Sprint 6: Visualization
For this sprint, I learned how to use Tableau in order to practice information visualization. I chose my user to be a member of the Cascade Bicycle Club and considered answering the research question: Based on the City of Seattle’s 911 Incident Response data, how can members of the Cascade Bicycle Club create an infographic that minimizes the amount of bicycle thefts in Seattle?
I learned how to use Tableau through the tutorial from studio. Working through the pre-studio and in-studio assignments taught me how to import a dataset (in this case, the City of Seattle’s 911 Incident Response Data from 2015) assign filters, and visualize data in variety of different ways. For example, I used a mix of map locations, heatmaps, and bar graphs in order to depict the specific data relevant towards bicycle theft in Seattle. After creating three different visualizations, I wrote a memo in order to describe and justify them while directly addressing the initial research question.
My first visualization displays the most common places for bicycle theft to occur in Seattle. I did this by marking the median locations of where bicycle thefts were reported in each sector of the city. This method made the most sense to me, as the median locations would show the heart of where bicycle theft occurred. My second visualization depicts the number of bicycle theft occurrences in each sector, through a heatmap of the median locations. The darker points represent sectors with more bike theft occurrences while the lighter points represent areas with less occurrences. Additionally, hovering over each point gives the exact number of bicycle thefts in that sector in 2015. My third visualization is a bar graph that sorts the number of bicycle thefts by each hour. Again, darker colors mean more occurrences while lighter colors mean less occurrences. Using all three visualizations, a biker from the Cascade Bicycle Club could most likely determine when and where it would be the most unsafe to leave their bicycle unattended in Seattle.
So What? (B):
Personally, this was one of the most interesting sprints for me. I really enjoyed being able to learn how to use Tableau effectively in order to create visualizations. Working on visualizing data and finding the best way to represent that data was challenging but exciting at the same time. It was easy for me to see the applicability of creating visualizations using Tableau, because of how widely used it is in the industrial world.
I also learned about how data can be communicated in a variety of ways. Some visualizations are better than others at making data easily understandable to a specific audience. Additionally, I learned about how some visualizations can misrepresent a dataset, causing confusion to an audience. My experience in this sprint taught me how to sort through a large dataset with relative ease and display information that is relevant to a certain research question, while maintaining a truthful representation of the data.
In the future, I would like to explore the other features of Tableau as well. I saw many different types of visualizations that interested me. However, they seemed less appropriate to use when addressing the research question and scenario that I came up with. I would also like to work with different datasets, since I found visualization so interesting.
There is so much data that exists in the world today and each day, more and more data is generated. This leads to the problem of sorting through all of that data and communicating it effectively to a particular user or audience. Visualization is a powerful tool that communicates data in a way that is easily understood. For example, my visualizations for the Cascade Bicycle Club filtered out all sorts of information in a dataset of all 911 Incident Response reports in Seattle (2015). The dataset included information on incidents such as traffic violations, public misdemeanors, and many more reports. However, I only needed access to the bicycle theft reports. Tableau made it easy for me to isolate the information surrounding bicycle thefts and to visualize them in a way that was relevant to my users.
In the future, I would use visualization to narrow down a dataset in order to answer a specific research question. This would make the data more understandable to an audience, by eliminating any excess or distracting data. Another use of visualization could be through creating a different representation of data. In this scenario, altering how a dataset is represented could potentially communicate the data more effectively.
I think that almost any project involving data could lend themselves to the use of visualization. There are many advantages to creating effective visualizations including enhanced clarity and engagement with the audience. A successful visualization would stimulate the viewer to think about the data being presented. However, visualization can also have drawbacks if the visualization representing the data was distorted or lacked integrity. Visualizations, whether purposefully or not, can exclude certain data that alters how a viewer understands the dataset. They can also be uninformative, or displayed in a way that confuses the viewer. Therefore, I believe that visualizations of data should only be created when particular data can be communicated effectively with clarity and integrity.