My Challenge
This week my goal is to explore the field of data visualization. Living in the era of big data, data contain powerful information. However, simply having data itself is not enough. It is difficult for normal people to recognize a pattern among a bunch of unorganized data. Therefore, the way to present the data — data visualization — becomes important. I used Tableau, a new enterprise standard software, to help my user persona, the event coordinator of Cascade Bike Club, to find an appropriate date and location for a summer cycling camp. The data I used for the visualization is the public data from https://data.seattle.gov/Public-Safety/Seattle-Police-Department-911-Incident-Response/3k2p-39jp. In order to help answer the research questions of “Where is the location with least bike-related crimes in U-district”, “Which month is the month with least bike-related crimes in U-district”, and “What time period within a day is the time with most bike-related crimes in U-district”, I created three data visualization. The chart I created to address these questions are a symbol map, a bar chart, and a line chart. The map can help visualizae the “density” of crimes happened in a particular spot, the bar chart clearly show the number of crimes for each month in a non-continuou way which is consistent with the month unit, and the line chart can show the trend of crime number because of its continuity over time. For all of these charts, I applied Tufte’s priniples to make the visualization more convincing. For example, I used size in the map and different color range in the bar and line chart to emphasize the difference in number of crimes. This can help uesrs to compare the difference of data pieces. I uploaded my final work to Tableau Public and here is the link:
https://public.tableau.com/views/HCDE210Visualization/Dashboard1?:embed=y&:display_count=yes
Critical Reflection
The most difficult problem I encountered in the data visualization is about how to choose the appropriate chart type to represent data. New to Tableau, I’m not familiar with the chart type it has to offer and it took me several tries to find the best way to represent data. For example, when trying to find a chart to visualizr the number of bike-related crimes by hour, I first tried the packed bubble as shown below.
However, this is not a good choice. Althought the size of the bubble can represent the number of crimes, the pattern these bubbles being displayed is random and hard for user to quickly recognize. As I continue explore the chart type, I found that the line chart is a better choice because its chronological order and continuity can help visualize the relationship between different hours, which fit Tufte’s principle of “comparing” different pieces.
But as I continue exploring, I found a even better tool for visualization — color. By using “warm” orange for high value and “cold” blue for low value, I successfully introduce the color contrast, which can further help distinguish betwen different data pieces.
The lesson learned here is that sometimes first try might not be the best way to represent the data. As designers we should explore more possiblities of different ways to represent data so that the visualization can be more informative and easy to recognize for our audience. The tool here can be used include chart type, color, font size, etc. In my future data visualization, I’ll continue trying different combination for data visualization.
Future Application
Data visualization is the tool that bring data “alive”. It is applicable in every aspect that links to data, especially for explaining design purpose. One important thing in the design field is to point out why there is a need for the design and this process usually evolve some statistics. Data visualization can be a powerful tool here. For example, suppose I am showing why we should develop self-driving car. In my speech, simply tell you “self-driving car is good” is not enough. I plan to use statistic to show that self-driving car is much safer than human-drivers. To show the graph below is much more convincing in that you can “visualize” the curve of self-driving is much steeper than the classic car so that you will have a better sense exactly how much safer self-driving car can be and therefore understand the need of designing a autonomous driving system.
Therefore, whenever I need to present my design purposes, especially those connected with data, data visualization is always a powerful tool.
