RESEARCH RESOURCES

Data & code.

Computational tools, research workflows, and project materials.

METHODS IN CONTEXT

From political language
to research evidence.

I use computational methods to study changes in political rhetoric and to compare discourse across speakers, countries, and time.

My projects draw on word embeddings, semi-supervised dictionaries, and cross-national datasets. Explore my GitHub profile or contact me about materials for a specific project.

Visit GitHub (opens in a new tab)
01

REPLICATION DATA · HARVARD DATAVERSE · 2025

Fossil fuel subsidy reforms

Replication data for Fossil fuel subsidy reforms have become more fragile, with Paasha Mahdavi and Michael L. Ross. The dataset is linked in the published article’s data-availability statement.

Open dataset (opens in a new tab)
02

ANALYSIS CODE · CODE OCEAN

Reproducing the published analysis

Code to replicate the study’s findings, linked in the published article’s code-availability statement.

Open analysis code (opens in a new tab)
03

ONGOING WORK

Populist rhetoric & UNGA discourse

These projects are in development. Please contact me about the underlying text corpora and analysis workflows.

Inquire about materials