Introduction

One of the most exciting aspects of living in NYC is hoping you’ll run into a celebrity on the streets or walk in the background of a film shoot and unintentionally make your big-screen debut. It’s not surprising, given that NYC is known as the media capital of the world (Richter, 2015) and is one of the most popular filming locations in the world, specifically Central Park and Times Square (Davidson, 2021). I remember the excitement at Pratt when  “And Just Like That” was being filmed a couple streets over from the Manhattan campus last year. As a pop culture aficionado, I am very interested in all things media, entertainment, and celebrities. Hence, for my final project, I decided to create visualizations to show:

1. The number of filming events that took place across the 5 NYC boroughs from 2021 – 2022 (the complete dataset is available only for these 2 years; I did not include 2023 in my analysis since it only has a few month’s worth of data) 
2. The types of filming events that took place in NYC from 2021 – 2022 
3. The difference between the number of filming events that took place in NYC boroughs from 2021 – 2022 by month

Methodology

I found the Film Permits dataset on the NYC OpenData website. Since producers have to obtain a film permit from the NYC Film Office before any major shoot, I knew that this dataset would have the information necessary to visualize my topic. The dataset contains 13,000 rows and I looked at the following variables:

Categorical	Time-based
Event Type	Start Date Time
Borough	End Date Time
Category
Sub-category

The dataset did not require any cleaning and I created my visualizations in Tableau Public. Since the goal of information visualizations is to “promote insight and understanding” (Banissi, 2014), it’s imperative that visualizations are “perceivable, comprehensible, and useful” (Banissi, 2014). Hence, as Banissi et al. argue, evaluating information visualizations in terms of their usability are an integral part of the research and development process.

I conducted user research on my visualizations to understand whether my visualizations are intuitive and clear to viewers, and whether viewers are able to take away the main insights from the visualizations around where filming events took place in NYC, what types of filming events took place in NYC, and whether there were any differences in the number of filming events and filming locations over time. 

My first step in the research process was to conduct a design audit on my end to address quick design considerations such as making sure the axes were labelled correctly, the chart titles were written in simple and clear language, the legends did not contain duplicate colours, tooltips/labels were enabled where appropriate, and that the visualizations were simple and not overwhelming.

I then recruited 2 participants from my personal network who met the following criteria:

• Over 18 years old
• Has not taken an info viz course (to keep the user testing insights generalizable to the average viewer) 

For my user research sessions, I took a qualitative approach. I asked my participants to look through the visualizations and asked them to explain to me what they took away from the particular visualization and if anything was confusing to them. I wanted to understand if my participants’ main takeaways matched what I wanted to visualize. In addition, in some instances, I made 2 different visualizations for the same research question. In these instances, I asked my participants if they preferred one over the other, and why. 

Overview of the Visualizations

My first visualization answers the question: How many filming events took place across the 5 NYC boroughs between 2021 – 2022?

I decided to visualize this question through a simple bar chart because the difference between the number of filming events that took place in Manhattan vs Brooklyn is not that large, and a bar chart shows the number of filming events that took place in each borough more clearly at a first glance, as compared to other visualizations e.g. a treemap. 

I created a treemap for this question as well and showed it to my participants during the user research sessions. Both of my participants preferred the bar chart over the treemap because they found it clearer and easier to digest. They both mentioned how the information in the chart is easily digestible from a first glance in the bar chart, whereas in the treemap, they had to hover over the boxes to actually see the data. However, one of my participants called out that the treemap was more visually engaging than the bar chart. Hence, instead of  displaying all the bars in the bar chart in one colour, I colour-coded them by the different NYC boroughs in an effort to make the visualization more visually-appealing. 

The second visualization answers the question: How do the number of filming events that took place in NYC from 2021 – 2022 differ by month? 

I selected a line graph to visualize this data since it contains a time-based measure. Line graphs are the appropriate choice to depict how variables change over time (Yi). Both of my participants found this chart straightforward and intuitive. They both were able to infer that the most popular months for filming in NYC are October and March (which makes sense since the weather in NYC is best during the spring and autumn months). One of my participants mentioned that they found the labels that show the total number of filming events above each point for a month very helpful. One participant called out that the chart title could be made clearer by specifying that this visualization is for all of NYC and does not show the data split by boroughs. This demonstrated to me the value of having people look over your work because they are more likely to find small adjustments required to make the visualization clearer or more intuitive since they analyze the visualizations from a fresher perspective.

The third visualization answers the question: What types of filming events took place in NYC from 2021 – 2022?

