Portfolio
Connor Armstrong
Hi! I’m Connor, an economics and data analysis enthusiast passionate about using statistical tools to uncover insights in real-world data. I recently graduated from Michigan State University with a B.S. in Economics and a concentration in econometrics and data analysis. This portfolio showcases two projects where I applied economic reasoning, programming in R, and data science methods to explore important real-world questions.
📊 Projects
1. Economic Theory vs Reality: Testing GDP Formulas
Goal: Test whether the economic identity GDP = C + I + G + (X - M) holds in real U.S. economic data from 1947 to 2023.
Data Source: FRED (Federal Reserve Economic Data)
Tools: R, pandas, data wrangling, time series analysis, visualization
🔍 In this project, I downloaded quarterly economic indicators and built a theoretical GDP estimate based on the textbook formula. I compared it to actual GDP data, plotted discrepancies over time, and identified that government spending was consistently overestimated, leading to a growing divergence.
📈 Key Findings:
- Theoretical Net Exports matched exactly with reported data.
- Theoretical GDP was higher than reported GDP in most years.
- Government expenditure appears to be the primary driver of this divergence.
2. Do NBA Players Miss More Free Throws in the Clutch?
Goal: Measure if “clutch” situations—defined as close games in the final 5 minutes—significantly impact free throw percentage.
Data Source: NBA Play-by-Play Data (2018–19 season)
Tools: R, tidyverse, string manipulation, logistic regression, fixed effects modeling
🏀 I created a dataset of all free throw attempts, engineered features such as score margin and time left, and defined a clutch indicator. Then, I ran regressions to estimate the effect of pressure on FT performance.
📊 Key Results:
- Non-clutch FT%: 76.7%
- Clutch FT%: 75.2%
- Regression analysis (w/ player fixed effects) found a small, statistically insignificant decline in FT% under pressure.
💡 Skills & Tools
- Programming: R, Python, SQL
- Analysis: Econometrics, data wrangling, statistical modeling, hypothesis testing
- Visualization: ggplot2, matplotlib
- Data Sources: FRED, NBA.com API, CSVs, public datasets
- Languages: English, Spanish (bilingual)
🌐 About Me
I’m aspiring to become a quantitative researcher or data analyst, ideally combining economics, data science, and policy analysis. I’m currently exploring master’s programs in Data Analytics or Financial Engineering (MFE) to further deepen my skills.
If you’re interested in my work or want to collaborate, feel free to connect with me on LinkedIn or check out the repositories below!
Thanks for visiting my portfolio!
📬 Contact
- 📧 armst421@msu.edu
- 📱 (734) 259-9637