Although New York City ranks globally as one of the most modern cities in the world, it is also an aging city, one that constantly undergoes wear and tear, and in need of constant maintenance. In order to request for maintenance or report less than desirable conditions, people may call 311 to file a request. The aim of the overall project was to gain an insight for the metrics on how many requests are created, and the nature of these requests.
While the dashboard itself was created as part of an assignment, the intended audience for a dashboard such as this would agencies that oversee and delegate resources and manpower to other agencies, such as that of the Department of City Administrative Services (DCAS), or the Mayor’s Office. They would leverage the insights provided by the data and use it to drive decisions in terms of spending money, contracting third party vendors, and scheduling projects.
The dashboard primarily centers itself around counts. The primary table in the upper middle and upper left of the dashboard is a two-way contingency heat table that lists the overall counts of commonly requested service categories by each borough. Having access to these counts allows for more advanced calculations such as finding the rate of completion, which is detailed along the upper right side of the dashboard.
The bottom of the dashboard is relegated to charts and line graphs. The charts help to give a better sense of how each category stacks up against one another, since numbers, especially numbers in a contingency table can be intimidating to a first-time viewer or a novice. The line graph itself is a high level time series overview that catalogues how many incidents were called in over a five year period.
The tables and charts chosen for the dashboard were for simplicity’s sake. It was meant to be easy for the average office worker to take a glance and glean information rapidly. The heatmap was used in particular to draw attention to the darker colors, since those have higher counts, and would need more attention. In addition, bar charts are a staple in any dashboard in that they are simple, and clean, and the size of the bars deliver a quick and easy idea of how one category stacks against another category.
For metrics such as the average or the completion rate, no charts were needed since those were single number metrics. As such, putting them in chart format would have been overkill and awkward to fit the information accordingly, so the most logical course of option was to display them as single numbers accompanied by some contextual words.
Of course, the dashboard is not without its limitations. The issue of finding enough space versus potentially overloading the end user with too much information is always present. Furthermore, a more complete picture would be painted ten years’ worth of data were utilized in comparison to five years, but that is a limitation due to hardware memory (an expensive problem to fix, out of the budget of a grad student). Finally, because the data is coming from a multitude of different sources, the data needed to be molded into a more usable form. One example of this is the categorization of the data; there were many overlaps such as there being multiple plumbing categories because of the fact that they went to different agencies, and New York City agencies are not exactly one cohesive unit even with an administrative backbone like DCAS or the Mayor’s Office.
Because of these limitations, this is not a one and done project. The project will have multiple iterations and updates in order to convey the data in a manner that finds the optimal balance between providing enough information, and being simple enough for users to understand it.
A quick wishlist of things to follow up with, including, but not limited to:
– More interactivity (filters, search bars)
– Additional insights such as which time of the day, or which day of the week receives the most requests
– Which zip code/neighborhood sees the most volume
– Figuring out better color schemes to balance between drawing attention, and readability
