Suzanne Papik
Projects and Reports:
Recent Projects
I recently ran a marathon and used Strava to track every training run leading up to the race. After running the marathon, I wanted an easier way to reflect on my training and explore my marathon data more deeply — this dashboard was the result!
The first tab highlights insights from the marathon itself. The value boxes at the top show key finisher stats, and the animated GIF — created from my Strava GPX file — visualizes my pace, elapsed time, and elevation changes throughout the course. Additional plots provide a closer look at elevation shifts, pace at each mile and split, and average heart rate by mile.
The second tab focuses on the training period (5/12/2025–10/25/2025) using my running data pulled from the Strava API. Value boxes summarize total runs, miles, and minutes trained. A calendar heatmap visualizes my running schedule, mileage, and pace, with darker shades indicating longer runs and light grey or white squares showing rest or sick days. Below that, histograms display the distributions of key training metrics such as distance, pace, and heart rate.
I plan to create similar dashboards for future marathons to compare training cycles and performance over time — and to use these insights to refine my training and recovery strategies.
Graduate Projects & Reports
For my Stat 484 project, I explored an R package to illustrate its utility in data analysis. I selected DataExplorer for its versatility in assisting exploratory data analysis, and used NHL Playoff data to demonstrate the package’s features.
The purpose of this project was to complete a statistical analysis in R while utilizing different R packages and functions. I analyzed housing data in R to model home prices based on various housing characteristics. I performed a linear regression, checked model assumptions, and visualized the results with multi-panel plots to explore how different features impact housing values.
For this project and report, I designed an experiment to investigate how flour type, butter consistency, and leavening agent affect the rise of vanilla cupcakes. Using a 2×2×2 factorial design, I baked and measured cupcakes under controlled conditions and applied a three-factor ANOVA to explore main and interaction effects. This analysis allowed me to quantify how ingredient choices impact cupcake height and identify combinations that produce the best rise.
For this project and report, I designed an experiment to investigate how flour type, liquid type, and leavening agent affect the height of pancakes. Using a 2×2×2 factorial design, I prepared and measured pancakes under controlled conditions and applied a three-factor ANOVA to examine main and interaction effects. This analysis allowed me to quantify how ingredient choices impact pancake fluffiness and identify the combination of seltzer water and baking powder as producing the tallest pancakes.
For this final recorded presentation, I analyzed datasets containing housing information from multiple neighborhoods to build a predictive model for home sale prices. The project involved data cleaning, feature engineering, model fitting, and model selection, and my final model achieved the most accurate predictions in the class.
For my final capstone, I analyzed sales trends to enhance inventory management strategies, optimize operational efficiency, and align staffing resources with fluctuating transaction volumes influenced by seasonal and weather patterns for a small business. Using three years of transaction data combined with local weather information, I investigated customer traffic patterns, seasonal variation in product categories, and the impact of precipitation on product sales. I applied Poisson regression, decision tree modeling, chi-square tests, and exploratory data analysis to answer these questions, providing actionable insights to support staffing, inventory, and sales planning decisions.
Undergraduate Projects & Reports
Stat 184: Predicting if a NHL Player will be inducted into the Hockey Hall of Fame
For my Stat 184 project, I explored hockey data to see what player stats predict Hall of Fame induction. Using decision tree models, I assessed odds of induction for each position, evaluated predictions for current players, and investigated why some top stars have dramatically different chances of being inducted. It was an extremely simple analysis, but it produced some interesting results.
Predicting Income and Unemployment in the United States Presentation Predicting Income and Unemployment in the United States Report Predicting Income and Unemployment in the United States R Code
For Stat 462, regression analysis, myself along with my four other group members, decided to use the 2015 American Community Survey from the Census Bureau to see if we could develop models that could predict income and unemployment at a county level. The final model for predicting income used elastic net regression, while the final model for predicting unemployment rate used LASSO regression.