Beth Kolko manages the Design for Digital Inclusion research group at the University of Washington, a group that includes undergrads, grads and faculty across fields, focusing on a wide variety of topics: technology in Central Asia, non-instrumental uses of techology, technology and autism, games for development, and other topics.
Uniting her work is a basic questions about technology use in different communities: What ICTs (information and community technologies) are adopted in diverse communities and why? The “what” gets very complicated in this question – a technology may be used very differently in one community than another. The overarching questions focus on what people in diverse communities do with ICT, and how can we design better technologies and policies?
There are lots of people in different academic fields looking at these questions – Beth points to user-centered design, learner-centered design, and value-sensitive design. Her goal is to contribute to how people think about design, and to incorporate more varied perspectives and functions, especially “discourse at the margins”. She argues that diversity lends robustness to environments, and helps you avoid designing brittle, fragile systems.
Beth’s field, information and communication technologies for development – ICT4D – focuses on technology in areas of resource constraint. She points out that these constraints aren’t static – they tend to change over time, sometimes even within a day. Constraints are also not purely geographic – Beth believes that some of the work she’s been doing in Central Asia has direct applicability for the Yakima Valley, for instance.
Most of her work relies on two sets of data, a yearly survey of 4,000 people in four Central Asian nations, with questions on media confidence as well as technical use. Only 11% of the survey respondents identify as Internet users, which gives a small sample for those questions, but it’s carefully demographically controlled and should be an accurate representation of the net-using population. She and her team also run qualitative, ethnographic studies based on interviews with teachers, businesspeope, and mobile internet users in Central Asia.
Her observations are organized into two categories, “form and function”:
Form – internet as weather dependent:
In nothern Cambodia, Beth wanted to interview customers at a local cybercafe. It was empty. Beth asked the proprietor why – it had just rained, and regular customers knew that the internet usually went down for a couple of hours during a rainstorm. “Wait two hours and come back – we’ll be full.” They were.
She argues that this complicates what is the internet – unlike in developed markets, it’s not a ubiquitous technology. It also forces us to rethink what an internet user is – “it’s often defined as someone who’s once touched the internet.” Understandings of internet users need to recognize that the net is not ubiquitous, and there’s a vast gap in how people use the internet, how often and how long they get to use it for.
Form – Internet as a shared public resource:
In Central Asia, internet use is primarily in shared spaces like cybercafes. It’s not always net use as we think of it. In a Kyrgyz cybercafe, we see pricing structures, which charge 2.5 times as much for web usage – downloading web pages – as for playing LAN games or chatting. The constraint is bandwidth, which is the cafe’s most expensive cost. Other behaviors, like Voice over IP, tend to be shaped by policy environments.
Form – Phones as banks:
Phones weren’t designed to be a banking device, but in countries like Kenya, they’ve become one. This raises questions about phones and security – people access their money, their email and other sensitive information on their phones, and often don’t set passwords. “How transparent is password protection on your phone?” Beth wonders.
Function – internet as strengthener of ties:
Beth and her team have been trying to do surveys of internet non-users as well as of users. Non-users generally offer three reasons why they’re not online: it’s difficult, it’s expensive, and there’s nothing to interest them online. She observes that everyone characterizes the Internet as appropriate for “the youth” – and that everyone defines the youth as themselves and younger, whether they’re 20 or 40 years old. Controlling for demographics, she and her team see a small but visible correlation between people with strong social networks and trust in those social networks and use of the internet.
Function – mobile phone as purveyor of fraud
It’s currently fashionable to look at the mobile phone as the platform of the future for the developing world. While Beth acknowledges the power of the mobile platform, she warns that it’s a dangerous platform – she points to a spam message she got in Kenya, promising her that she’d won a Safaricom contest. She returned the call, and her scammer proceeded to demand her passport number, which she refused. Her message – don’t assume that these platforms are always going to be used the way we’d want them to be used.
Function – SMS as weapon
With protests in Iran taking place as Beth gives her talk, it’s worth considering the ways SMS can be used – and is actually used – at moments of conflict. Analyzing the Lemon revolution in Kyrgystan, she explains that SMS wasn’t really used for planning demonstrations. Instead, it was used to warn people to stay away from riots, or to organize family and friends to protect businesses from looting. There is speculation – which she couldn’t confirm – that SMS was used by looters to coordinate their work. In Kenya, during election violence, SMS was a very effective platform for disseminating virulent ethnic hatred, using stories that were mostly true – and therefore credible – but had been exaggerated to be inflamatory.
Function – games as tech training
Since games are cheaper than the internet in many countries, people who aspire to modernity often gravitate to them first as an introduction to an online existence. As such, they end up being an alternative pathway towards technical training for users.
Based on these observations, Beth has tried to inform her design processes. This involves design ethnography – exposing aspects of a design to end-users and studying their reactions – as well as to modified design methods and prototyping. Her lab has worked on two main projects, MoSoSo and Starbus.
MoSoSo is a yellow pages – business directory – service for mobile phone users. It tries to take advantage of the realworld social network behaviors Beth observed in Central Asia to build trust in the system. It includes a general directory of businesses, where any user can rate the businesses. It also has a private, password-protected directory which users share with family and friends – the theory is that this closed directory will be believable and reliable in a way that the public directory might not be.
Starbus attempts to solve a common developing world problem – waiting for the bus. In Seattle, where Beth lives, there are great mobile phone and online tools designed to let riders figure out where the bus is and when it’s arriving. In developing nations, where transport tends to be adhoc and shared, there’s very little information on routes and timing. She and colleagues designed a tool called “the starbox” – a GPS receiver attached to a bus which communicates to a server via SMS – users can access the bus position on the server or query its location via SMS.
The system worked great in Seattle as she prototyped it, but failed utterly in Bishkek for the first several days. She and colleagues had made a design decision to use as little power as possible, to conseve battery life – a classic developing world strategy. Unfortunately, GSM towers in Bishkek weren’t as powerful as in Seattle – she needed to use more power to get the tool to work. Once they adjusted the signal strength, users of the system loved the new functionality and urged her to find a way to bring the system to Bishkek permanently.
One of the takeaways for me from Beth’s talk is the sheer complexity of reframing thinking about design in developing world environments. I’ve offered my set of suggestions for designing from constraints, which has some overlap with Beth’s observations, but not much. My guess – there’s no hard and fast rules for designing for constrained environments, just lots of practice, observation and careful information sharing…