. Author manuscript; available in PMC: 2014 Jun 25.

Published in final edited form as: Twin Res Hum Genet. 2009 Feb;12(1):8–18. doi: 10.1375/twin.12.1.8

Table 2.

Effects of Violating Assumptions on Parameter Estimates From the Classical Twin Design (CTD), Nuclear Twin Family Design (NTFD), Stealth, and Cascade Models

	Assumptions	Models	Typical biases if assumptions are not met
	Assumptions	Models	Overestimated	Underestimated
1	Either V_C = 0 or V_D = 0	CTD	V_A	V_D, V_C
2	CV(spouse) = 0	CTD	V_C	V_A, V_D
3	CV(A, C) = 0	CTD	V_C	V_A, V_D
4	Either V_S = 0, V_F = 0, or V_D = 0	NTFD	V_A	V_F, V_D
5	CV(spouse) due to primary phenotypic assortment	NTFD, Stealth	variable	variable
6	V_Epi = 0	All	V_D, maybe V_A (CTD), V_A (Stealth)	V_Epi, maybe V_A (CTD), V_S (Stealth)
7	V_A _× _C = 0	All	V_A	V_C, V_A _× _C
8	V_A _× _E = 0	All	V_E	V_A, V_A _× _E
9	V_A _× _age = 0	All	V_A (CTD), V_D (Stealth), V_S (Stealth)	V_A _× _age (CTD), V_A (Stealth), V_A _× _age (Stealth)
10	V_TW(MZ) = 0	All	V_A, V_D	V_TW(MZ)

Note: V_x = variance of X; CV (X, Y) = covariance between X and Y; X × Y = nonscalar interaction between X and Y; A = additive genetic; D = dominant genetic; F = familial environment, due to vertical transmission from parents; S = sibling environment, C = S + F = common environment; E = unique environment; Epi = epistasic; TW(MZ) = special MZ twin environment.