I created 2 visualization – a treemap (shown above) and highlight table. While this is the perfect use-case for a treemap since the data for this visualization represents part-to-whole relationships through categories and sub-categories (Rowe, 2022), I wasn’t sure if the treemap is overwhelming. In my opinion, a table is a little easier to skim through. However, both of my participants found the treemap informative and liked how it condenses a lot of information in one visualization while ensuring the data is still digestible. Hence, I decided to go with the treemap. I colour-coded the boxes by the broader category and viewers can see the number of filming events that took place in NYC from 2021 – 2022 by each sub-category.

The biggest takeaway from this visualization is filming for TV shows episode are the most common types of film shoots that took place in NYC from 2021 – 2022. It’s interesting that in comparison to TV shows, not that many movie shoots took place since I was expecting there to be a lot of movie shoots that took place in NYC as well.

The last visualization answers the question: How do the number of filming events that took place in NYC boroughs from 2021 – 2022 differ by month? 

I used a line chart to visualize this data as well since it contains a time-based component. Again, both participants found this chart intuitive and straightforward, and liked how the lines are colour-coded by borough. I did not include data labels in this chart since there are 5 lines instead of 1, and I do not want to clutter the visualization with too much information. It’s interesting to see how the number of filming events that took place in Manhattan steadily increased over time whereas both Brooklyn and Queens saw a dip in the number of filming events during the months of June and July. All boroughs except for Staten Island also saw a decline in the number of filming events in December.

Summary of Key Insights From the Visualizations

1. Manhattan had the most number of filming events from 2021 – 2022, followed by Brooklyn.
2. March and October were the most popular months for filming from 2021 – 2022.
3. Majority of the filming events that took place in NYC from 2021 – 2022 were for TV shows, specifically, episodic series.
4. The number of filming events that took place in Manhattan steadily increased over time, whereas both Brooklyn and Queens saw a dip in the number of filming events during the summer months of June – July.

Summary of Key Insights From the User Research

1. Stick to simpler visualizations (e.g. bar chart instead of a treemap) to make the data more digestible.
2. Use of aesthetic strategies e.g. colour-coding by categorical variables can elevate even simple charts and make them more engaging.
3. Make chart titles as specific as possible.
4. Adding data labels (where appropriate) can make the process of digesting information from a visualization easier and less cognitively-demanding.

Reflection and Critique

Overall, I am very pleased with my final visualizations. I definitely felt more comfortable working in Tableau this time around compared to the first time I created visualizations in it, and I also felt more confident in terms of selecting the appropriate visualizations for the questions I was trying to answer. I was happy that no major pain points came up during my user research (both during my personal design audit and the user interviews I conducted) and overall, my participants found the visualizations clear, intuitive, and informative.

The one thing I would have loved to do would have been to create a map which shows exactly where the filming events took place across the boroughs. Unfortunately, I experienced issues with loading CSV files as a layer into QGIS as well as plotting data from the CSV file onto a shape file, so I wasn’t able to create the map. While the map isn’t critical (since my visualizations include a bar chart which shows the breakdown of the number of filming events from 2021 – 2022 by NYC boroughs), it would certainly be a nice way to supplement the data in the bar chart and provide viewers with additional context. Hence, if I were to expand on this work, I would certainly include a map which plots exactly where in NYC the filming events took place, categorized by the category of the filming event.

I would have also preferred if the dataset included information on filming events from years prior to 2021 and 2022 to compare the number of filming events in NYC pre and post pandemic. Hopefully, this dataset is updated to include previous and subsequent years so that I can visualize a more holistic picture of the frequency and locations of filming events in NYC. Lastly, it would have been super interesting if the name of the project e.g. “And Just Like That” had been included in the dataset. The addition of this data would have certainly made my visualizations more exciting and engaging.

References

Ebad Banissi, E., Camilla Forsell, E., & Francis T. Marchese, E. (2014). Information Visualisation: Techniques, Usability and Evaluation. Cambridge Scholars Publishing.

Davidson, B. (2021, February 28). The Most-Filmed Location in Every Country. NetCredit Blog. https://www.netcredit.com/blog/most-filmed-location-every-country/

Film Permits | NYC Open Data. (n.d.). Data.cityofnewyork.us. Retrieved May 2, 2023, from https://data.cityofnewyork.us/City-Government/Film-Permits/tg4x-b46p‌

Richter, F. (2015, March 11). New York Is The World’s Media Capital [Review of New York Is The World’s Media Capital]. Statista; Statista. https://www.statista.com/chart/3299/new-york-is-the-worlds-media-capital/

Rowe, S. (2022, September 6). What is a Treemap? [Review of What is a Treemap?]. Storytelling with Data. https://www.storytellingwithdata.com/blog/what-is-a-treemap

Yi, M. (n.d.). A Complete Guide to Line Charts. Chartio. https://chartio.com/learn/charts/line-chart-complete-guide/

‌