Michael Best on Liberian storytelling

This post is from a MacArthur foundation sponsored workshop organized in Abuja, Nigeria to support Nigerian civil society organizations in using the internet.

Mike Best of Georgia Tech studies computing in post-conflict environments, countries recovering from civil war, genocide or other disasters. These countries are almost always low resource areas. In Liberia, where Mike has worked extensively, there’s no electric grid and no electricity, except via generators. There’s no telecommunications network other than wireless. And most of the people Mike works with were forced out of school by war and are illiterate.

We see a video of Samuel W. Togbeh, discussing his experiences during the Liberian civil war. He lives in Congotown, a poor neighborhood near Monrovia. While he’s had some primary school education, his schooling ended before fifth year and he’s not literate. When he was demobilized from the civil war, he did some agricultural training, but he’s never used a computer. He works as a goalkeeper for an amputee soccer team.

Thinking of Sam, Mike observes, “the more constraints, the more interesting your environment is,” from a design point of view. Building information systems for Sam requires thinking through power and connectivity issues and issues of literacy, but also dealing with a very difficult physical environment. Liberia has the wettest rainy season in Africa and a dry and dusty Harmattan season – these present unique challenges for deploying technology. And physical security is an issue – “if it’s not bolted down, it will be stolen.”

Within these constraints, Mike and his team developed a tool called MOSES – the Mobile Story Exchange System. MOSES allows participants to record and browse videos using an interface that uses pictures and speech, though no text. The system is portable and was moved throughout the country, tested in different areas. Mike’s team uses a model called HDF – heuristic evaluation, diaspora evaluation and field evaluation – to sharpen the designs. This allows a team based in Atlanta, Georgia to try and develop tools that can work in Liberia. One resource they used is a “cultural capsule”, a Liberian restaurant in Atlanta called “Mena’s Kitchen”, which allowed for partial field testing with expatriate Liberians in the US.

This testing helped Mike and team determine that videos couldn’t be organized in hierarchial trees, as they usually are in computer systems. The idea of drilling down to find content isn’t very intuitive and was hard for inexperienced users to understand. The final system lets users scan across the contents instead of forcing people to move down through a tree.

Testing in the diaspora doesn’t always reveal the way a tool will be received in the field. The team found a Liberian radio personality living in the diaspora. Diasporans thought he had a terrific voice and delivery and so the team wanted him to voice the system. But Sam, our friend from Congotown, couldn’t understand the radio man’s voice. Instead, the system was voiced by a Liberian computer technologist at Georgia Tech, Dickson Fuly.

The system has generated hundreds of videos, and thousands of Liberians have participated. Mike has been evaluating the impact of the process using a tool called “the Pre and Post General Self-Efficacy Scale”. This evaluation tool asks participants to evaluate their self-esteem. Mike has found 25% increases on this particular scale for people who used the MOSES system as compared to people who simply worked with rich media, but not in a way that focused on reconcilliation.

One surprise from the work was the effectiveness of “embodied conversational agents”, like the cartoon figure who led users through the MOSES system. While most Liberians had experience with video, as “video clubs” showing DVD movies are a very popular form of entertainment, few had encountered cartoons before. But cartoon figures turned out to be very easy for users to engage and identify with.

His work in low-resource environment has pointed Mike towards some interesting solutions. He points to some early work he’s done with voicemail and directory systems – these systems can deliver complex information to people who are not print literate. We see a teacher’s training college in rural Tanzania, where a computer lab is powered by biogas, fueled by the waste of 20 cows and some occasional diesel fuel. A computer lab in a Tanzanian refugee camp is built around solar panels, VSAT internet access and lightning arresters – these technologies collectively allow limited but functional internet access virtually anywhere.