. 2020 Dec 10;51(3):279–288. doi: 10.1007/s10519-020-10033-9

Table 3.

Effects of violating assumptions on parameter estimates

Assumption	Models	Violation	Typical biases if assumption violated
Assumption	Models	Violation	Overestimated	Underestimated
No residual stratification	0, 1, 2	Residual strat	$V_{F}$ , w, v
No $G \times A g e$ ( $δ_{o} = δ_{*}$ )	0, 1, 2	$δ_{o} > δ_{*}$	$V_{P G S }$ , $V_{L G S }$	$V_{F}$
$f_{p} = f_{m}$	0, 1, 2	$f_{p} \neq f_{m}$		$V_{F}$
No dominance effects	0, 1, 2	Dominance effects	$V_{ϵ}$
No epistasis	0, 1, 2	Epistasis	$V_{ϵ}$ (0,1), $V_{LGS}$ (2)	$V_{F}$ (2), v(2)
$r_{P G S, t 0}^{2} = h_{t 0}^{2}$	0, 1	$r_{P G S, t 0}^{2} < h_{t 0}^{2}$	$V_{F}$ , w (when AM)
${\hat{a}}^{2} = a^{2}$	2-NP	${\hat{a}}^{2} < a^{2}$	$V_{F}$ , v	$V_{LGS}$
$\frac{g}{δ} = \frac{h}{a}$	2	$\frac{g}{δ} > \frac{h}{a}$	$V_{LGS}$	$V_{F}$ , v
$\frac{w}{δ} = \frac{v}{a}$	2	$\frac{w}{δ} < \frac{v}{a}$	$V_{F}$ , v	$V_{LGS}$
$μ = 0$	0	$μ > 0$		$V_{F}$ , $V_{PGS}$ , w
Eq. AM	1e, 2e	Diseq. AM	$V_{LGS}$ (2e), $V_{F}$ (1e), w(1e)	$V_{F}$ (2e), $V_{PGS}$ (2e), w(2e), v(2e)
Diseq. AM	1d, 2d	Eq. AM	$V_{F}$ , $V_{PGS}$ (2d)	$V_{PGS}$ (1d)
No social homogamy	1, 2	Social homog.	$V_{LGS}$ , $V_{PGS}$ (2)	$V_{F}$ , w, v
No genetic homogamy	1, 2	Genetic homog.	$V_{F}$ , w, v	$V_{LGS}$ , $V_{PGS}$