Geolocated Twitter Users and Polarization Correlation During COVID-19

A dataset analyzing over 50 million tweets from late 2020 in the United States and Canada. It measures subnational partisan polarization related to lockdowns, masks, and vaccines, created by Zachary Yang and last updated in April 2026. The analysis reveals a strong negative correlation between regional vaccination rates and polarization in vaccine discussions within the U.S.

Use Cases

• Modeling the correlation between online polarization and regional vaccination rates based on the described negative relationship.
• Analyzing temporal spikes in partisan discourse linked to specific political events and policy announcements mentioned in the description.
• Comparing cross-national differences in polarization dynamics between the United States and Canada as highlighted in the study.
• Investigating geographic heterogeneity in polarization patterns across conservative and liberal regions described in the analysis.

Strengths

• Based on a large-scale analysis of over 50 million tweets.
• Provides subnational geographic granularity for both the United States and Canada.
• Focuses on a critical period of polarizing discourse during the COVID-19 pandemic in late 2020.
• Links online polarization metrics to tangible public health outcomes like vaccination rates.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data reflects the discourse of politically engaged Twitter users, which may not be representative of the general population.

Provenance

Source

figshare

Collection Method

Computational analysis of Twitter data.

Time Range

Late 2020

Freshness

Last updated 2026-04-20 17:58:09; freshness should be verified.

Geography

United States and Canada