed that every child would go to school. The current target year for this to happen, according to the UN, is 2015. Because “school” is equal to “education”. While this may be undoubtedly true, many may invert this equation to - “no school” is equal to “no education”. I think we have a problem with this second equation. Education does not happen only through schools. Learning certainly does not. This article is about the problems of schools for UPE and about some alternative learning scenarios.

Let’s look at schools first. I used to live in India and will use Indian numbers for my examples in the hope that these numbers will correctly depict the situation in other countries, when scaled up or down appropriately.

There are over 600,000 schools in India. While this is a large number, the population of India (currently 1.2 billion) is large enough to overcome it. There are 122 million children in India who do not go to school. In order to have the right number of schools for all children in India to go to school, we would need to triple the number of schools and the number of available teachers. There is not enough money to do this, nor enough time, or management capability or institutions. The problem is just too large to solve with traditional, linear methods of scaling up. 122 million children growing up without an education spells trouble in the 21st century. Trouble with a capital T.

Let us, for a moment, assume that we could, somehow, build the necessary number of schools and train the necessary number of teachers, in time. We would suppose that all the schools we have built have a uniform level of quality, performance and efficiency and then, I am afraid, run into a quality of education problem. And that one is even harder to solve than the numbers problem. Here is how the argument goes.

Schools in remote areas do not have good enough:
Teachers: because good teachers tend to migrate to urban schools for better salaries and standards of living
Retention of teachers: The occasional good teacher stays for a while before moving to urban schools
Infrastructure: Local infrastructure in remote or rural areas are constrained by the size and economics of the market available to them. There is not as much competition and variety as would be in an urban area.
Maintenance of infrastructure: It is expensive to maintain infrastructure in remote areas because such maintenance would usually come from the nearest urban area. This also results in a higher mean-time-to-repair (MTTR).

In other words, the farther a school is from an urban area, the worse off it is. The quality of education from a remote school will usually be less than that from its urban cousin. This is a human problem, not an economic one. After all, cities were built for comfort, safety, convenience and so on. So, people want to live in cities. Surely there are exceptions, but most people anywhere in the world tend to congregate towards cities and their suburbs. In the developing world, where rural and remote infrastructure is weak or non-existent, the problem is all the more acute.

A recent survey in India (Annual Status of Education Report for Rural India 2005) shows that 51.9% of children aged 7-14 cannot read grade 2 texts and about 65.5% of these children cannot perform simple arithmetic operations.

Looks like UPE is caught in a bizarre stalemate. Not all the money in the world will correct the problem of lower quality education in remote schools. Schools with absent teachers, no teachers, wicked teachers, sick teachers are abundant throughout Africa, India, South America and other large parts of the world. The more money we spend on teacher development, the more they move to the cities. Without good teachers and administrators, the schools crumble and break. Children remain in schools, only on official records.

The developed world is somewhat better off because their remote areas often have infrastructural facilities comparable to their cities. But only to a limited extent can that attract good teachers. A school in New Jersey will always have a wider choice of teachers, equipment and infrastructure than a school in Alaska, or even Newfoundland. The problem of quality will remain.

It is in this context that we need to look for alternative methods for primary education that are relatively independent of teachers and infrastructure. Educational technology (ET) has traditionally been piloted and introduced first in good schools in urban areas. Since these areas have good teachers, the performance of students is, in any case, good. Hence, not much improvement is noticed in learning, and the role of ET has often been questioned. Many teachers consider the use of technology in education as an over-hyped and under-performing idea.

In view of this, I think that the most advanced educational technology should go first to the most disadvantaged learners. Any advantage to such a learner would be a benefit. For example, a learner in an urban school with good teachers may score 80% in some test of performance. A learner in a remote school with poor resources may score 30% (failed) in the same test. If the introduction of educational technology the urban learner increases the score to 90%, it may be considered too expensive for the value it provides. On the other hand if the same technology were to increase the scores for the disadvantaged learner to 40% (passed), it may be considered vital and very good value for money. A little improvement, at the “bottom of the pyramid” affects larger numbers, permanently.

For technology to be introduced first in rural and remote areas, it must be designed such that it is easy to transport and requires little maintenance.

The “Hole in the wall” experiments

Groups of children can learn to use computers on their own, irrespective of who or where they are.

Groups of children, given access to shared, publicly accessible computers in playgrounds and other public areas, will teach themselves to use the technology on their own.

We found this through a set of experiments conducted from 1999 onwards and often referred to as the “Hole-in-the-wall” experiments.

We found that children given unsupervised access to computers in public or play areas would become:

1. Computer literate – in their own way, with their own vocabulary, but highly effective nevertheless.
2. Better at math and English – I don’t know why, maybe because they learn to analyze and solve problems in groups.
3. More social and cooperative – because they learn that knowledge, unlike material objects, grows with sharing.
4. More interested in school – if the computer is near or in the school premises.
5. Less likely to drop out of school – because they want their computer.
6. Less interested in petty crime – mostly because all their free time is spent at the computer
7. Generate local goodwill – parents like the idea that the child is learning something and not creating trouble at home.

It took us five years of rigorous measurements across the Indian subcontinent to verify these results amongst 40,000 of the world’s poorest children. Almost half of them, girls.

The data based outcomes showed:

Acquisition of functional computer literacy
Improvement in academic performance
Increase in confidence and self-esteem
Increased collaborative behavior

Apart from data-based findings, there is consistent anecdotal evidence of large-scale impact on school enrollment, retention, concentration, attention span and problem-solving ability.

