By Ziyue Li
23/5/17
What did we do?
In this week’s studio, We were learning Tableau by using pair programming technique which is surprisingly helpful for the whole learning process, and I personally improve a lot from skills aspect to communication aspect. Then we worked together on a research topic based on the Seattle 911 Incidents Dataset. First, we decided our user group and research question which is UW students who are looking for off-campus housing in the university district. After deciding our users, we put ourselves into the users’ shoes and imagined what if I were the users, what factors I need to consider and look for. Then, we started to take a close look at data and determine several data that we want to use for our user group, for example, assault, residential burglaries, noise complaints and theft and so forth. After that, We narrowed those factors down by using Tableau to visualize the data and have a better understanding and perspective of the data. For example, we chose the the number of assault incidents reported in the university district not only because it is closely related to students safety on their commute from school to their living space, but also we saw the interesting trends in the three different zones which students may need to to see and take that information into consideration. Because we can not see the pattern just directly look at the numbers, but after putting the data into the line graph that we can know the trends for the number of assault in three zones, which provides a different angle to consider the best location. While considering the data we chose, we also seek for the best and various data type to present data, using the line graph, map and bar graph. We believed that the graphs that we chose well-represented the data unbiasedly and provide our user group important information to solve their problems.
What questions did it raise?
During the process of creating the different data visualization, I was trying to figure out the best way to present data. Choices are ranging from the line graph to filled map. Each one focus on the different aspects of the data. For example, I used the line graph to present the assault data to show the trend of assault in university district from 2014 to 2015. The audience can tell the trends by simply looking at the lines going upward or downward. From the line graph, I found out that as the numbers assault reported in zone U2 and U3 are decreasing, the number of assault happened in U1 is increasing. However, when I tried to use a bar graph to show the data, users might get a different feeling about U1 and U2, although the number of assault happened in U2 is decreasing, the base number is way larger than the other two areas. Zone U1 is the other way around, the number is increasing, but the base number is fairly small. Trying different ways of visualizing data makes me think that even if we use the real data, presenting in different graphs or models can also mislead people, which leads me to the questions: How to present data in an unbiased and objective way? I think the target audience, the intention of the producer and the data itself can all be factors that influence the data visualization. People in news industry or some high technology companies are very likely to manipulate those data to achieve their goals which is a current phenomenon. I’d like to know how to avoid distorting data but to present in the most informative and understandable way for all the audience and let people make their judgements.
Application in the future
This week’s project provided me a better understanding of data visualization. I can see myself applying this technique in a lot different situation, For example, doing a presentation which requires me to present a large amount information in a fairly short time and in the most accessible way. I think the data visualization can be applied to any data-related projects like newspaper articles, a presentation about a product, analysis of the users’ information (like a sleep app that analyzes sleep patterns). People who use data visualization may aim to imprint a powerful image to the target audience about the data itself or the message they want to deliver. I think all the data-related projects are appropriate for using the visualization technique because it is such powerful way to present information and get different perspectives in using different models when looking at the data. However, data visualization can be easily skewed and altered from the unbiased and honest data. Therefore, I don’t think there are any projects not appropriate to use this technique to present data, but the people who make the visuals should have the responsibility to present data visualizations appropriately.
