Statistical learning tools: simple circles explain complex numbers

UN Chronicle,  Sept-Nov, 2004  by Nguyen Tang Le Huy Quoc-Benjamin

• E-mail
• Print
• Link

A professor of international health at the Karolinska Institute in Stockholm, Hans Rosling recognized an ironic situation while teaching young students in Sweden about health problems in rural Africa: while research and statistical data were, expectedly, more easily available to people in rich countries than in poor ones, he felt that those who did not suffer from such hurdles knew much less about their "neighbours" than those who did. "How can we take the numerical data that most countries and the United Nations have compiled and make it available to the public in a more appealing and understandable way?" Dr. Rosling asked. Feeling that it would be easier to understand and remember statistics through visual images, he envisioned a graph that could demonstrate the relationship between health and wealth of countries, symbolized by "bubbles", and their sizes proportional to population. He wanted to enable people to see instinctively where and in what conditions the majority of the world live.

[GRAPHIC OMITTED]

Rosling's idea was made feasible by one of his students who wrote a computer programme that could plot the statistical data into the graph. The professor was also intrigued by a question most often raised by his students: "How did they [the less developed countries] move up?" He thought that if the bubbles could be made mobile, his graph would better demonstrate the gradual development of each country. One day, his son Ola, then a college student in economic history and art at Gothenburg University, asked him if he could "borrow" and make use of the idea about a visual display that could demonstrate the growth of countries over time. Without hesitation Rosling agreed, and in the fall of 1998 the first prototype came out. He recalled how he, his son and the first viewers were struck by the amount of information they could absorb from these moving "bubbles"--fathoming the twenty-year development progress of the world in twenty seconds.

Having witnessed the impact of a simple prototype on audiences, the Roslings decided to promote this innovative idea and make it of benefit to everyone. They applied for funding from the Swedish Fund for IT (Information Technology) in Learning, but were turned down, and so they decided to use the family's own funds to improve the prototype. In the spring of 1999, an updated prototype plotted a world health chart to visualize international health development and enable better use of health data for learning, advocacy and hypothesis generation.

After finishing the initial design of the chart, Dr. Rosling presented it to the Statistics Department of the World Health Organization. WHO was willing to sponsor further development of the chart, whose support enabled the Roslings to complete their first full beta version. The WHO recognition also gave them valuable credence to apply for more funding from the Swedish International Development Agency, which eventually became the main sponsor of their other programmes.

In March 2003. Dr. Rosling met with staff of the UN Statistics Division of the United Nations Development Programme. With sufficient funding, the Roslings were able to produce a number of advanced prototypes also on other topics, ranging from education, health and income charts to the Millennium Development Goals achievement and human development trends. Speaking to the UN Chronicle, Dr. Rosling noted that the software was very cumbersome to develop and required the formation of a non-profit company consisting of programmers and graphic designers. The mission of the new company named Gapminder was to mind the gaps in the world.

Speaking to the Chronicle, Paul Cheung, head of the UN Statistics Division, expressed support for the idea of making the database of the United Nations more visible and accessible to the public: "We are always happy to see people use our data for any kind of analysis. But especially with his creative software, Prof. Rosling has made an impact on the international audience, and we are particularly pleased to see it [the software] become successful and widely accepted today." Mr. Cheung added that it was always the first priority of the Division to produce an accurate numerical database, but academia and research organizations were encouraged to make those statistics more tangible to the public through their own innovative devices.

Gapminder is undoubtedly among the pioneers in this enterprise. Visitors to its web site (www.gapminder.org), can view and download free software that "visualizes" human development. Dr. Rosling and his colleagues believe there has been a market failure in distributing global data: many people want to know more facts about the world but cannot afford the data. The advanced skills required to analyze the statistics cut down further the number of people who have access to data and are able to understand it. Thanks to the Internet and the concept of "global public goods", that market failure might be overcome.