Hey! My name is Brosha--some people call me B. I'm currently enrolled in the MPM program here at Heinz and I will hopefully be graduating in December. I've lived in Pgh for most of my life and currently reside on the North Side. My pronouns are she/her.
What I'm hoping to get out of this course is a much better understanding of how to present data in a clean way. I'm not a very creative person, or really a data person, but I hope to get a baseline sense of how to use different methods to convey dense information in a way that everyone can understand. This is a skill I would be able to apply to my work. Conveying information regarding budgets, trends, observations gathered from census tract information, etc. is essential in conveying a message which may result in operation/policy change—something I try to achieve at work. I'm currently employed full-time so I do not have immediate plans after graduation, but I hope to start using the these skills to improve my work.
Here's where I will post my work.
https://datawrapper.dwcdn.net/NWO9R/1/
OIG: DHS Needs to Address Dangerous Overcrowding and Prolonged Detention of Children and Adults in the Rio Grande Valley
When creating visuals from the perspective of the data journalist, I first sought to define my target audience and publication type. I created the content for a publication like TIME magazine, one that is widely read by the general public and which does not have a leaner audience like Washington Post, NYT, Economist, etc. The broad audience forces me to work harder in order to catch the eye. Furthermore, readers may be individuals who lack interest in politics. Knowing that a) I was trying to reach a wide audience and b) that I needed/wanted to elicit an emotional response I explored data sets and visualization options which serve well to demonstrate a stark comparison or simplify the message in order to be effective.
I chose to visualize data found on page 3 of the report—statistics related to the number of detainees who had been held over 72 hrs, as well as more than 10 days. The numbers are impactful in that they are simple and straightforward ratios that represent an incredibly complex problem of overcrowding on a large scale. I chose to use a pictorial visualization for two reasons. First, because it achieves the goal of humanizing the numbers by literally representing them in a human form. This assists in eliciting an emotional response. Second, because I was trying to highlight a subset of a whole in a way that would grab the attention of the audience the pictorial was able to do that in a nontraditional representation with minimal color and element use. The graphic is able to grab attention, plainly demonstrate the data, as well as humanize a statistic.
<script id="infogram_0_efd9b9c5-e536-4dc8-a4fd-5fefa932435b" title="Number of Individuals Held" src="https://e.infogram.com/js/dist/embed.js?l9E" type="text/javascript"></script>Working from the perspective of an analyst creating data visualizations for DHS I chose work with the data presented in a table on page 2. The table is attempting to convey a lot of important information—number of detainees, categorized, as well as present the significant increase by using percent. I attempted to see if I could approach the data in two different ways while still conveying a lot of information to decision makers in a way that would force them to see the need for policy change, etc. In the first attempt, I created two identical bar charts and populated them with FY 17 and FY18 information. I assigned a different color to each category in order to present each category equally and without bias. The colors, as well as identical charts allows the reader to interpret the data, as well as assign importance, on their own. I was not attempting to highlight that apprehension of ‘family units’ had increased 269%--I was attempting to demonstrate that all categories saw dramatic change. The two charts allow the readers to see each FY’s data individualy, as well as in relation to the prior year.
In my other attempt to effectively convey the data I used a grouped bar chart. I grouped the numbers from each FY together in order to compare change in the number of detainees. By having the bars representing different years right next to each other, the visual does all the work in presenting the stark fluctuation in numbers. I included the specific numbers since this article is for decision makers—precision and inclusion of details is warranted.
[Link to the orignal chart:](https://btkachev.github.io/A6/Screen Shot 2019-07-16 at 10.08.38 PM.png)
The chart I chose for my critique was very well done (visual presentation) and one of the things I thought to experiment with is color. Seeing if color can be reduced while maintaining all of the information. The orignal chart had two separate categories for damage--internal and external. 'Harm to user' was broken down to health damage, mental impairment and dependence, mortality, loss of tangibles and relatinships while 'Harm to Others' was broken down by community, econ and environmental costs, crime and injurty, and family adversities. I thought that depending on the intention of the article, the author would want to simplify color. I also thought that by going dark grey for the category of 'harm to others' gave some negative emphasis to the externality. This exercise was particularly difficult when attempting to repackage very complex information without a solid understanding of the metrics behind the scoring system.
<iframe src="https://data.oecd.org/chart/5CSA" width="860" height="645" style="border: 0" mozallowfullscreen="true" webkitallowfullscreen="true" allowfullscreen="true">OECD Chart: General government debt, Total, % of GDP, Annual, 2015</iframe> *Figure 1*AUSAUTBELCANCHECHLCOLCZEDEUDNKESPESTFINFRAGBRGRCHUNIRLISLISRITAJPNLTULUXLVAMEXNLDNORPOLPRTSVKSVNSWETURUSA199619982000200220042006200820102012201420162018 Figure 2
For the third visualization, I chose to use a bump chart to convey data in Figure 2. My goal for this visualization was to use a chart type which, like the bar chart, excels at not only displaying fluctuations within data, but also comparing data sets—in this case comparing debt % of GDP among OECD nations and demonstrating relative standing. The bump chart, I think, does both. For the first visualization, which was a bar chart, I chose to highlight 3 nations—Canada, US, and Mexico—to compare North American countries to all OECD nations. I continued to highlight those three nations in the bump chart by choosing to only use colors for those three, while keeping all other nations in different shades of grey. The three colors I chose were the same ones as those in Figure 1. in order to maintain continuity for the reader. For all the other nations I chose not to use the same shade of grey in order to draw the attention of the viewer to the fact that we were comparing more than a dozen nations.
What I think is most useful about the bump chart is that it is able to demonstrate a number of variables without making the visualization burdensome. The bump chart is able to demonstrate magnitude, as well as the relative size of the GDPs. While comparing the percentage of national debt to total GDP, it is important not to view the nations as homogeneous. Their GDPs vary wildly—it is important to show the significant change each nation underwent while not ignoring important background information. While this visualization is busy the grey minimizes the work the eye has to do, while clearly presenting trends.
Mexico: Blue; Canada: Red; United States: Purple.
Disprarity and Fluctuation among North American nations, % of Debt to GDP 1996-2016.
<path class="layer" d="M0,172.71396744233434L6.806729357