Table 4.
Results of different ICC estimation methods: Filipino data set
| Study arm | Estimation method | ρ̂ | SE(ρ̂ ) | Wald 95% CI | Linear 95% CI |
|---|---|---|---|---|---|
| Overall | ANOVA | 0.113 | 0.067 | (0.021, 0.278) | (−0.018,0.244) |
| Fleiss-Cuzick | 0.110 | 0.060 | (0.024, 0.252) | (−0.008,0.228) | |
| Pearson | 0.127 | 0.072 | (0.033, 0.303) | (−0.014,0.268) | |
| GEE model | 0.033 | 0.045 | -- | (−0.055,0.121) | |
| Random intercept logistic model | 0.070 | 0.061 | -- | -- | |
| Intervention1 | ANOVA | 0.072 | 0.100 | (−0.030, 0.351) | (−0.124,0.268) |
| Fleiss-Cuzick | 0.064 | 0.088 | (−0.030, 0.305) | (−0.108,0.236 | |
| Pearson | 0.072 | 0.102 | (−0.014, 0.381) | (−0.128,0.272) | |
| GEE model | 0.077 | 0.103 | -- | (−0.125,0.279) | |
| Random intercept logistic model | 0.102 | 0.105 | -- | -- | |
| Intervention2 | ANOVA | 0.073 | 0.083 | (−0.030, 0.288) | (−0.090,0.236) |
| Fleiss-Cuzick | 0.065 | 0.078 | (−0.031, 0.270) | (−0.088,0.218) | |
| Pearson | 0.066 | 0.083 | (−0.019, 0.312) | (−0.097,0.229) | |
| GEE model | 0.072 | 0.069 | -- | (−0.063,0.207) | |
| Random intercept logistic model | 0.095 | 0.095 | -- | -- | |
| Control | ANOVA | −0.070 | * | * | * |
| Fleiss-Cuzick | −0.076 | * | * | * | |
| Pearson | −0.070 | * | * | * | |
| GEE model | −0.067 | 0.037 | -- | (−0.141, 0.006) | |
| Random intercept logistic model | 0.000 | † | -- | -- |
*
Valid variance estimate could not be obtained; standard error and confidence interval are not available.
†
Point estimate is truncated to be 0; corresponding standard error and confidence interval are not available.