UFO Sightings & Political Alignment

Statistical Analysis of U.S. UFO Reports

Statistics
Machine Learning
Data Analysis
NLP and statistical analysis of 100K+ public UFO reports.
Published

December 12, 2025

Do political differences influence reported UFO sightings?

Using over 100,000 public UFO reports, I investigated whether political alignment is associated with differences in reported sightings, narrative characteristics, and reporting behavior across U.S. states.

View Source Code View Research Report View Research Poster ← Back to Projects

Role

Statistical Researcher

Domain

Public Data Analytics

Stack

R · Tidyverse · NLP

Methods

Hypothesis Testing · Regression · Text Mining

Project Snapshot

Question

Do Democratic and Republican states differ in UFO reporting behavior?

Approach

Combined statistical hypothesis testing, regression modeling, and text analysis on national UFO sighting records.

Findings

Political alignment was associated with reporting frequency, while narrative characteristics remained remarkably similar.

Overview

Thousands of UFO sightings are reported across the United States every year. While these reports are often studied individually, much less attention has been given to broader geographic and political patterns.

This project investigates whether political alignment is associated with differences in UFO reporting frequency, reported object characteristics, and witness narratives using publicly available data from the National UFO Reporting Center.

Research Questions

Reporting Frequency

Do Democratic and Republican states report different numbers of UFO sightings?

Statistical Significance

Are observed differences statistically significant or likely due to chance?

Narrative Analysis

Do witnesses describe sightings differently depending on political alignment?

Predictive Modeling

Can political alignment explain variation in reported sightings after accounting for statistical uncertainty?

Analysis Pipeline

1
🛸

Data Collection

Imported national UFO sighting records.

2
🧹

Data Cleaning

Prepared state-level observations and standardized variables.

3
📊

Statistical Testing

Performed Mann–Whitney and Chi-square hypothesis tests.

4
📈

Regression

Fit Poisson regression models to estimate reporting differences.

5
💬

Text Analysis

Compared narrative descriptions across political groups.

6

Interpretation

Summarized statistical evidence and practical implications.

End-to-end statistical workflow combining inference, modeling, and NLP.

Technical Stack

  • R
  • Tidyverse
  • Hypothesis Testing
  • Poisson Regression
  • Mann–Whitney Test
  • Chi-Square Test
  • Natural Language Processing
  • Data Visualization
  • Statistical Inference

What I Built

Data Pipeline

Prepared and cleaned a nationwide UFO reporting dataset for statistical analysis.

Statistical Analysis

Applied multiple hypothesis tests to evaluate differences between political groups.

Predictive Modeling

Built Poisson regression models to quantify the relationship between political alignment and reporting frequency.

Narrative Exploration

Analyzed witness descriptions using natural language processing techniques to compare reporting patterns.

Key Findings

🛸
100k+

UFO reports analyzed

📊
3

Independent statistical methods agreed on reporting differences

💬
NLP

Narrative descriptions were broadly similar across political groups

What I Learned

This project strengthened my experience applying statistical inference to large observational datasets. Rather than relying on a single statistical test, I combined nonparametric testing, generalized linear models, and exploratory text analysis to evaluate the same research question from multiple perspectives.

It also reinforced the importance of distinguishing statistical significance from practical significance. Even when differences exist in reporting frequency, interpretation requires careful consideration of potential confounding factors such as population size, reporting culture, and geographic variation.

Research Deliverables

The complete statistical report and research poster are available below.

View Source Code

Research Report (PDF)

Research Poster (PDF)