- see [[220331_173513 summed badness|summed badness]], [[220331_183210 fraction bad|fraction badness]]
# parameters
- bad/goodness measures: fc, mbfc (factual, bias, mean(factual, bias), min(factual_bias)), afm (reliable, bias, mean(reliable, bias), min(reliable, bias)) (see [[different measures of domain quality]])
- goodness thresholds: $[5, 10, 15...90, 95]$
- each domain with value **below** threshold is counted as $1$ (domains above threshold not counted)
- outcome: bad domain **count** (range: $[0, \infty]$)
- winsorize user summed values (both at t1 and t0): $[0.85, 0.90, 0.95, 0.99, 1.00]$ (1.00 = no winsorizing)
- winsorizing applied **before** *filling in missing values with 0*
# results
models
- sum badness, **quasipoisson** model: `feglm(t1 ~ conditionC * t0LC | block, dt1, family = "quasipoisson", vcov = "HC1")`
- `conditionC` $[-0.5, 0.5]$
- `t0` time0/pre-campaign badness (not shown because all estimates are highly significant)
- sum badness: `log(t0 + 1)` then mean-centered
- `block` as fixed effect
- red: statistically significant effect (p < .05)
![[_count_winsorize-1.00 1.png]]
![[_count_winsorize-0.99 1.png]]
![[_count_winsorize-0.95 1.png]]
![[_count_winsorize-0.90 1.png]]
![[_count_winsorize-0.85 1.png]]