GEOG 486
Cartography and Visualization

Part III: Visual Thinking

PrintPrint

Although cartographers have not focused their attention on designing maps to promote visual thinking until recently, visual thinking with maps is certainly not a new phenomenon. Two historic examples of scientists using maps to think are Dr. John Snow's cholera map of 1854 and Alfred Wegener's illustration of his theory of continental drift in 1915.

While investigating an outbreak of cholera in London, Dr. John Snow noticed that a large proportion of deaths occurred in households near the Broad Street water pump. He theorized that the water from this pump was the source of contamination and lobbied to have the pump handle removed so that people could not drink the water. Although it's unlikely that this action stopped the epidemic (it was already waning when the pump handle was removed), the map he later constructed to visualize where the cholera deaths occurred [See Figure 1.3.1] provided strong support for his hypothesis (i.e. confirmation).

 

A portion of John Snow's map of Cholera deaths in London
Figure 1.3.1 Portion of John Snow's (1854) cholera map in London.

 

Although Alfred Wegener was not the first person to notice that the coastlines of South America and Africa fit together remarkably well, after noticing their similar shapes on a map, he was inspired to seek out other geologic and paleontologic evidence that would support his hypothesis that the two land masses were once joined together. His ideas eventually matured into what we now know as the theory of continental drift, which he published in 1915 [See Figure 1.3.2].

 

Wegener's map, which illustrates his theory of continental drift.
Figure 1.3.2 Wegener's maps illustrating his theory of continental drift.
Credit: Wegener, 1922

 

In 1987, an NSF-sponsored report on Visualization in Scientific Computing that described the potential for using computer technologies to support scientific inquiry by making scientific data and concepts visible prompted David DiBiase (1990) to develop a functional model of visual methods that are used in geographic inquiry. In his model, DiBiase proposed that visual methods serve several functions in scientific research, ranging from acts of visual thinking performed in the private realm (exploration and confirmation) to acts of visual communication performed in the public realm (synthesis and presentation). MacEachren (1994) expanded this model of map use by adding another dimension: interaction [See Figure 1.3.3].

 

DiBiase's model of visual methods that are used in geogaphic inquiry.
Figure 1.3.3 (Cartography)3
Credit: MacEachren and Taylor, 1994

 

MacEachren introduced a new way of thinking about map use, which he called geographic visualization or geovisualization. Geographic visualization, as you can see from the figure, is a type of map use that can be described as: being performed in the private realm (in other words, the map is being manipulated for one's own goals, and not to be used for others); is focused on revealing unknowns in the data; and involves a high degree of human-map interaction (primarily through direct manipulation of the data being mapped, as opposed to solely mental interaction). Often maps used in a visualization context are also used in conjunction with other visual aids, such as statistical graphics. Visualization and communication are complementary aspects of map use; in other words, all maps can be used for communication and/or visualization. It is the map user that determines to what degree a map is used for one over the other.

Visual thinking is a cognitive process and is used in geographic visualization to produce insights on patterns, relationships, and/or anomalies in data. MacEachren (1995) presented a model of insight as a "seeing-that then reasoning-why" process, where the map reader "sees" features, patterns and relationships by combining the graphical marks on the page with mental categories she brings to the viewing process and then comparing what she sees with what she knows. This recent change of emphasis in cartography to include the use of maps for visual thinking as well as visual communication has prompted cartographers to consider whether the symbol-referent relationships that have been developed for communication purposes work equally well for all types of map users and map user tasks. This is still an open question in cartography, and one that you will need to consider when designing a map depending on your map purpose. We will explore this idea in more detail in Part IV: Map Purpose and Audience.

Figures 1.3.4 and 1.3.5 are examples of maps that can typify each end of the spectrum.

