In South Africa, perhaps not unlike in the rest of the world, most parents find themselves unable to afford school fees for their children. Yet a surprising observation that can readily be made is that most students do have access to a cellular phone and can make proficient use of the device. The explanation for this oddity isn't obvious, but the potential use of the cellular phone already in the pocket of many school-going children is an enticing prospect for those looking to improve the quality of education. This is especially true in a country where the school-leaving qualification program has a pass rate of only 60 percentthis with the definition of a pass as low as 30 percent for some subjects. The work reported in this article took place with students who attend schools where class sizes are 40 to 60 and whose parents are not able to cover the full cost of school fees and textbooks.
MXit (www.mxit.co.za) is a South Africabased instant message service designed to run on almost every cellular phone currently in use. The result is that most phones with at least GPRS and Java capabilities can run a version of MXit allowing the majority of cellular users access to a much cheaper chat platform than the already popular SMStypically, a character sent by MXit is one-thousandth the cost of a character sent by SMS. It is no surprise then that this platform is popular with children of all backgrounds: It allows them to chat with their friends instantly and cheaply.
However, MXit is not uniformly welcome in schools. There have been many sensationalist newspaper reports of MXit being used for underhanded purposes. These reports are reminiscent of the early days of the Web, when many negative stories circulated about the potential harm it would inflict upon our children. Like the Web before it, MXit has huge potential as a positive medium, and it is our responsibility as technology designers and researchers to leverage that potential.
Many researchers have already commented on how the value of a technology to a person will force them to overcome any hurdle as prosaic as a poorly designed interface. To use MXit, users have formidable tasks: The phone must be set up to use data services; the interface is crammed onto the small screen, and it is often impossible to see who sent the most recent message; there are handset inconsistencies between how text is entered for an SMS and how text is entered in the MXit client.
Despite these difficulties, MXit users have managed to learn and use it to achieve their overall goal of communicationwithin South Africa at the moment, the number of MXit users is greater than the total number of landlines installed in the entire country!
Like any new social network technology, MXit had a quiet period before it gained sufficient traction to make it viable as a communication tool. Viewed with frustration by adult users, teenagers were spurred by the value proposition and persevered to the point where almost every high school pupil in South Africa with an appropriate handset has MXit installed.
When we started investigating ICT interventions in South African high schools back in 2006, MXit was already highly popular and seemed like a good infrastructure around which to base our work. Primed by the press reports, we were worried about the abuses of a social network composed almost entirely of teenagers. However, when we started to interview the children, it became clear that their friend lists were made up almost entirely of friends from school. They treated those people whom they knew only from MXit with suspicion, and the children were much more guarded in their communication with these friends than they were with the people they knew in real life. Showing that teenagers are the same the world over (and across time periods, as well), most of their conversations on MXit were about parties or how their friends were doingjust a few hours after spending all day with them at school. Many children reported that their parents did not know how to use MXit, giving this medium an exclusive feel and opening the way for conversations that could be free from parental observation.
In order to further explore the potential of mobile ICT systems in education, the principle researcher started work as an IT teacher in a high school in South Africa. As part of this exploration process, he registered as a MXit user. Most of the students were surprised by his presence on MXit; some were threatened by what they considered to be an invasion of their space. This presents an interesting dilemma. On the one hand, we want to use a technology the students are familiar with, one they're excited about using and that fits in with their daily routine. But if our intervention leeches out their enjoyment of MXit, then we will have ruined our relationship with the students in destroying a technology they see as theirs.
As an initial, tentative intervention, extracurricular support sessions were offered via MXit. At first the students did not make use of this facility for asking question on how to do an assignments, instead they asked questions like "how are you?" and "can you get me this girl's number?"
After observing and interviewing the students about this behavior, it became clear that they were upset that not only was an adult invading their space but also that schoolwork was being pushed into their social space. Moreover, some students reported concern that their friends might notice they were communicating with their teacher, which is "uncool."
One unexpected observation was in relation to the latency in the MXit system. To overcome this, most children have multiple chat sessions running concurrently so that they have the potential to get a message from one of the available friendsdespite their legendary thumb-typing prowess, replies from friends could take more than 20 seconds to be delivered, which they deemed too long. However, some of the students felt that they must maintain total focus while chatting to the teacher about school. They felt that in talking to their teachers it meant that teachers should be on the top of their reply list, disrupting the way in which they normally communicated.
