# Minority Stress Experienced by LGBTQ Online Communities during the COVID-19 Pandemic

**Authors:** Yunhao Yuan, Gaurav Verma, Barbara Keller, Talayeh Aledavood  
**arXiv:** [2205.09511](https://arxiv.org/abs/2205.09511)  
**Date:** 2022-05-19  
**Summary page:** https://www.summarizepaper.com/en/paper/2205.09511v3/

## Key points

- COVID-19 pandemic has impacted minorities, especially those in the LGBTQ community due to pre-existing social disadvantages and health disparities.
- Few studies have focused on the pandemic's effects on the LGBTQ population.
- A paper titled "Minority Stress Experienced by LGBTQ Online Communities during the COVID-19 Pandemic" develops and evaluates machine learning classifiers to identify Twitter posts exhibiting minority stress unique to members of the LGBTQ population due to their sexual and gender identities.
- Best pre-pandemic and during-pandemic models show strong and stable performance for detecting posts containing minority stress.
- Anger words are strongly associated with minority stress during the COVID-19 pandemic.
- Members of the LGBTQ community experience a greater increase in cognitive words usage but worsened observable attribute in positive emotion words usage compared to a group with similar pre-pandemic behavioral attributes from the general population.
- Results have implications for public health domains and policymakers to provide adequate support, especially with respect to mental health, for members of the LGBTQ community during future crises.

## Summary

The COVID-19 pandemic has had a significant impact on the lives of minorities, particularly those in the LGBTQ community, due to pre-existing social disadvantages and health disparities. While extensive research has been conducted on the pandemic's effects on various aspects of the general population's lives, few studies have focused on the LGBTQ population. In this paper titled "Minority Stress Experienced by LGBTQ Online Communities during the COVID-19 Pandemic," Yunhao Yuan, Gaurav Verma, Barbara Keller, and Talayeh Aledavood develop and evaluate two sets of machine learning classifiers using pre-pandemic and during-pandemic datasets to identify Twitter posts exhibiting minority stress unique to members of the LGBTQ population due to their sexual and gender identities. The authors demonstrate that their best pre- and during-pandemic models show strong and stable performance for detecting posts containing minority stress. They also investigate linguistic differences in minority stress posts across pre- and during-pandemic periods, finding that anger words are strongly associated with minority stress during the COVID-19 pandemic. The authors explore how the pandemic has affected the emotional states of the LGBTQ population by adopting propensity score-based matching to perform a causal analysis. Their findings reveal that compared to a group with similar pre-pandemic behavioral attributes from the general population, members of the LGBTQ community experience a greater increase in cognitive words usage but worsened observable attribute in positive emotion words usage. These results have implications for public health domains and policymakers to provide adequate support, especially with respect to mental health, for members of the LGBTQ community during future crises.