Model: count_time1 ~ condition x count_time0
- `glm(count _time1~ conditionC * count_time0, family = poisson(link='log'))`
- 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 (models take forever to fit when I try to include block as dummy)
- replaced missing values with 0 (no tweets)
Todos?
- [ ] winsorize outcome (count) before fitting models?
- [ ] look at *all* tweets, instead of just tweets with domains that have quality ratings
# Interaction effect - Oct 4 to Oct 24
![[coef_conditiont_combine-interact_count_Oct04-Oct24.png]]
## Probe `quality01_tweet`
```r
term results
1: (Intercept) b = 1.75, SE = 0.002, z(32884) = 763.90, p < .001, r = 0.43
2: time0C b = 0.009, SE = 1e−05, z(32884) = 898.51, p < .001, r = 0
3: conditionC b = 0.08, SE = 0.005, z(32884) = 16.37, p < .001, r = 0.02
4: time0C:conditionC b = −0.007, SE = 2e−05, z(32884) = −342.99, p < .001, r = 0
SIMPLE SLOPES ANALYSIS
Slope of conditionC when time0C = -1.794666e+01 (- 1 SD): (users who share less during pre-campaign)
Est. S.E. z val. p
------ ------ -------- ------
0.20 0.00 41.48 0.00
Slope of conditionC when time0C = -7.176886e-17 (Mean):
Est. S.E. z val. p
------ ------ -------- ------
0.08 0.00 16.37 0.00
Slope of conditionC when time0C = 1.794666e+01 (+ 1 SD): (users who share more during pre-campaign)
Est. S.E. z val. p
------- ------ -------- ------
-0.05 0.00 -10.07 0.00
```
![[coef_conditiont_combine-interact_count_Oct04-Oct24 1.png|600]]
![[coef_conditiont_combine-interact_count_Oct04-Oct24_quality04_probe2.png|600]]
# Interaction effect - Oct 11 to Oct 24
![[coef_conditiont_combine-interact_count_Oct11-Oct24.png]]
## Probe `quality01_tweet`
```r
term results
1: (Intercept) b = 1.37, SE = 0.003, z(32884) = 492.23, p < .001, r = 0.35
2: time0C b = 0.009, SE = 1e−05, z(32884) = 723.64, p < .001, r = 0
3: conditionC b = 0.08, SE = 0.006, z(32884) = 13.61, p < .001, r = 0.02
4: time0C:conditionC b = −0.007, SE = 2e−05, z(32884) = −277.18, p < .001, r = 0
SIMPLE SLOPES ANALYSIS
Slope of conditionC when time0C = -1.794666e+01 (- 1 SD):
Est. S.E. z val. p
------ ------ -------- ------
0.20 0.01 34.27 0.00
Slope of conditionC when time0C = -7.176886e-17 (Mean):
Est. S.E. z val. p
------ ------ -------- ------
0.08 0.01 13.61 0.00
Slope of conditionC when time0C = 1.794666e+01 (+ 1 SD):
Est. S.E. z val. p
------- ------ -------- ------
-0.04 0.01 -8.14 0.00
```
![[coef_conditiont_combine-interact_count_Oct11-Oct24_quality04_probe.png|600]]
![[coef_conditiont_combine-interact_count_Oct11-Oct24_quality04_probe2.png|600]]