Model: count ~ condition
- `glm(count ~ condition, family = poisson(link='log'))`
- replaced missing values with 0 (no tweets)
- each panel is one type of tweet
- each row is one measure of domain quality: see [[different measures of domain quality]]
- haven't accounted for blocks
Todos?
- [ ] winsorize outcome (count) before fitting models?
- [ ] look at *all* tweets, instead of just tweets with domains that have quality ratings
# Condition effect - Oct 4 to Oct 24
Treatment condition tweeted less frequently?
![[coef_conditiont_combine-count_Oct04-Oct24 3.png]]
# Condition effect - Oct 11 to Oct 24
![[coef_conditiont_combine-count_Oct11-Oct24 1.png]]