What type of study is this?

This is a Quantitative analysis study.

September 10, 2025

Analyzing Bias in Large Language Models: A Quantitative Study Using Sentiment and Demographic Metrics

View Full Paper

Ask AI

Bookmark

Authors

Quantitative study shows demographic bias in large language models, indicating ethical implications for applications.

Demographic bias was detected in large language models, with sentiment scores showing variation by gender and race.
Sentiment analysis revealed higher polarity scores for queries about females compared to males in all models tested.
Statistically significant sentiment variations by demographics were confirmed using ANOVA across the models examined.
These findings highlight the need for methods to address bias in large language models to promote fairness.

Ramya Mandava (2025) studied this question.

View Full Paper

Ask AI

Bookmark