Skip to main content
. Author manuscript; available in PMC: 2010 Sep 21.
Published in final edited form as: Biometrics. 2010 Sep;66(3):783–792. doi: 10.1111/j.1541-0420.2009.01323.x

Table 1.

Power calculations of the Weighted Cumulative Geographic Residual Test and Weighted Spatial Scan Statistic for different weights and effect size.

Scenario CumGeoRes Spatial Scan*
Weight Var σZ2 c Pow Sens Accuracy Pow Sens Accuracy
η0 ηZ
Change Effect Size 1 1 1 0.5 0.20 0.76 0.45 0.25 0.60 0.50
1 1 1 1.0 0.68 0.90 0.46 0.88 0.92 0.89
1 1 1 1.1 0.77 0.91 0.46 0.93 0.82 0.84
1 1 1 1.2 0.86 0.93 0.45 0.98 0.88 0.89
1 1 1 1.3 0.90 0.95 0.43 0.99 0.91 0.92
1 1 1 1.4 0.94 0.97 0.41 0.99 0.93 0.94
1 1 1 1.5 0.96 0.98 0.40 1.00 1.00 1.00
1 1 1 2.0 1.00 1.00 0.33 1.00 1.00 1.00
1 1 1 3.0 1.00 1.00 0.30 1.00 1.00 1.00

Change Z* Weight 1 2 1 1.5 1.00 1.00 0.32 1.00 0.99 0.99
1 10 1 1.5 1.00 1.00 0.33 1.00 0.97 0.95
1 100 1 1.5 0.17 0.49 0.56 0.34 0.65 0.56
1 1000 1 1.5 0.03 0.37 0.67 0.12 0.43 0.40

Change Z*c Weight 2 1 1 1.5 0.57 0.88 0.47 0.99 0.99 0.98
10 1 1 1.5 0.07 0.57 0.32 0.08 0.56 0.44
100 1 1 1.5 0.05 0.25 0.13 0.05 0.21 0.14
1000 1 1 1.5 0.05 0.22 0.13 0.05 0.19 0.40

Change Z* Variance 1 1 2 1.5 0.91 0.97 0.42 0.99 0.95 0.98
1 1 10 1.5 0.55 0.89 0.48 0.76 0.66 0.90
1 1 100 1.5 0.15 0.78 0.57 0.44 0.41 0.75
1 1 1000 1.5 0.08 0.77 0.61 0.34 0.35 0.67

Change All Weights 10 10 1 1.0 0.70 0.89 0.47 0.88 0.75 0.76
100 100 1 1.0 0.66 0.89 0.46 0.85 0.73 0.74
5000 5000 1 1.0 0.67 0.91 0.46 0.84 0.72 0.73
10000 10000 1 1.0 0.65 0.90 0.47 0.88 0.76 0.77
*

Results for Spatial Scan Statistic when available were done and reported by Huang et al. (2009). Otherwise, were performed through simulations as outlined for the Cumulative Geographic Residual Method.

Model Framework: YiZiN(c2Zi,σZ2Zi+Zic),wiZi=η0Zic+ηZZi, and Zic=1Zi

power=11000j=11000I(Pj<0.05)

sensitivity=11000j=11000I(Pj<0.05)I({CritRegjZ})

accuracy=11000j=11000I(Pj<0.05)1CritRegjk=1CritRegjI({CritRegjkZ})