TABLE 2.

CPH model: simulation results for the ith tested locus in a backcross

	Effect		C.P.:	Estimated effect ± SE:
Type					% P < 0.05
EA	0.55	0	3	0.16 ± 0.12	24
	0.55	0	5	0.24 ± 0.12	50
	0.55	0	7	0.33 ± 0.13	80
	0.55	0	9	0.37 ± 0.13	82
	0.55	0	11	0.38 ± 0.13	89
LA	0	0.55	3	0.37 ± 0.13	81
	0	0.55	5	0.29 ± 0.13	61
	0	0.55	7	0.21 ± 0.13	32
	0	0.55	9	0.16 ± 0.12	23
	0	0.55	11	0.08 ± 0.13	12
IA	0.55	−0.55	3	−0.27 ± 0.12	52
	0.55	−0.55	5	−0.006 ± 0.12	7
	0.55	−0.55	7	0.06 ± 0.13	10
	0.55	−0.55	9	0.15 ± 0.13	21
	0.55	−0.55	11	0.23 ± 0.12	50
PA	0.55	0.55	3	0.59 ± 0.13	100
	0.55	0.55	5	0.59 ± 0.12	100
	0.55	0.55	7	0.57 ± 0.12	100
	0.55	0.55	9	0.61 ± 0.12	100
	0.55	0.55	11	0.58 ± 0.13	100

Estimation results using the CPH model are shown. Types EA, LA, IA, and PA refer to early-, late-, inversely, and proportionally acting QTL, respectively. C.P., change point. The standard error (SE) is empirically derived from the distribution of the parameter estimates (based on N = 100 independent experiments). % P < 0.05 refers to the percentage of the experiments that resulted in significant detection of an effect.