I'm an undergraduate student at American University, studying Mathematics, Economics, and Political Science. My passion involves applying quantitative methods to Political and Economic policy questions, whether it be traditional statistics, Bayesian methods, or machine learning concepts.
These are some of the areas I'm currently working in:
- Adaptive Experimental Designs and Multi-Arm Bandits
- I'm currently working on analyzing the efficacy of adaptive experiments in public policy field experiments and other political science applications.
- I've developed
{whatifbandit}, an R package that re-analyzes randomized controlled trials as adaptive designs. We support Thompson Sampling and UCB1 decision algorithms, along with other customizations to tailor the simulation. - Available on CRAN. Development version on GitHub.
- Looking to expand support for more data types and modelling frameworks in future updates.
- Economic Impacts of Regional Autonomy
- Small case study involving Northern Italian Regions, and 4 German Länder to isolate the impact of regional autonomy on GDP, GDP Per Capita, and GDP Growth
- Utilizing the Regio-EU 1977-1996 economic dataset, and panel data methods such as random effects, fixed effects, and mixed-level models.
- Repo will be public soon
These are some of the projects I've completed:
- Machine Learning:
- I'm presenting at the Southern Political Science Research Conference 2026 on my paper examining the utility of using survey data and simple machine learning models to perform political donor classification. Currently looking for publication opportunities. The replication repo with the data and code is available here
- Causal Inference with Observational Data
- I've completed a paper studying the impacts of city managers on municipal turnout, using the most recent dataset available, and propensity score matching techniques to draw causal inferences from observational data. The paper was just published in the Spring 2025 edition of American University's undergraduate journal, Clocks and Clouds, and it can be found here. Repo. The repository with the code is also public.
Feel free to checkout my LinkedIn, and contact me here at noahochital@icloud.com