Measures of variation
|
Model 1a
|
Model 2b
|
Model 3c
|
Model 4d
|
---|
Community level
|
Variance (SE)
|
2.676 (0.091)*
|
0.556 (0.085)*
|
1.117 (0.074)*
|
0.479 (0.086)*
|
PCV (%)
|
Reference
|
79.2
|
58.3
|
82.1
|
ICC (%)
|
44.9
|
14.5
|
25.3
|
12.7
|
MORe
|
4.73
|
2.03
|
2.73
|
1.93
|
Model fit statistics
|
DIC (−2log likelihood)
|
4462
|
2984
|
4142
|
2953
|
AIC
|
4466
|
3050
|
4170
|
3043
|
- SE Standard Error; PCV Proportional Change in Variance, ICC Intraclass Correlation Coefficient, MOR Median Odds Ratio, DIC Deviance Information Criterion, AIC Akaike’s Information Criterion
- aModel 1 is an empty model, a baseline model without any explanatory variable
- bModel 2 is adjusted for individual-level factors
- cModel 3 is adjusted for community-level factors
- dModel 4 is final model adjusted for both individual and community-level factors
- eIncreased risk (in median) that one would have if moving to a neighborhood/cluster with a higher risk
- *P-value < 0.001