To keep computers working in, mostly, outdoor environments, we had to design several pieces of hardware and software. In five years a design emerged that is reliable and low on maintenance. The design is resistant to vandalism and undesirable adult access. Interestingly, both vandalism and adult access is automatically low in public places where children are present. We were also able to design software to remotely monitor all activity at these “playground” computers.

We found much more effective use of the computers already owned by schools—200 children can become computer literate using one playground computer—making it an effective and affordable method.

Without adult intervention or supervision, 40,000 village children experimented with computers and software to acquire an enduring understanding of the information age.

How it Works

Computers are made available to children in shared, public spaces, free of charge and no structure is imposed on when, how or what they learn.

Shared outdoor public computers, preferably connected to the Internet, incorporating self-protective hardware and software are combined with voluntary group self-learning by all the children of a given community, whether in or out of school. In five or six years’ time, the oldest of these children, now 13 or 14, will be the first computer-literate adult generation of their communities.

The computers, typically located in a government school playground or in similar safe, public areas, are unsupervised and are available to the children at least eight hours a day.

Working in self-organized groups and helping each other, the children typically navigate within minutes and begin to browse in about an hour. Within three months they achieve basic computer literacy, and by nine months have achieved the proficiency level equivalent to the skills of most modern office workers. They also pick up a considerable amount of the English language from common multimedia software.

Educational and other games and content tested with other children of the same age group provides the “minimally invasive” educational input that causes change in educational outcomes.

Several types of school-related content and links are also provided to help with schoolwork. Teachers have been very positive about the whole thing because of the children’s increased interest in learning, higher enrollments, and their concentration on higher level tasks like mentoring students and leading class discussions.

What would it cost the World?

I cannot resist doing this calculation. It is an exercise in fantasy, but one that you might enjoy!

The Earth’s current population: 6 billion
No. of children below 15: 1 billion – usually 15% of the total population
No. of “hole in the wall” computers required: 50 million playground computers, 200 children share each
Cost of computers and related infrastructure: USD 60 billion, each facility with three computers costs USD 1200 to build
Total other costs: USD 6 billion to take care of contingencies in varied terrain
Recurring costs: USD 6 billion per year (that is, less than 2 US cents per child per day)
Time to make operational: Could be done in less than 10 years if an International Agency gets the act together.

Result: No child left behind. Irreversible social change in 15 years as one generation grows up.


Gender issues

It is important to mention here that these results are obtained when the computers are placed in a safe, public location, such as a playground. The same computers, when placed in a room inside the school do not produce the same results. The children do not perceive computers in schools as their property and playthings. They suspect some hidden agenda and are afraid of some “catch” in using them. Girls tend to avoid going into closed rooms with boys, unless there is some adult supervisor.

In most outdoor locations, the number of girls and boys are about equal. However, there are some locations, particularly in slums and very poor areas, where there are very few girl users. Ensuring and perceiving safety is important for attracting girls to public computers.

While the girls learn as quickly as the boys, their activity patterns tend to be different. Girls are more practical, they will learn those things that they can use immediately. While boys will experiment, for example, with a lasso (irregular cutting) tool in a painting program, girls will wait until the boys have learnt the tool, then learn it from them and begin using the tool immediately to create drawings.

While both boys and girls use games as the most frequent application, the type of games played differ with both age and gender.

Very young boys and girls play the same games. These are usually simple clicking games such as catching a ball with a net, or playing a sound by clicking on a picture.

In children over 10, boys tend to concentrate on “action” oriented games such as racing and pinball, while girls seem to be interested in more conceptual games. Girls would play chess or games with arithmetic more than boys.

Self organizing pedagogy

Minimally Invasive Education (or MIE, as we call it) was developed by observing and analyzing natural collaborative learning of computer skills among Indian children in 12 ethnically and linguistically different states, from Ladakh to Tamil Nadu (3000 Kilometres apart, north to south), and from Rajasthan to West Bengal (also 3000 Kilometres apart, west to east). Having isolated the common cross-cultural factors in learning, the researchers then focused on enhancing them in ways that were “minimally invasive”, or largely invisible to the children, such as testing and selecting the most effective content, allowing some children to observe a technician performing a task so that they would later teach the others, and occasionally by developing animated tutorial software in which a cartoon character coaches children through a particularly difficult task, such as signing up for an e-mail account. Throughout, all instruction by adults and older youngsters was rigorously avoided, to prevent the children becoming dependent on scarce and expensive resources. Participating teachers were encouraged to assign tasks to be performed at the computers, and to lead class discussions about computer learning, but were asked to avoid direct instruction. As a result, “child teachers” emerged at each of the experimental sites—typically, talented 6-8 year old boys and girls who took on the “teacher” role and taught 3 or 4 “generations” of children to use the computers. In effect, every learner was a teacher and vice-versa.

Highly interconnected systems, such as the Internet, or biological structures such as us, show self-organisation. This field of study, called Complex Systems is mostly studied by Physicists and is relatively new and unexplored. It seems to contain the answers to how life, cognition and consciousness work. Self organizing systems show what is called “emergent behaviour”, that is, behaviour that was not programmed or expected from it. Small changes in the inputs to such systems can cause disproportionately large changes in its output. It seems to me that these are properties that could be extremely desirable for mass education. Could there be a pedagogy based on self organization?

The “hole in the wall” experiments seem to suggest an alternative, inexpensive and reliable method for bringing computer literacy and primary education to those areas where conventional schools are not functional.

Such facilities are not meant to replace schools and teachers, they are meant to supplement, complement and stand-by for those areas of the earth where good schools and good teachers are, for whatever reason, absent.