The realization that students were responding purely out of politeness, in a way that disrupted their normal use of the system, led us to explore other uses for MXit in education. One lead we pursued stemmed from the observation that MXit users experienced a lot of delay in using the system. In fact, depending on the particular child and the friends that were online, up to 50 percent of conversation time could be spent waiting. We therefore built a bot that would send information to any user who sent it a blank message"bot" is the term for any program participating in an online chat. The information related to the material being taught in class.
The children started to "talk" to the bot whenever they were waiting. As the bot could respond faster than any human, they could use it to fill in the gaps in conversation time left by their friends. Although the bot was clearly invading their social space with school information, the students still kept using it. It would seem that the relief from the tedium of waiting, which the bot provided, was sufficient to overcome the fact that it was sending nonsocial communications.
While it would be hard to prove any direct correlation between the students' receiving communication from the bot and their grades improving, the system grew in popularity. In fact, the bot started to be used by students from other schools. In the two weeks preceding the final examinations, there were more requests received than in the previous year. Clearly, the students perceived the bot service to be beneficial.
After the examination period, we interviewed a number of the students. Most of them felt that the service was useful and that they would not have been able to answer some of the examination questions without the information the bot provided. Whether or not this is strictly true is hard to say, but it was clear the bot had played some part in their motivation to learn. Having seen this initial bot, the students were able to make suggestions about services and features of future bots, which inspired us to take the intervention further.
The potential use of the cellular phone already in the pocket of many school-going children is an enticing prospect for those looking to improve the quality of education. This is especially true in a country where the school-leaving qualification program has a pass rate of only 60 percent.
The first new bot was the multiple-choice bot, which not only asked questions related to the subject but also provided feedback on both correct and incorrect answers. When the student answered correctly, he or she received further information expanding the concepts behind the question; an incorrect answer prompted an explanation of why the answer was wrong.
Following up from there, we decided to expand the service by introducing equation-solver, dictionary, and Wikipedia bots.
With the equation solver, students were able to type in a quadratic equation to get a step-by-step guide on how to solve that equation. The dictionary and Wikipedia bots are essentially reference services; both allow students to follow up on words or concepts they do not understand by using MXit as an interface to online services.
The bots lend themselves well to content-based subjects. At present we are using them in information technology and mathematics, but we have not attempted to introduce bots to support other subjects. Rather than attempt this translation ourselves, we have instead been working with other teachers to empower them to create their own bots to support the teaching of their subjects. Our hope is to create a suite of bots that will be accessible to anyone using MXit.
Ostensibly, this project is about using a mobile chat service (MXit) as an educational tool. However, as we develop and reflect on the project, we realize there are more general lessons for others trying to make effective ICT interventions in the developing world.
Conceptualizing the Internet. For most of our users, MXit is the Internet. They have no notion of sophisticated Web applications and browsers; if it isn't accessible through MXit, it does not exist. If you are trying to reach users with a new service, then we suggest you look at the ICT they are already familiar with and use that to give them a hook into the new system. It may be they use a system similar to MXit.
Local market forces. Within South Africa we have relatively cheap data plans that are accessible to pay-as-you-go customers as well as customers on contract. On current plans, users can access data at 8 cents/Mb. Most customers can therefore conduct all their communications for a few cents per day! It may be that other countries have similar low-cost data plans, or perhaps there are other low-cost ways of communicatingfor example, populating voice mailboxes with information in response to an SMS query.
The many handsets problem. Anyone who has tried to do a wide-scale deployment of mobile applications will know the headache involved in trying to write software that will run on any platform. Had we attempted an intervention in the school on our own, it would have been cheaper to buy the students identical handsets rather than invest the programmer time to create software for every handset the students have. Fortunately, we could leverage the MXit infrastructure already in place. Again, if MXit is not available, then it may be worth seeing how much you can achieve with, say, a WAP interface before going the painful route of creating your own client software.
Jakkaphan Tangkuampien is working on his Ph.D. in computer science at the University of Cape Town, South Africa under Gary Marsden. He also teaches the subject of information technology to high school students. His research interest is in exploring the potential of mobile devices in education in the developing world where computing devices are not as readily available.
©2009 ACM 1072-5220/09/0100 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2009 ACM, Inc.