Binary sequential tests

Share:

Description

Design and analyze binary sequential tests

Details

Package: binseqtest
Type: Package
Version: 1.0.2
Date: 2016-10-12
License: GPL (>=2)
LazyLoad: yes

The package creates designs for testing a series of binary responses sequentially. It allows checking after every response, or grouped sequential tests. Gives exact confidence intervals and p-values. Has an option for non-binding futility boundaries.

There are functions for creating the binary sequential boundaries or binary grouped sequential boundaries (see designOBF), creating tables of statistics (estimates, confidence intervals, and p-values) at specific stopping points in the boundary (see link{stopTable}), modifying the boundaries (see modify), and plotting the boundaries (plot-methods).

For details see Kirk and Fay (2014).

Author(s)

Jenn Kirk, Michael P. Fay

References

Kirk, JL, and Fay, MP (2014). An Introduction to Practical Sequential Inferences via Single Arm Binary Response Studies Using the binseqtest R Package. (to appear in American Statistician).

Examples

1
2
3
4
5
6
7
8
9
# create an O'Brien-Fleming-type design, with 2.5 percent error on each side with max N of 50
B<-designOBF(50)
# plot it
plot(B)
# create a table for N (total samples) values between 20 and 25
stopTable(B,Nrange=c(20,25))
# modify the boundary if you missed looks at N=30 through 35
Bmod<-modify(B,missN=30:35)
plot(Bmod)