Design

Mapping Distributions

For this lab assignment, I chose to depict the Maternal Mortality Ratio in the Middle East. I chose this group of countries to depict a statistic concerning women’s health as it is a region that most people associate with a large gender imbalance and I thought this assignment would be a way to investigate this stereotype. First, I toyed with the idea of using different classification methods but most of the other methods were more appropriate for data that was either linear or fit a curve, which my data does not. So, natural breaks (jenks) were used for all three maps.


The first method of presentation is in terms of a color gradient, like in the census lab. This is a very straightforward presentation, appropriate if you are attempting to illuminate the data for more objective and less pedagogical reasons- a simple presentation of the facts.


The second method of presentation uses dot density and is probably the least appropriate for this data set. Using a physical unit on the map to denote a unit or amount of units is not appropriate when the country size varies so differently. It also might give one an incorrect idea that the data actually has that geographic shape when it is really the work of the computer.


The third method of presentation uses graduated symbols. This is my favorite way to present the data, as we can instantly get a sense of the scale of the differences between countries. The amount of deaths in Afghanistan dominates the map and drives that statistic home, something that is lost in the first map. It seems like a good compromise between the two other sets of data- visually striking but still factual.

nytimes.com Election Map Replica

My changes to the map were straightforward (I mean really, it’s the New York Times here and they know what they’re doing) but I do think that they improve upon the original map. Things like enlarging Alaska so that it is closer to its real area, darkening and enlarging the popular vote so that it is more prominent (as it is a valuable statistic), and labeling the states on the Eastern seaboard add valuable information in a simple way that I thought the original had not done an adequate job of conveying to the reader. I added pictures of the candidates, their full names, title in government and their parties which I think adds to the content of the map substantially- this is after all an important election in terms of race and adding pictures of the candidates makes visual reference to that. I thought formalizing their names and titles added a serious element that seems appropriate for the Times. In addition to these simple additions, I denoted the key battleground states vis a a vie yellow hatching and borders, hoping to draw attention to the states where the races would be closest and could ultimately be most important in deciding the election. In addition, I made more explicit the unique electoral vote splitting situation in Nebraska by also using a read and blue hatching fill and asterisk. Now, I’m assuming that this map improvement could still be capable of the same interactive features that it had online- showing previous elections, state by state break downs, etc. Adding more detailed voter information on the map template seemed like it would be too visually cluttered; the point of the map is to show which states went for which candidates- not to visually convey all available statistical information. If this were not the case then I might have added more information about states that went from red to blue or blue to red from the last election, or made a county by county breakdown and a red and blue scale. I hope my assumption is right and my improvements to the nytimes.com map are considered as such.

First Lab- 2000 US Census Project

These three thematic maps characterize three separate racial categories in their percentages (relative to the total county population) as obtained from the 2000 US Census. In descending order, they show the percentages of people who identified as White alone, Native American or Alaskan Native alone, and those who identified as two or more races. Classifications were based on natural breaks or Jenks and more classes were added to highlight variation.



In the map for whites, some interesting patterns appear. The lightest green denotes areas with the highest percentage of whites, the darkest purple those who chose this category the least. The southwest, west, and eastern seaboard are generally purpler than the interior of the country, though many counties show a surprisingly deep hue of purple- especially in states like Nebraska and South Dakota. Also, in the South where there is surely a high black population there is more purple, (though again a surprising amount to me). Arizona also has an interestingly deep purple part of the state.



When we look at the map of the Native American and Alaskan populations, the oddities of the first map make more sense. The counties in the Midwest and Arizona that are deeply purple in the first map are deeply green in this one, indicating a high percent of Native Americans. There is a high concentration in Montana, North and South Dakota, Oklahoma, Arizona, and New Mexico. Most of the Deep South and Eastern Seaboard lacks a substantial Native American population, a fact that fits the historical context.



The last map shows the percentages of people who claimed two or more races. Purple dominates the western half of the country, especially in California and New Mexico. There is also a purple section in Oklahoma which (based on the previous two maps), seems to be due to the high number of Native Americans living there who have likely intermarried outside of their “race”. Florida, Michigan, and even a bit of the North East also appear to have intermingling- though not quite in the same numbers as the west.