. Author manuscript; available in PMC: 2021 Jan 1.

Published in final edited form as: Stat Biopharm Res. 2019 Nov 4;12(2):164–175. doi: 10.1080/19466315.2019.1677494

Table A1.

Conjunctive and marginal coverage probabilities of average 95% predicted intervals calculated under the assumption where the planned effect size of 0.2 for both of the two endpoints or the observed effect size is true at the interim monitoring at 50% information time. The planned sample size of 516 is calculated based on a one-sided test with targeted power of 80% at the significance level of 2.5%, assuming no correlation between the endpoints.

True effect size	ρ	Conjunctive coverage probability (marginal)
True effect size	ρ	Planned effect size	Observed effect size
(0.2, 0.2)	0.0	0.989 (0.995)	0.697 (0.834)
	0.3	0.990 (0.995)	0.705 (0.835)
	0.5	0.989 (0.994)	0.720 (0.835)
	0.8	0.991 (0.995)	0.757 (0.833)
	0.99	0.994 (0.995)	0.818 (0.835)

(0.1, 0.1)	0.0	0.900 (0.949)	0.694 (0.833)
	0.3	0.905 (0.949)	0.704 (0.834)
	0.5	0.911 (0.950)	0.720 (0.835)
	0.8	0.923 (0.949)	0.757 (0.834)
	0.99	0.944 (0.950)	0.817 (0.835)

(0.0, 0.0)	0.0	0.479 (0.692)	0.696 (0.834)
	0.3	0.517 (0.691)	0.704 (0.834)
	0.5	0.547 (0.691)	0.719 (0.834)
	0.8	0.601 (0.691)	0.760 (0.836)
	0.99	0.674 (0.693)	0.819 (0.836)