. Author manuscript; available in PMC: 2023 Jan 12.

Published in final edited form as: Stat Sin. 2022;32:1983–2005. doi: 10.5705/ss.202020.0248

Table 2:

Experimentation and recommendation percentages for the poSPM and the PIPE design. Column containing the MTD in bold. Scenarios 1 to 4 correspond to scenarios A, C, E, G in Mander and Sweeting (2015). they can be found in Appendix A.

Scenario	Method	Experimentation %					Acc	%DLTs
Scenario	Method	[0, 0.10)	[0.10, 0.15)	[0.15, 0.25]	(0.25, 0.30]	(0.30, 1]	Acc	%DLTs
1	poSPM	9.0	25.8	56.9	7.7	0.4	0.35	16.9
	PIPE	2.8	7.5	59.4	26.1	4.2	0.39	22.0
2	poSPM	0.0	26.4	45.0	9.4	19.1	0.88	23.5
	PIPE	0.0	6.2	21.2	11.5	61.0	0.63	36.0
3	poSPM	12.2	27.0	36.1	22.8	1.9	0.38	18.0
	PIPE	0.0	6.6	21.8	11.5	60.2	0.33	35.9
4	poSPM	31.2	19.4	32.4	13.7	3.2	0.67	14.2
	PIPE	12.3	14.0	34.0	26.0	13.4	0.55	22.1
		Recommendation %					Acc	${\bar{N}}_{M T D}$
		[0, 0.10)	[0.10, 0.15)	[0.15, 0.25]	(0.25, 0.30]	(0.30, 1]	Acc	${\bar{N}}_{M T D}$
1	poSPM	2.0	19.9	71.6	6.3	0.0	0.63	3.73
	PIPE	0.3	5.2	70.3	23.7	0.4	0.64	2.23
2	poSPM	0.0	17.8	54.3	12.8	15.0	0.92	2.60
	PIPE	0.0	8.8	38.4	17.4	35.3	0.84	1.44
3	poSPM	3.8	29.1	36.6	29.5	0.9	0.44	3.18
	PIPE	0.0	9.8	37.7	17.5	35.0	0.28	1.44
4	poSPM	20.1	22.6	39.1	16.8	1.2	0.77	4.2
	PIPE	17.4	23.5	38.0	19.1	2.0	0.77	2.77