A portion of the Penn State main campus parking map.
Figure 1.3.4 Penn State parking map. Example of a special-purpose map used in a communication context.
Credit: Courtesy of Penn State
A screen capture showing an example of a geovisualization image. The image is described in the text below and shows how an uploaded dataset provides a map and other information visualizations.
Figure 1.3.5 GeoVISTA Studio. Example of a map used in a geovisualization context.
Credit: Courtesy of GeoVISTA Center, Penn State

 

The map in figure 1.3.4 shows a Penn State campus map with the purpose of visually communicating where a map reader can park. Most maps that an average person comes across from day to day are maps like this, which communicate information known by the cartographer to others (i.e. the public, which refers to someone other than oneself) with little opportunity for physical interaction that would reveal new information. Maps made and used for visual communication do not need to be of a specific type. For example, a thematic map showing disease incidence rates can be made for visual communication as much as a reference map or a map made for navigation. Any map made with known information to present to other people - no matter the representation - incorporates visual communication.

Figure 1.3.5 is showing a screen capture of GeoVISTA Studio, open source software for geovisualization created by the GeoVISTA Center at Penn State. The user loads a dataset into the software and is able to explore the phenomenon of interest visually with the map and other information visualizations. The dataset being shown in this image is of general health status indicators (e.g. cancer incidence, mortality rates, doctors available, smoking prevalence, etc.) broken down by gender, race and age, and is of the U.S. by county. The user interacts with the data by physically manipulating what is represented on the map, e.g. by selecting, highlighting or parsing variables of interest, segments of population, and/or geographic areas.  The user could explore clusters of high cancer rates, correlations between rates and behavior (e.g. smoking), rates and other variables like environmental contamination, or how different rates affect different populations in different places. The map use here is private (i.e. the map is made for the user alone), interactive (e.g. the user can change the data being looked at), and the map can potentially reveal unknowns.

As discussed earlier, whether a map is used more to visually communicate or visually think is up to the user. For example, a map user can take a map made for visual communication, e.g. a static reference map, and plot information on top of it with a pencil, essentially interacting with the map privately to potentially reveal new information to the user. Conversely, a user could take a map from a geovisualization tool, or interactive online map made for visual thinking, and present the image in a context to visually communicate information to others.

With technological innovations and access to high speed internet, we are seeing more examples of interactive maps available online. These interactive maps are representative of the upturn in designing maps for visual thinking. They still incorporate visually communicating data to the user, but allow for user interaction with the data to create private maps particular to the user's queries. Following are three examples of online interactive maps that both communicate data and provide opportunity for visual thinking through interaction.

Wellbeing Toronto is an online mapping tool containing city variables for Toronto on topics including socioeconomics, health, education, crime, environment and more. Users can look at variables individually or combine them into one visual variable, placing weights (for importance) on each. Several reference layers can be added for context as well, like parks, day care centers, health facilities, political boundaries, etc. 

A screen capture from World Freedom Atlas, an online mapping tool.
Figure 1.3.6 A screen capture from Wellbeing Toronto.

 

The Bike Share Map by Oliver O'Brien provides an interactive map of 100s of different bikesharing programs around the world. It permits users to explore individual programs, like Washington D.C.'s as shown in the figure below. Users can find out live information, such as how many bikes are available at each station, or how many open spaces are available.

A screen capture to show Washington, DC's Bike Share Map
Figure 1.3.7 A screen capture of Washington D.C.'s bike sharing stations from the Bike Share Map

270toWin is one of several election mapping tools available where the user can explore different scenarios with regard to an upcoming presidential election. Users are able to change outcomes shown on the map, in order to visually think about ways to get to 270 electoral votes.

A screen capture of the web site 270 to win
Figure 1.3.8 A screen capture from 270toWin, http://www.270towin.com/.

 

Recommended Readings

If you are interested in investigating this subject further, I recommend the following:

  • Snow, J. (1855). On the Mode of Communication of Cholera. London: John Churchill.
  • Wegener, A. (1966). The Origin of Continents and Oceans. Translated from the German by John Biram. New York: Dover.
  • DiBiase, D. (1990). "Visualization in the earth sciences." Earth and Mineral Sciences, Bulletin of the College of Earth and Mineral Sciences, Penn State University 59(2), p. 13-18.
  • MacEachren, A. M. and D. R. F. Taylor. (1994). Visualization in Modern Cartography. New York: Elsevier Science Inc.
  • MacEachren, A. M. (1995). How Maps Work. New York: Guilford Press.