Fork me on GitHub

Project Team #9: Environment, DS4200 F20

Derek Brenner, Julia Nitschke

Service-Learning Course Project as part of DS 4200 F20: Information Visualization, taught by Prof. Cody Dunne, Data Visualization @ Khoury, Northeastern University.

Abstract

This project is about the relationship between economic behavior, education, and outcomes, and particularly their relation to environmental concern. The idea that people should become more aware of environmental issues is not enough; we need pro-environmental action. We are particularly interested in what kind of information impacts people's perceptions of their actions, and the trigger points that make people act on their knowledge and awareness. In this project, we compare survey respondents' recycling perceptions to actual recycling data from public data sets. We also look at how people's recycling actions relate to their overall environmental concern, as well as willingness to participate in other positive environmental action. Our partner is a research team at Northeastern led by economics professor Dr. Venkatesan. The data used in this project is from a survey conducted by Dr. Venkatesan and a group of Northeastern students.

Visualization

You can interact with the public dataset graphs by highlighting either points on the line graph or rows in the table. The corresponding data points in the other graph will be highlighted as well.

You can interact with the Northeastern University survey result graphs by hovering over the pie chart sections or bars to get details on the precise values in the dataset.

Public Dataset Results

Data on Plastics in Municpal Solid Waste (MSW) by Weight (in thousands of U.S. tons)
Data from Envrionmental Protection Agency

In 2018, only around 8.5% of plastics produced were recycled. A majority of these plastics ended up in landfills.



* Note that the time between years is not consistent and is based on the data that is publicly available.

Northeastern University Survey Results

Demo Video

Visualization explanation

Public Data Visualizations:

For our public data visualizations, we chose to include a line graph and a table with correlating data for the viewer to visualize how much waste has been recycled over time. For the first visualization, it is a line graph depicting the percentages of waste recycled over time. The data we used here was parsed from the existing public data set, using a simple division calculation for ease of understanding for the audience. We chose a line graph because they best depict trends over time, which is the most important factor for this visualization. The line graph also changes in size depending on the size of the browser, so that it is always easily readable no matter the dimensions. When viewing the line graph, there are several points that are shown as small circles, which are the exact data points used to construct the graph. These circles can then be selected by the user to see the corresponding data in the adjacent table (more information below). The encodings we used were the y-coordinate of a point as a representation of the percentage, and the progression of the x-axis as the change in time.

The second visualization pertaining to the public data set was a table showing all of the values given from this data set, including the year, how much waste was generated and recycled, and the percentage of recycled waste. The table provides clarity to the audience on the exact number of tons that were generated and recycled, and can provide context for the line graph right beside it. Users can also analyze and link the data from the line graph to the table using brushing and linking, where if the user clicks and drags, they can create a box that will match data from the table to the line graph, or vice versa. This method makes it easy for users to draw analyses based on the information from the line graph and the table with the public data, as well as understanding which data points are linked to each other.


Northeastern University Survey Results Visualizations:

To visualize the Northeastern survey results, we used a series of pie charts and stacked bar charts. The data is mostly qualitative and proportions matter, so these graphs satisfy both of these data elements. We kept the color encoding consistent across the visualizations, with the light red/pink color denoting not routeinely recycling and the light green meaning routinely recycling. We used light colors so saturation isn’t too high and the text is easy to read. We kept the colors consistent across graphs because they represent the same attribute. We made these specific color decisions because of the general association of red meaning stop/no and green meaning go/yes, but also because green has an association with positive environmental action.

In each graph, the user hover over the pie chart sections or bar chart sections to see more detail. The more precise values associated with that section will show to provide more information. The user can compare how answers to the questions change based on whether or not the respondent routinely recycles, and also look at overall sentiment toward the questions.

Link to presentation: Google Slides

Acknowledgments

List here where any code, packages/libraries, text, images, designs, etc. that you leverage come from.