HCDE210 Sprint-08 Data Visualization

2/28/17 Creating a visualization for Seattle Noise Disturbances on 911 data

By Qianlin Luo
This is a screenshot from my Tableau.
Who is using my design?

Those who will benefit from my visualization are Citizen scientists who are studying the patterns between noise pollution and human activities in Seattle.

These scientists are trying to look at the potential cause of noise pollution in Seattle from multiple lenses.

In other words, they are finding and making an assumption of when, where, and how noise normally occurs. Through the 911 incident data, they want to explore how occurrence of noise correlates with different hundred block locations. Through my visualization, they will finally be able to answer these questions: do disturbances happen more often in higher density neighborhoods? Does location near bars, restaurants and nightclubs have more incidents than others?

In addition, based on the different dates and specific times of the 911 calls, these scientists are also curious of the correlation between time and noise incidents. Through my visualization, they can study the noise pattern thoroughly and answer all these questions one by one: what time of a day does noise occur more often? Which day of the week do noise incidents take place frequently? This data of time is important because it also implies the daily schedule of Seattle residents and provides a big amount of valuable information for these scientists to study.

What matters?

In this scenario, a question these citizen scientists may ask is

“According to the 911 incident data, what are the causes behind the pattern of noise pollution in relation to time and different locations in Seattle?”

This relational pattern between noise, time, and locations will effectively guide these scientists to identify the actual causes of noise pollution. For instance, through my visualization map between location and number of incidents and chart of hours vs average incidents, scientists could answer whether bars and nightclubs near a place are the essential cause of noise incidents at night. However, this question can be sometimes difficult to answer because there are many different factors that may also potentially lead to a noise incident. For instance, Scientists are trying to prove if having restaurants is a big cause for noise pollution. In reality, when scientists are looking at one location of a restaurant, a noise incident might simply have happened because a couple had a fight at their apartment above this restaurant.

How my visualization works?

In order to help users identify the possible causes for noise pollution, I focus on a visualization that portrays the correlation between noise on hundred block locations, weekday and hours. Guiding by Tufte’s principles discussed in class, I designed the visualization in this way:

This is my Disturbance Map.

Viz #1 — Noise Disturbance Map

For my first visualization, I made a map that portrays the correlation between noise incidents and hundred block location. It addresses the research question by showing where the noise incidents occur more often in Seattle. Then, according to Tufte visualization principles, I induce the viewers to contemplate on the map: why are some locations darker? The bar underneath the map helps the viewer to interpret the data based on the color. The varying numbers of incidents in different locations have been reflected in varying orange colors. The darker the orange color indicate the more incidents recorded in this location. I think this visualization is effective because it allows the user to easily make a comparison between two locations based on colors and easily identify the pattern of noise in relation to locations. Therefore, users could use the location pattern to predict the possible causes of noise incidents and answer my research question.

Viz #2 — Hours vs Noise Incidents graph

For my second visualization, I made a graph that shows the relationship between noise incidents and different hours of the day. It addresses the research question by presenting when the noise incidents occur more often in the day. Then, according to the Tufte visualization principles, I avoid mixing up data from different hours together and give them contrasting colors. The chart under the graph helps the user to identify the hour based on its related color. I find this illustration effective because it allows users to easily make a conclusion and come up with a pattern by comparing different heights of bars. After understanding when the incidents occur more often, the users could predict the possible causes (work, party, etc.) for noise incidents and answer my research question.

Viz #3 — Weekday vs Noise Incidents Graph

For my third visualization, I made a graph that presents a clear connection between noise incidents and different days of the week. It addresses the research question by portraying when the noise incidents occur more often in the week. Then, according to Tufte visualization principles, I also encourage the viewers to compare data from different days of the week: why are some days darker and bigger? The bar under the graph helps the user to calculate the number of incidents based on the color. The darker and bigger one square is the more incidents recorded within this day. I believe this visualization is effective because it allows the user to easily identify the pattern of noise in relation to the weekday based on the size and color of the square. Therefore, users could use the weekday pattern to predict and eliminate the possible causes of noise incidents and answer the research question.