[Lab #4] Exploratory Data Visualization

For this week’s lab, I used Rawgraphs.io to visualize baby name data from New Zealand. I really enjoyed the Rawgraphs.io interface, and turning tabular data into something visual and understandable was quite fun! My final data visualization is below.

Data + Graph Style

I chose to visualize the top ten most popular baby names in New Zealand by gender in the year 2010. I decided to show the data in a sunburst diagram. I think the sunburst diagram is an effective and appropriate graph style because it allows the viewer to understand the graph in layers. First, they can look at the inner most ring and understand that the data is split apart by gender. Then, they can move to the next ring and see that the data is further split by rank, where 1 represents the most popular name and 10 represents the tenth most popular name. Finally, the viewer can move to the outer most ring and see which name is associated with each particular ranking for both genders. The ability to break down the data into meaningful layers is what made me choose the sunburst diagram as my visualization tool.

Changing the Data to Improve Clarity

When I was first working with the sunburst diagram, I found it incredibly hard to extract anything meaningful from the visualization because the full data set was simply too dense to fit on a single graph. In an attempt to make the data cleaner and clearer, I decided to go into the original CSV file and edit the data so that the file only contained data on baby names from the year 2010 (the most recent recorded year in the data set). Making this change to the data set greatly improved clarity, and when I made a new sunburst diagram, there was enough space in the visualization to actually be able to read and interpret what was being displayed. I also changed the colors of the diagram to something a little easier on the eyes.

Reflection on Lin’s Lecture

In her lecture, Lin mentioned that there are two steps to a successful data visualization: exploration and explanation. Exploration is finding the stories in a given data set, while explanation is showing the stories in the data set. In this lab, I found the exploration step was mostly done for us because the data file we were given was clearly labeled and nicely communicated what kind of story the data was telling. In the case of our data set, the story was about baby name popularity by year. So, our task of visualizing the data fell under the category of explanation. What type of graph should we use to inform the viewer about the meaning of this baby name data in the best way possible? What labels and columns of the data should be included? From an aesthetic standpoint, what colors and font styles should we use? These are all questions I had to consider while creating my data visualization.

I think digital humanities as a field tends to focus more on the explanation step, too. As Lin mentioned, oftentimes the data that humanities fields draw on is already vaguely understandable. If analyzing a text, for example, the stories in the data set are already displayed in front of us. The author has written legible words that we can in some way have an immediate understanding of. What digital humanists tend to focus more on is how they analyze and explain the data. What is the significance of the data? What parts of the data should be highlighted? How might we present the data in a way that allows the viewer to think about meaning rather than technical understanding of the visualization? In this sense, this lab really reminded me of the kind of work that digital humanists do.