Skip to main content
. Author manuscript; available in PMC: 2014 Mar 14.
Published in final edited form as: Biometrics. 2013 Dec 9;70(1):153–163. doi: 10.1111/biom.12113

Table 3.

Simulation setup II: Summary of variable selection.

Methods Stepwise Pen1 Pen2 Pen3 EN1 EN2 EN3
two-stage procedure
main effects
average # of effects found 29.63 19.17 12.27 19.48 19.18 12.27 19.78
average # of true effects found 5.78 5.71 6.10 5.77 5.71 6.10 5.76
TPR 0.578 0.571 0.610 0.577 0.571 0.610 0.576
FPR 0.265 0.150 0.069 0.152 0.150 0.069 0.156
interaction effects
average # of effects found 14.09 9.85 14.09 9.80
average # of true effects found 4.10 2.23 4.10 2.20
TPR 0.342 0.186 0.342 0.183
FPR 0.002 0.002 0.002 0.002
simultaneous procedure
main effects
average # of effects found 24.29 20.23 12.27 20.23 20.24 12.27 20.27
average # of true effects found 5.95 5.93 6.10 5.93 5.93 6.10 5.93
TPR 0.595 0.593 0.610 0.593 0.593 0.610 0.593
FPR 0.204 0.159 0.069 0.159 0.159 0.069 0.159
interaction effects
average # of effects found 17.09 18.93 17.12 18.94
average # of true effects found 4.60 3.40 4.60 3.40
TPR 0.384 0.284 0.384 0.283
FPR 0.003 0.003 0.003 0.003

NOTE: TPR, true positive rate; FPR, false positive rate. The total number of main effects is 100, whereas the number of true main effects is 10. The total number of two-way interaction effects is 4950, whereas the number of true interactions is 12.