Skip to main content
. 2011 Jan 6;10(1):2. doi: 10.2202/1544-6115.1553

Table 1:

Simulation results comparing GEE and tVIM-RM with n=50, 100, 500 and σy = 1, 10: provided are the mean value (μβ) and standard error (SEβ) for β estimates over the 500 iterations, the mean standard error estimate (μSE) for the influence curve based standard error estimate from 500 iterations, and the percent of time the true β value is included in the 95% confidence interval (CI95%) based on the standard error estimate over the 500 iterations.

n=50, σy = 1 tVIM-RM GEE
Q Confounding μβ SEβ μSE CI95% β μβ SEβ μSE CI95%
true N –1.997 0.080 0.075 0.944 –1.997 0.080 0.075 0.942
true Y –1.997 0.080 0.076 0.942 –1.997 0.080 0.075 0.942
wrong N –2.005 0.079 0.083 0.962 –2.005 0.079 0.083 0.956
wrong Y –2.000 0.080 0.078 0.950 –1.735 0.055 0.057 0.002
n=50, σy = 10 tVIM-RM GEE
true N –1.990 0.251 0.238 0.944 –1.990 0.251 0.238 0.942
true Y –1.990 0.251 0.241 0.942 –1.990 0.251 0.238 0.942
wrong N –1.999 0.250 0.241 0.944 –1.999 0.250 0.241 0.940
wrong Y –1.993 0.253 0.242 0.944 –1.734 0.173 0.169 0.622
n=100, σy = 1 tVIM-RM GEE
true N –1.998 0.048 0.051 0.956 –1.998 0.048 0.047 0.936
true Y –1.998 0.048 0.052 0.956 –1.998 0.048 0.047 0.936
wrong N –2.003 0.049 0.057 0.970 –2.003 0.049 0.052 0.952
wrong Y –1.999 0.048 0.053 0.960 –1.760 0.034 0.036 0.000
n=100, σy = 10 tVIM-RM GEE
true N –1.993 0.153 0.162 0.956 –1.993 0.153 0.148 0.936
true Y –1.993 0.153 0.164 0.956 –1.993 0.153 0.148 0.936
wrong N –1.997 0.154 0.164 0.960 –1.997 0.154 0.150 0.938
wrong Y –1.994 0.153 0.164 0.960 –1.756 0.109 0.109 0.364
n=500, σy = 1 tVIM-RM GEE
true N –2.000 0.023 0.023 0.936 –2.000 0.023 0.023 0.934
true Y –2.000 0.023 0.023 0.930 –2.000 0.023 0.023 0.934
wrong N –1.990 0.024 0.026 0.962 –1.990 0.024 0.025 0.960
wrong Y –1.995 0.023 0.023 0.946 –1.746 0.016 0.017 0.000
n=500, σy = 10 tVIM-RM GEE
true N –2.001 0.074 0.074 0.936 –2.001 0.074 0.072 0.934
true Y –2.001 0.074 0.072 0.930 –2.001 0.074 0.072 0.934
wrong N –1.990 0.074 0.074 0.944 –1.990 0.074 0.073 0.942
wrong Y –1.996 0.073 0.072 0.936 –1.747 0.050 0.050 0.000