Only study 1, 2, 4 has concordance and veracity.
- Study 3: only false news
- Study 5: new covid headlines without concordance info
Negative four-way interaction! See last panel.
- `share ~ ideology * aot * veracity * concordance
```r
m2 <- brm(share ~ ideology * aot * veracity * concord + (1 + veracity * concord | study / pid) + (1 | stimulus),
data = dt1,
prior = priors,
sample_prior = TRUE,
iter = iter,
warmup = warmup,
chains = chains,
family = bernoulli(link = "logit"),
file = fname,
cores = cores
)
```
![[_4-3-2-way_interactions 1.png]]
Probe four-way-interaction
- third column (compare high vs low AOT people)
![[_probe_four-way_politics-aot-veracity-concord 1.png]]
![[_probe_three-way-discern_politics-aot-concord.png]]