As a geography professor, the incorporation of effective visuals is essential to my teaching. Often, these visuals consist of charts or maps that show how some phenomenon of interest varies geographically around the world. For example, a map of life expectancy by country can reveal the distinct inequities that persist between countries in the developed and developing parts of the world; similarly, a line chart of birth rates for countries by year can ably illustrate variations in the demographic contexts of different societies.
Locating effective visuals, however, can often be a significant challenge for an instructor. I, like many other instructors, have combed through the pages of Google Image Search trying to find the right chart or map for my course topic. Often, available visuals are of poor graphical quality; are out of date; or do not fit precisely with what I want to communicate to my students in class. Further, these visuals are usually static images; as such, the potential to foster an interactive experience for students around these graphics is limited. In response, I started experimenting with creating my own interactive data visualizations for my courses last spring, which has now become a major initiative in my teaching.
In the past three years, many new frameworks have emerged for the development of interactive, web-based visualizations that run in the browser. While some of these frameworks require knowledge of programming languages, others are accessible to practitioners at varying levels of technical skill. An excellent example of this is Plotly (https://plotly.com), a new web-based visualization and data sharing tool, which I have been using in my World Regional Geography course over the last month. Plotly includes a web-based interface that allows novice users to upload or enter their own data for rapid visualization; for more advanced users, Plotly interfaces with several popular programming languages, such as R and Python, enabling the creation of more sophisticated charts. Some examples of my Plotly charts are available at this link: https://plot.ly/~walkerke/.
Interactive charts like these serve as an invaluable resource for communicating with my students. Static images are necessarily limited to one specific view of a dataset, whereas interactive charts allow me to actively explore the data behind the chart with my students. For example, I frequently produce line charts using data from the World Bank’s open data catalog (https://data.worldbank.org/), which show how demographic or economic characteristics of countries have varied over time. In one example (https://plot.ly/~walkerke/12), I show variations in fertility levels by countries in South Asia. As this chart is interactive, I can zoom in to different parts of the chart to create a specific focus on particular time periods. Additionally, the interactivity in the chart gives me direct access to my data. When I hover my mouse over a data point, Plotly reveals the specific value of that point. In turn, my students and I can interactively explore the data in the chart, and I have information at my fingertips to help me respond to students’ questions.
As a geography instructor, I also frequently use maps in my instruction. Interactive maps are immensely useful instructional tools for many of the same reasons I outlined earlier regarding interactive charts; I can zoom and pan around the world as needed, and click on countries or places to get a pop-up with specific information. The tool I rely on the most is ArcGIS Online (https://www.arcgis.com); an example showing projected population change to 2050 by country is available here: http://bit.ly/1mIFNfl. In my World Regional Geography course, I also have used this tool to put interactive visualization in the hands of my students. I have prepared geographic datasets on a variety of topics (e.g. gross domestic product per capita by state in Mexico, and female educational attainment by country in sub Saharan Africa), and asked students to create interactive shaded maps with these datasets and write brief reports analyzing the geographic patterns they have revealed. These assignments have successfully given introductory students experience working with data and mapping, and allowed them to explore our course topics in additional depth using materials they created themselves.
I am encouraged by the potential of interactive data visualization in the classroom, and I am actively working to make my materials publicly available so that others can use, adapt, and modify them for their own purposes. I am documenting some of my examples on my fledgling blog, “Teaching with Interactive Data Visualization,” available at http://walkerke.github.io. This website includes tutorials and walkthroughs that show how I have created the visualizations I use in my teaching. For more advanced visualizations that require writing code, I host all of my code in a related GitHub repository, available at https://github.com/walkerke/teaching-with-datavis. My hope is that teachers and practitioners who want to get started writing code to create their own visualizations can use my examples to give them a head start.
I also look forward to working with faculty and staff at TCU who are already using data visualization in their teaching, or who are interested in learning more. This fall, I will be leading a Faculty Interest Group on teaching with data visualization, in which I will introduce participants to the different visualization frameworks I use. If you are interested, please contact me at kyle.walker@tcu.edu.
This article was written by Kyle Walker, Department of Geography, for the Fall 2014 Issue of Insights Magazine.