| We report below four indicators of emotional well-being based on the textual content of the tweets of a fixed population (39,970 users) located in France: Anger, Anxiety, Sadness and Mental Health. |
| Anger, Anxiety and Sadness are constructed from the averaged daily frequency (across users) of words in the tweets that can be matched to the lexical field of anger, anxiety and sadness, respectively. |
| Mental Health is constructed from the averaged daily frequency (across users) of all words that are matched to the lexical field of anger, anxiety and sadness in all tweets. |
| Increases in these series are associated with a deterioration of the corresponding indicator. |
| Data span from January 1st 2019 to March 22nd 2023. |
