Journalism 677: Concepts & Tools for Data Analysis and Visualization

This project looks at the factors that impact college students’ mental health to give institutions and administrators a better idea of how to support specific mental health struggles. As students, who understand how stressful college can be, we wanted to create meaningful visualizations that drive real change in institutions. These mental health struggles included stress, anxiety, and depression, and the variables we used in our research was course (or major), GPA, and financial stress (also rated on a scale of 1-5).

Real Chemistry provided us with ~100 Reddit posts within the nephrology community, and we were tasked with providing branding insights for a pharmaceutical team’s new nephrology drug treatment. We used R to analyze post content and group the discussion into 5 main topic categories, also noting the overall sentiment or “tone” of each post. The large volume of negative posts inspired us to visualize the main frustrations and concerns of the community, supporting our proposal of an empathetic and patient-centered branding approach. Additionally, we noticed that several patients were seeking answers to clinical questions in subreddits rather than asking their nephrologist. This also inspired us to push for a standard of care that focuses on ensuring patients are thoroughly informed throughout their treatment journey.

Our poster is about helping policymakers advocate for the trees in Queens, NY. In NYC, trees are crucial for shading sidewalks, filtering air pollution, reducing street temperatures, and bringing a sense of calm to one of the world’s busiest urban environments. We identified that there are trees that are in danger in Queens, that the danger is geographically concentrated, and that some species are in more danger than others. We determined the best course of action on strategic maintenance and planting low risk tree species to ensure New York, Queens trees remain in good health.

We analyzed Spotify’s subscription data to understand what actually drives churn across plans and global markets. This infographic shows where cancellations are happening and why. Family plans lead in churn, retention shifts by country, and listening behavior does not predict churn, which pushes teams to focus on pricing, plan structure, and regional strategy.

We wanted to explore motor vehicle crash statistics to find specific danger areas using Wisconsin DOT crash records. While exploring this data we found that several crash conditions proved to be the most fatal, despite frequency. Those conditions were alcohol and drug-related crashes. Thus, upon further analysis we aimed to understand other aspects of these conditions, including frequency, severity, and time trends to ultimately generate conclusions that advance understanding of impaired driving trends in Wisconsin and assist in later research for policy and societal interventions that effectively reduce impaired-driving fatalities overall.

Border crossing has been prevalent in American history since the country’s very founding. Early American settlers crossed the colonial borders of the Appalachian mountains into the Ohio River Valley. Furthermore, Manifest Destiny motivated American settlers again West past their country’s borders into lands owned by Native Americans and Mexicans. In modern times, American political debates ring back and forth over border security and immigration. With such a storied topic still so important after all these hundreds of years, our group decided to focus on the more contemporary chapter in this saga. As such, we devised two separate research questions to guide our study. Our first research question is: “How have U.S. border crossings changed over time, by region and method of entry, and what insights can stakeholders use to improve transportation planning, resource allocation and policy decisions?” Our second research question is: “What operational patterns are shown from comparing U.S.-Mexico and U.S.-Canada border activity?”

In order to better understand the success of our campaigns across customer segments, we developed a variety of graphs that examine conversion rates within target segments, ROI by campaign type and customer segment, and ROI versus conversions narrowing on a specific customer age segments.

Using a sample of 50,000 reviews of European hotels from Booking.com, we coded five visualizations in R language to demonstrate trends in review sentiment and habits of different reviewers from around the world.

This project looks at how geographic location and gender influenced voting patterns in the 2016 presidential election. Using state and county data, we examined which places leaned toward each candidate and what factors might explain those differences. By comparing vote shares, total votes, and gender turnout across states, we aimed to get a clearer picture of the demographic and geographic trends that shaped the election.