. Author manuscript; available in PMC: 2017 Oct 3.

Published in final edited form as: Genet Epidemiol. 2015 Dec 1;40(1):35–44. doi: 10.1002/gepi.21940

Table 2.

Mean (SD) of causal effect estimates from two IV models

IV strength	No IV	No missing		Missing scenario 1^a		Missing scenario 2^b

		SNP	Haplotype	SNP	Haplotype	SNP	Haplotype
Endogeneity ~10%
20%	0.39 (0.06)	0.32 (0.10)	0.32 (0.10)	0.32 (0.11)	0.32 (0.10)	0.32 (0.10)	0.32 (0.10)
10%	0.39 (0.06)	0.31 (0.15)	0.32 (0.15)	0.31 (0.15)	0.32 (0.15)	0.31 (0.15)	0.32 (0.15)
5%	0.41 (0.07)	0.33 (0.23)	0.33 (0.21)	0.33 (0.24)	0.34 (0.21)	0.33 (0.23)	0.33 (0.21)
Endogeneity ~20%
20%	0.44 (0.06)	0.31 (0.10)	0.32 (0.10)	0.31 (0.11)	0.32 (0.10)	0.32 (0.10)	0.32 (0.10)
10%	0.47 (0.06)	0.32 (0.16)	0.33 (0.16)	0.32 (0.17)	0.33 (0.16)	0.34 (0.15)	0.34 (0.15)
5%	0.48 (0.07)	0.32 (0.25)	0.36 (0.22)	0.32 (0.25)	0.36 (0.22)	0.36 (0.20)	0.37 (0.20)

Two percent of genotypes is missing for each SNP.

Two causal SNPs are completely not typed.