Validating Daily Social Media Macroscopes of Emotions

Explore a 44-minute lecture on validating social media macroscopes for emotion analysis. Dive into the methodological challenges of using sentiment analysis on social media text to study emotions at a macroscopic level. Learn about a large-scale survey conducted at an online newspaper to compare user self-reports of affective states with sentiment analysis of user discussions. Discover the replication study using Twitter data and the strong correlations found between text analysis results and self-reported emotions. Examine the combination of supervised and unsupervised text analysis methods for accurate emotion measurement. Investigate the application of these macroscopes in studying the relationship between COVID-19 cases and emotions. Gain insights into the potential of social media text to track macro-level dynamics of affective states, complementing traditional survey methods. Understand how this research contributes to the World Happiness Report 2022 and its implications for future studies in affective science and social media analysis.

Introduction
Background
Word Count
Classifiers
Bad name
Evidence
World Happiness Report 22
Data sources
Bot filter
Survey
Baseline correction
Extended analysis
Analysis
External validation
Summary
Publication
Conclusion
Negative sentiment