This summary is a brief description of the most recent phase of my research into new approaches to data visualization. One approach, Information Recast: A Flexible Map, is part of my ongoing information visualization series, Information Recasting. One of the primary aspects of the work in this series is the design of interfaces that give users control over the visual presentation of data. In this project in particular, special emphasis is placed on creating information visualizations that convey compelling narratives and ideas about the data that would otherwise be silent, hidden away because the data set is too voluminous and multifaceted to consider in a single space in a single moment.

The Flexible Map software is a tool for analyzing 2004 United Nations data on global immigration. The squares on the screen at run time represent the countries of the globe arranged by latitude and longitude. Users can mouse over any square to see the name of a particular country. The moving dots represent the amount of immigration from a country of origin to a country of destination. In this case, there are nine destination countries. They are among the major immigrant destinations, countries like the United States, Australia and Canada. (see Figure 1 below)

Figure 1, Mouse rolled over country (box), reveals "Canada"

On the interface control panel are a variety of features that allow users to alter the way in which data is displayed. They can change the arrangement of the countries on the screen from a latitude and longitude configuration to a circular one. The color of the moving elements can be altered so users can highlight all the immigration traffic headed to specific destinations. The paths of the moving elements can be charted by trail markers. The color scheme of the entire viewing environment can be changed. The map can be viewed from three-dimensions, and so on. (see Figure 2 below).

Figure 2, Control Panel. (more below)

During testing it became clear that this degree of control over the display of data creates a highly engaging user experience. It also became clear that this high level of engagement encourages more extensive and much faster exploration of a data set than if the users were to look at the same data in spreadsheet format. Users commented on dozens of interesting observations that they made during short periods of use; but the most interesting piece of feedback was the following: “This device turns people who normally only look at data visualizations into users who make them.” The analytical skills of each viewer add to the value of the visualization. Figures 3 through 9 below demonstrate some of the ways that this data can be viewed.

Figure 3, The dots representing immigrants moving to Canada are changed to the color red, and trails indicating their individual pathways are added. In this way, the user quickly gets a sense of the overall traffic and trajectory of immigration to Canada. (more below)

Figure 4, The overall color scheme of the environment is altered to white. (more below)

Figure 5, All the countries are arranged in a circle around a single destination. The dots representing immigrants to Canada are still red. (more below)

Figure 6, The same circular arrangement in 3D perspective, tilted backward. (more below)

Figure 7, This is the same circular arrangement in 3D, tilted backward, but the user has zoomed in and changed some of the colors of immigrants going to other destinations. (more below)

Figure 8, The map is returned to the latitude and longitude configuration in 2D. The names of all the countries are switched on. (more below)

Figure 9, This is the same configuration in 2D with the country names switched on, with a population density map overlay. (more below)

A compelling data narrative, a story told by a data visualization because it is expressing an idea about the data that is greater than the sum of its parts, begins to emerge when the user selects the "GDP per capita" arrangement on the control panel. In that configuration the map is organized with longitude on the horizontal axis and GDP per capita along the vertical axis on the screen. (see Figure 10 below)

Figure 10, GDP as the y axis in place of latitude. (more below)

Conclusion

Whether generated by sensors, keystrokes or code, the amounts and types of information being recorded in all aspects of every sphere of activity continue to increase, as does our demand to retrieve and interpret the information being collected. Furthermore, the increasing ease with which processors can render complex graphics on relatively light-weight platforms makes it easier for designers and analysts to quickly develop visualizations. In short, appreciation for data visualization and how it allows users to interface with and present data is on the rise. The greatest results will occur when users are encouraged to explore data and create visualizations in environments where it is easy to do so rapidly. That many of these visualizations can be created by professionals without design and programming knowledge increases the likelihood that a given visualization will elegantly advance a complex argument about data, an argument that would otherwise get lost in massive spreadsheet tables.

Visit Information Recast: A Flexible Map

You might also be interested in:
Information Recasting presentation during thesis week at ITP
Information Recast: A Flexible Matrix Description and Demo
Buzzmap, created for presentation at Where 2.0 Conference in San Jose, 2007
Paul Kedrosky blogs A Flexible Map
Presentation Designer, Laura Fitton blogs A Flexible Map
Resume