Skip to main content
. Author manuscript; available in PMC: 2023 Sep 28.
Published in final edited form as: Biometrics. 2022 Dec 8;79(3):2551–2564. doi: 10.1111/biom.13803

TABLE 2.

Estimated odds ratio (bootstrap standard error) associated with the average treatment effect and each exposure-month treatment effect based on three models, for the data application in Section 3

Original data
Data with induced heterogeneity
Month Model 1 Model 4 Model 5 Truth Model 1 Model 4 Model 5
Average (Δ) 1.19 (0.12) 1.39 (0.25) 1.19 (0.12) 1.19 0.86 (0.09) 1.49 (0.24) 1.13 (0.21)
1 1.19 (0.12) 1.20 (0.15) 1.19 (0.11) 0.82 0.86 (0.09) 0.88 (0.11) 0.85 (0.11)
2 1.19 (0.12) 1.21 (0.18) 1.19 (0.14) 0.92 0.86 (0.09) 0.93 (0.12) 0.85 (0.11)
3 1.19 (0.12) 1.28 (0.22) 1.19 (0.13) 1.03 0.86 (0.09) 1.24 (0.20) 1.07 (0.18)
4 1.19 (0.12) 1.53 (0.35) 1.19 (0.18) 1.06 0.86 (0.09) 1.24 (0.24) 1.01 (0.18)
5 1.19 (0.12) 1.45 (0.34) 1.19 (0.14) 1.20 0.86 (0.09) 1.68 (0.39) 1.26 (0.34)
6 1.19 (0.12) 1.24 (0.35) 1.19 (0.12) 1.80 0.86 (0.09) 2.32 (0.63) 1.52 (0.46)
7 1.19 (0.12) 1.95 (1.14) 1.19 (0.18) 1.91 0.86 (0.09) 3.32 (1.70) 1.56 (0.75)