. Author manuscript; available in PMC: 2019 Feb 20.

Published in final edited form as: Stat Med. 2017 Oct 19;37(4):519–529. doi: 10.1002/sim.7528

Table 4.

Simulated coverage on μ for post-validation interval of the form y ± s, using n=10,000 replications. Ideal coverage is 68.27% and is theoretically acheived when the constant SD assumption holds using $s = t_{m (n - 1)}^{- 1} (0.8413) \hat{σ}$ (equation 2). We also try s = σ̄(y, 0.80) and s = σ̄(y, 0.90) for 4:6:80% and 4:6:90% procedures.

(σ₁, σ₂, σ₃, σ₄)

s = t_{m (n - 1)}^{- 1} (0.8413) \hat{σ}

s= σ̄(y, 0.80)

s = σ̄(y, 0.90)

(1.0, 1.0, 1.0, 1.0)

68.3

75.5

79.4

(1.0, 1.0, 0.9, 0.9)

65.9

73.3

77.3

(1.0, 0.9, 0.8, 0.7)

61.1

69.6

73.6

(1.0, 1.0, 0.5, 0.5)

57.0

69.2

73.2