TABLE 3.

Power comparison with results of Abecasiset al. (2000a, Table 4)

		No. of families/sample size N
One sib per family: 480/1440		Two sibs per family: 240/960		Three sibs per family: 160/800		Four sibs per family: 120/720		Five sibs per family: 96/672		Six sibs per family: 80/640		Eight sibs per family: 60/600
D′ %	No. of simulated data sets	AbAw	F1,a, F̂1,a, LRT	AbAw	F1,a, F̂1,a, LRT	AbAw	F1,a, F̂1,a, LRT	AbAw	F1,a, F̂1,a, LRT	AbAW	F1,a, F̂1,a, LRT	AbAw	F1,a, F̂1,a, LRT	AbAw	F1,a, F̂1,a, LRT
0	0.2	0.1	0.0	0.1	0.1	0.1	0.1	0.1	0.2	0.1	0.1	0.1	0.0	0.1
1,000	0.1	0.1	0.1	0.0	0.1	0.1	0.0
1,000	0.1	0.1	0.1	0.0	0.1	0.1	0.0
20,000a	0.105	0.10	0.105	0.09	0.105	0.095	0.09
20,000b	0.105	0.10	0.105	0.09	0.085	0.085	0.09
25	2.1	33.1	1.8	15.4	2.0	10.6	2.6	8.5	3.0	7.4	2.1	6.7	2.1	5.8
1,000	33.0	14.8	11.2	7.3	8.2	5.3	4.3
1,000	32.9	14.7	10.9	7.2	7.4	4.9	3.8
50	19.5	99.2	22.9	89.4	24.8	78.6	26.7	70.7	26.7	65.2	27.2	61.2	23.9	56.0
1,000	99.4	90.5	76.7	69.9	63.7	55.5	47.5
1,000	99.4	90.4	76.4	69.0	62.8	54.1	45.0
75	69.3	100	72.6	100	74.2	99.8	76.9	99.3	76.0	98.7	76.5	98.1	75.4	96.9
1,000	100	100	99.9	99.2	98.9	97.3	94.6
1,000	100	100	99.9	99.2	98.8	97.2	93.8
100	97.4	100	97.7	100	98.3	100	98.4	100	98.2	100	98.4	100	98.5	100
1,000	100	100	100	100	100	100	100
	1,000		100		100		100		100		100		100		100

In the AbAw columns, the power (%) is taken from Abecasis et al. (2000a)(Table 4). In columns 4, 6, 8, 10, 12, 14, and 16 the power (%) of F_1,a is calculated on the basis of the theoretical approximation of noncentrality parameter λ_1,a of test statistic F_1,a at a 0.001 significance level; the empirical power (%) of F̂_1,a and LRT are calculated as the proportions of 1000 or 20,000 simulated data sets that give significant results at the 0.001 significance level on the basis of F_1,a and the likelihood-ratio test statistic, respectively. The parameters are the same as those of Abecasis et al. (2000a)(Table 4): , h² = 0.1, σ² = 100, , , . In addition, and .

^aResults of the row are calculated on the basis of F̂_1,a.

^bResults of the row are calculated on the basis of LRT.