How has COVID-19 impacted the Pratt Community?

By Janet Liu and Celine Yap

about our project

In light of recent events, there is no doubt COVID-19 has been on everyone's minds. It has affected the way we perform our daily routines, the way we travel, and even in the way we interact. As Pratt announced its transition to online classes in early March, our group was curious to know: How has COVID-19 impacted the Pratt Community?

UX Research

We conducted our user test by sending out a google survey to Pratt's listserv. There are 9 questions in total, with the survey constructed in a way that will allow us to visualize multiple kinds of variables (Figure 1). A total of 52 graduate students from Pratt School of Information participated in our survey.

Data cleaning

The datasets were cleaned on OpenRefine, then visualized on Tableau Public. Some of the detailed steps we took to clean our original google survey data includes:

1. Clustering mixed upper and lower case names into the same boroughs
2. Changing locations into to the format: city, state, country
3. Splitting column texts into multiple columns with counts (Figure 2)

DeSIGN CHOICES

We based our visualization design choices on the golden rules mentioned in MacDonald's "Using Color Effectively in Computer Graphics."

The five golden rules are:

1. Take account of human visual needs and expectations.
2. Conform to the color conventions for the application.
3. Design the screen layout considering all available visual variables.
4. Be consistent in the use of color throughout all screens in an application.
5. Use color sparingly, never more than is necessary for the task.

As MacDonald writes, "for effective design of presentations on computer displays, use color in conjunction with the other visual variables of: size, shape, value (lightness), position (x,y coordinates), texture, color, orientation." We followed this by incorporating the color yellow along with various sizes of circles and labels to make sure our visualizations can be interpreted by all.

Based on aesthetics, design, and cognitive theory, we made sure to use colors sparingly and meaningfully to avoid confusing viewers with too many colors. We did this by only using color when highlighting a specific point on the visualizations that were the most important (as demonstrated in the visualizations below). We paired the color "yellow" with gray to avoid using strong contrasting colors that would also leave readers with a preconceived perception or emotion attached to the color like blue or red normally would. Yellow is also a color that the Pratt community identifies with.

Findings

The first visualization is a histogram of the number of Covid-19 cases in New York City starting from the first recorded day (March 4th, 2020) to May 03, 2020- when the data was downloaded. We used this visualization as a timeline, marking important dates with the use of color, associated to Covid-19 in New York City as well as the Pratt Community.

We wanted to see how Pratt took action in response to Covid-19 in the city as well as events that occured within Pratt such as when a member within the community tested positive. The visualization showed that when the city announced the first case of Covid-19, Pratt hosted an infosession the day after. We decided to include the date that New York University announced its cancellation of in-person classes as we wanted a reference point to compare that to Pratt and see whether there was a big difference. Another point we wanted to highlight was March 12th, which was the earliest date recorded from our survey when a student from Pratt left the city.

This Tableau story is based on the data collected from our Pratt survey responses. We decided to present the data in a story format to correspond to the order of questions asked.

The first visualization is a map with the zip codes of areas the Pratt community lived prior to COVID-19. The darker yellow represents areas where more than one student shared the same zip code, while the lighter yellow represents areas where only one student lived. We have also included labels on top as another way to see the count instead of only relying on color. Users can also see the name of the borough if they hover over the zip code.

The second visualization is a map showing where students are currently located. The map includes specific points, labeled with the city and state/province. We decided to use larger marker sizes for users to immediately see which cities have a greater number of students currently living in. Users can also see the count if they hover over the pin point.

One of the questions we asked in our survey looked at whether students in Pratt personally knew anyone who tested positive to Covid-19 in New York City. To visualize the answer, we utilized a bar chart as we felt it was the best way to represent binary answers.

The fourth visualization looked at the question "What is your current preferred pass time activity". We decided to represent the answers in the form of a bubble chart. We found this chart to be suitable to show the different types of activities, with the marker size of the bubble as a dimension representing the number of responses. We decided to group and color the three most commonly chosen type of activity which was playing video games, baking/cooking and watching netflix or other streaming sites. Users can also hover over the bubbles if they wanted to see how many students chose that type of activity.

For our last question which was the final visualization, we wanted to find out how students were feeling about the semester coming to an end. On our survey, we allowed for multiple answers and we felt that the best way to present our findings in visualization was by using a word cloud chart which was a more fun and creative way compared to using a bar chart. Similar to the previous visualization, we used color to group and highlight the three most common feelings which was overwhelmed, nervous and happy. Our visualizations also allowed users to hover their mouse over the words which will show the number of students that chose that feeling.

FEEDBACK & RECOMMENDATIONS

We received a total of 8 responses to our audience feedback questionnaire. The majority of responses were positive, as all respondents found our visualizations easy to understand, aesthetically pleasing, and agreed that we used the right platforms and tools to visualize our data.

When asked on what other platforms we could use to visualize our data, one respondent suggested that ArcGIS may be a better choice for maps. We have made the following revision and included a new map for the pinpoints of where students are currently located: