Skip to main content

Table 3 Average efficiency relative to an updating interval of 20 participants according to the initial values for the acceptance probabilities

From: Preference-adaptive randomization in comparative effectiveness studies

Updating interval

Initial values for acceptance probabilities

 

Correct

More extreme

Less extreme

1

1.20

1.20

1.21

3

1.11

1.11

1.11

5

1.09

1.07

1.08

10

1.03

1.04

1.03

  1. Correct: Equal to the true acceptance probabilities. More extreme: Unequal to the true acceptance probabilities, with the initial value for the arm with the largest acceptance probability set to 1.2 times its true value and the remaining initial values set to 0.8 times their true values. Less extreme: Unequal to the true acceptance probabilities, with the initial value for the arm with the largest acceptance probability set to 0.8 times its true value and the remaining initial values set to 1.2 times their true values.
Back to article page

Trials

ISSN: 1745-6215