Correction: Biometrika (1966), 53, 73–84: ‘ Closed sequential tests for binomial probabilities’

Abstract

I am indebted to Dr L. S. Nelson for informing me of the following inaccuracies in Table 3, p. 83. Spicer's test, power function: θ = 0.65, 0.75; for 0.404, 0.809 read respectively 0.441, 0.828; average sample number: θ = 0.2–0.8; for 5.4, 6.2, 8.5, 12.3, 16.8, 18.8, 16.5 read respectively 5.0, 7.0, 10.1, 14.0, 17.7, 18.6, 16.0. Also there are several errors in notation: P. 78. Four lines below Fig. 2, for [ 1 2 ( s - 1 ) ] read [ 1 2 ( i - s ) ] ⁠ ; on the next two lines and one line from the foot, for [ 1 2 ( i - s ) ] read [ 1 2 ( i - s ) ] ⁠ . Second and third lines (foot), right-hand side, for n, read n – d– 1. P. 79. Line 2, for ∑ i - o n ( n + d + 2 i ) Θ i ( 1 - Θ ) n + d + 2 - i read ∑ i - 0 n - d - 1 ( n + 1 i ) Θ i ( 1 - Θ ) n - i + 1 ; and on line 5, for ∑ i - o n read ∑ i - 0 n - d - 1 . P.82. Line 22, for ( 1 2 ( n - s ) + i + 1 ) read ( 1 2 ( n - s ) + i + 1 i ) ⁠ ; line 23, for 1 2 ( n + s ) read 1 2 ( n - s )