# The Causality Inference of Public Interest in Restaurants and Bars on COVID-19 Daily Cases in the US: A Google Trends Analysis

**Authors:** Milad Asgari Mehrabadi, Nikil Dutt, Amir M. Rahmani  
**arXiv:** [2007.13255](https://arxiv.org/abs/2007.13255)  
**Date:** 2020-07-27  
**Summary page:** https://www.summarizepaper.com/en/paper/2007.13255v1/

## Key points

- COVID-19 pandemic has had a significant impact worldwide, with the US having the highest number of daily cases
- Study analyzed Google search trends for bars and restaurants to gain insights into the outbreak
- Objective was to determine if these search terms could predict increase in daily COVID-19 cases in the US
- Study focused on top 10 states/territories with highest and lowest daily new positive cases
- Researchers examined causation of search query trends for bars and restaurants using Pearson correlation
- Results showed that in states/territories with higher numbers of daily cases, search query trends for bars and restaurants significantly influenced daily new cases after reopening
- Example: California had peak in restaurant searches on June 7th, 2020, correlated with increase in new cases within two weeks
- Limited number of search queries considered, but Google search trends for bars and restaurants showed significant effect on daily new cases in regions with higher infections
- These influential search trends can provide additional information for predicting new case numbers in each region
- Predictive ability can assist healthcare leaders in managing and controlling impact of COVID-19 outbreaks

## Summary

The COVID-19 pandemic has had a significant impact on countries worldwide, with the United States experiencing the highest number of daily cases. To gain insights into the outbreak, this study analyzed Google search trends related to bars and restaurants using data mining methods. The objective was to determine if these search terms could predict the increase in daily COVID-19 cases in the US. The study focused on the top 10 states/territories with both the highest and lowest daily new positive cases. By examining the causation of search query trends for bars and restaurants on daily positive cases, the researchers aimed to measure their linear relation using Pearson correlation. The results revealed that in states/territories with higher numbers of daily cases, historical search query trends for bars and restaurants significantly influenced daily new cases after reopening. For example, California had a peak in restaurant searches on June 7th, 2020, which correlated with an increase in new cases within two weeks. Although only a limited number of search queries were considered, Google search trends for bars and restaurants showed a significant effect on daily new cases in regions with higher numbers of infections. These influential search trends can provide additional information for predicting new case numbers in each region. This predictive ability can assist healthcare leaders in managing and controlling the impact of COVID-19 outbreaks on society. By utilizing such information, they can better prepare for potential outcomes and take proactive measures to mitigate the spread of the virus.