I enjoyed Lin Winton’s lecture on Data Visualization. In the past two terms, I have been entrenched in statistics courses, and the class gave me a better idea of what graphs are best for what types of data. Also, I never realized that there were two types of graphs; exploratory graphs for the use of the creator and explanatory graphs to show other people. In high school, I was taught to use Google Docs’ or Sheets’ charts on everything I showed to other people, no matter how ugly or confusing the chart. I also thought it was fascinating that pie charts are not a preferred graph type due to reading differences between the pie’s relative slices. Once again, my middle and high schools liked it when we made pie charts to show data.
From this lecture, I realized that, in DH, charts have a purpose and must be carefully designed to make information accessible to the masses. In addition, I learned that, like words, numbers could lie if not accurately portrayed. This can affect a graph maker’s credibility and create confusion on the part of those who read the graph.
The lecture, Michael Friendly’s Lie Factor gallery, and Michael Friendly’s “Best and the Worst of Statistical Graphics” galleries all reminded me of the subtle problems that may be lurking in the design of graphs. We talked about starting a scale at a different level than the other axis, but there are so many possible problems that one wouldn’t immediately notice. I learned that potential problems might be the scale of items used on the chart, the color scheme used, or metrics measured. I can think of multiple times in my life where I have been guilty of using bad color schemes or inflating a graph by using a broken axis. I need to be very careful when creating graphs in the future to describe the data shown accurately. In my projects, I would like to make choropleth maps while using a different enough color scheme so that it is easy to interpret. I see so many choropleth maps on major news websites, but their charts do not differ in colors. Mostly, I want to create exploratory and explanatory charts to make the charts accessible to others.
I like what you said about using choropleth maps in the future. They’re such an effective visual. Very simple to interpret and much better than presenting huge tables full of boring numbers.
Hey Phoebe,
I also didn’t realize that there were two different types of graphs prior to Lin’s talk, but I am starting to learn about them now in this class as well as my statistics class. I think that the potential for a scale problem that you mentioned is very important to take into account when creating a graph because it can make the implications of the results seem very different.
Hello Phoebe,
I also didn’t realize that there were two different types of graphs prior to Lin’s talk, but I am starting to learn about them now in this class as well as my statistics class. I think that the potential for a scale problem that you mentioned is very important to take into account when creating a graph because it can make the implications of the results seem very different.
Hi Phoebe, I liked the comments you had about Lin’s lecture! I also didn’t know about exploratory vs explanatory graphs before the lecture. After looking through Michael Friendly’s Lie Factor gallery, I thought that maybe some of the subtle design problems may come from trying to make exploratory graphs more public-friendly, or explanatory. That process might push the creators to sacrifice content in order to make it more visually appealing.