. Author manuscript; available in PMC: 2010 Aug 16.

Published in final edited form as: Behav Genet. 2009 Jun 14;39(5):580–595. doi: 10.1007/s10519-009-9281-0

Table 6.

Type 1 error rates for tests of epistasis (Dominant Main Effect Pattern, Eq. 3)

Test	Error distribution	LD	α = 0.05	α = 0.01	α = 0.001
Additive only model (4)—1 df (Y)	Normal	0	0.05166	0.01046	0.00108
Cockerham model (5)—4 df (Y)	Normal	0	0.04880	0.00976	0.00098
Additive only model (4)—1 df (Blom)	Normal	0	0.04180	0.00724	0.00072
Cockerham model (5)—4 df (Blom)	Normal	0	0.04060	0.00754	0.00080
Additive only model (4)—1 df (Y)	LaPlace	0	0.05218	0.01102	0.00094
Cockerham model (5)—4 df (Y)	LaPlace	0	0.05116	0.01100	0.00124
Additive only model (4)—1 df (Blom)	LaPlace	0	0.05658^¥	0.01190	0.00110
Cockerham model (5)—4 df (Blom)	LaPlace	0	0.05244	0.01090	0.00102
Additive only model (4)—1 df (Y)	$χ_{(1)}^{2}$	0	0.05124	0.01074	0.00098
Cockerham model (5)—4 df (Y)	$χ_{(1)}^{2}$	0	0.05132	0.01248	0.00194
Additive only model (4)—1 df (Blom)	$χ_{(1)}^{2}$	0	0.22728^¥	0.08028^¥	0.01454^¥
Cockerham model (5)—4 df (Blom)	$χ_{(1)}^{2}$	0	0.20990^¥	0.07384^¥	0.01552^¥
Additive only model (4)—1 df (Y)	Weibull (1, 0.5)	0	0.05274	0.01210	0.00160
Cockerham model (5)—4 df (Y)	Weibull (1, 0.5)	0	0.05376^¥	0.01426^¥	0.00270^¥
Additive only model (4)—1 df (Blom)	Weibull (1, 0.5)	0	0.36864^¥	0.16258^¥	0.04032^¥
Cockerham model (5)—4 df (Blom)	Weibull (1, 0.5)	0	0.43908^¥	0.22136^¥	0.07180^¥
Additive only model (4)—1 df (Y)	Normal	0.3	0.48438^*	0.25542^*	0.08614^*
Cockerham model (5)—4 df (Y)	Normal	0.3	0.05158	0.01034	0.00106
Additive only model (4)—1 df (Blom)	Normal	0.3	0.55036^*	0.29514^*	0.09588^*
Cockerham model (5)—4 df (Blom)	Normal	0.3	0.03980	0.00712	0.00060
Additive only model (4)—1 df (Y)	LaPlace	0.3	0.48280^*	0.25694^*	0.08556^*
Cockerham model (5)—4 df (Y)	LaPlace	0.3	0.05238	0.01220	0.00198
Additive only model (4)—1 df (Blom)	LaPlace	0.3	0.61430^*	0.36058^*	0.13364^*
Cockerham model (5)—4 df (Blom)	LaPlace	0.3	0.04266	0.00848	0.00118
Additive only model (4)—1 df (Y)	$χ_{(1)}^{2}$	0.3	0.30734^*	0.13412^*	0.03490^*
Cockerham model (5)—4 df (Y)	$χ_{(1)}^{2}$	0.3	0.05576^¥	0.01684^¥	0.00428^¥
Additive only model (4)—1 df (Blom)	$χ_{(1)}^{2}$	0.3	0.80786^*	0.58828^*	0.29482^*
Cockerham model (5)—4 df (Blom)	$χ_{(1)}^{2}$	0.3	0.21216^¥	0.07300^¥	0.01480^¥
Additive only model (4)—1 df (Y)	Weibull (1, 0.5)	0.3	0.52270^*	0.30720^*	0.12274^*
Cockerham model (5)—4 df (Y)	Weibull (1, 0.5)	0.3	0.06164^¥	0.02498^¥	0.00992^¥
Additive only model (4)—1 df (Blom)	Weibull (1, 0.5)	0.3	0.96588^*	0.88262^*	0.67264^*
Cockerham model (5)—4 df (Blom)	Weibull (1, 0.5)	0.3	0.39810^¥	0.18200^¥	0.05258^¥

Note: 100,000 simulations

N = 200, P₁ = 0.50, P₂ = 0.50

Inflated Type 1 error rate due to misspecified model

^¥

Demonstrated significantly inflated Type1 error rate