“Mapping Segregation” from the New York Times
Date published: July 8, 2015
“Mapping Segregation” is a visualization created by Matthew Bloch, Amanda Cox and Tom Giratikanon on July 8, 2015, and published on the New York Times.
In this visualization, each race is represented by a color. The color patterns are composed of thousands of colored dots. Every colored dot represents a certain amount of people in the same race, for example, in figure 1.1, each dot represents 120 people in the same race and, in figure 1.2 as we zoom-out more, each dot represents 12,000 people. The patterns formed by the colored dots illustrate who is living in which area. I believe the intended goal and purpose of this visualization are to break down the U.S. residents of a variety of places by race and reveal patterns of racial segregation in different regions in the United States.
There could be a variety of intended audience for this visualization. To begin with, under the name of this graph, it states “new government rules will require all cities and towns receiving federal housing funds to assess patterns of segregation”, so governments and government workers can understand and assess patterns of racial segregation through this visualization. Another group of audience could be students; this graph is a great tool to help students understand the issue of segregation in the U.S. using real-world data. The audience can also be business people and business migrants. For example, I’m a marketing agent from a Korean food company, and I’m trying to expand my business from Korea to the U.S.; this illustration can be helpful for me to locate where most Asian people live in the U.S. Researcher and scholars can also be a group of audience when they need information about patterns of racial segregation in their studies. Last but not least, this visualization is published on the “New York Times”, so the general public is also a big group of audience.
The graph contains the races and ethnicities of the U.S. residents, the patterns of how they are distributed in different regions, and the arrangments of racial and population density.
“Mapping Segregation” encodes 3 broad dimensions: race/census group(nominal data), region(nominal data) and population density/size (quantitative data). The encoding mappings have transformed both the nominal and quantitative information properly. In this graph, race and census group are encoded by color/hue that can be perceived easily and pre-attentively by human eyes. Different regions are encoded by position and area on the U.S. map. Population size as a type of quantitative data is mapped by the density of the dot where more concentrated dots means more people. How the author encoded the information is fairly appropriate and effective. By looking at the graph, we can immediately understand the patterns, for example, there are more Hispanic and Asian people living on the West coast.
The graph is successful in several vital tasks: it enables the comparison among different cities and regions when you click the city tabs(Figure 1.2), for example, by looking at Figure 1.3, the population in New York is denser than Los Angeles, and there are more Hispanic and Asian people living in Los Angele.
Moreover, when you move your mouse to a certain area on the graph, a small window will pop up and informs the viewer how the population is composed in the area selected, see Figure 1.4.
In addition, another meaningful task from this graph is that, as you continue to zoom out, instead of keeping the same amount of clustered dots, the number of people that each dot represents will increase and the number of dots will decrease, making the graph clean and easy to view the stories. For example, in Figure 1.2, one dot represents 12,000 people whereas one dot only represents 120 people in Figure 1.1. This is helpful when the reviewer is trying to evaluate the entire graph. For example, I can immediately conclude that larger cities in the U.S. tend to have more diverse populations.
This graph has three major strengths. First of all, the choices of color/hue are appropriate where the five colors are very distinctive from each other, making the overall pattern easy to perceive. Second, the variation among population density and racial diversity is apparent and straightforward to perceive because of the great choice of encoding mappings. Third, this visualization is visually appealing and very successful in communicating the intended stories and patterns to the viewer in a memorable and effective way that is hard to obtain with the original text-based census data.
On the other hand, there are a few weaknesses of this visualization when paying closer attention. Firstly, when you randomly click a place on the map, a window will pop up and show how the population is distributed in the area you clicked. On the top of the window, it says “Census tract” and a series of numbers. It is confusing that the graph didn’t explain what does the number represent. Secondly, the graph cannot be zoomed out to view the entire U.S; as a result, reviewers won’t be able to view the entire map (see Figure 1.2). Thirdly, the more clustered the dots, the denser the population, but it is hard for human eyes to perceive and compare density/how clustered the dots are. We can confidently conclude that there is definitely more people in New York than in Atlanta, but when two places seem to have similar population density, we can hardly tell which region has more people by eyes.
This visualization has served its intended purpose successfully except a few minor confusions analyzed in the previous sections. What worth mentioning is that most of the visual attribute that the authors adopted can be perceived by human eyes pre-attentively which provides insights and, at the same time, does not require excessive cognitive effort to understand.
There is only one improvement I want to make: instead of using clustered dots to represent population density, I would like to use “area”(dots in different sizes) to encode population size and density (for example, larger dots mean more people living in that area), since “area” are generally more accurate and effective than “density”.
I personally really like the design of this visualization, from the use of visual attributes to the messages that it conveys. I believe this is a well-designed visualization and can inform a wide range of audience.
References:
Bloch, M., Cox, A., & Giratikanon, T. (2015, July 8). Mapping Segregation. Retrieved October 22, 2019, from https://www.nytimes.com/interactive/2015/07/08/us/census-race-map.html?_r=1.
