crosscut {DOS2}

Crosscut Test and its Sensitivity Analysis

Description

Computes the cross-cut test and its sensitivity analysis. The cross-cut test is a nonparametric test of dose-response correlation with good design sensitivity when used for causal inference in observational studies.

Usage

crosscut(x, y, ct = 0.25, gamma = 1, LS=FALSE)

Arguments

x

Doses of treatment.

y

Response.

ct

The quantile that defines the cross-cut. By default, the cross-cut is at the outer .25 of the data, the lower 25 percent and the upper 75 percent.

gamma

Sensitivity parameter, gamma>=1.

LS

If LS=TRUE, a large sample test is performed. If LS=FALSE, an exact test is performed. For LS=FALSE, the mh function in the 'sensitivity2x2xk' package is used. For LS=TRUE, the mhLS function in the 'sensitivity2x2xk' package is used.

Details

Performs a one-sided test of no association against positive association, together with a sensitivity analysis. The method is described in Rosenbaum (2016), used in Rosenbaum (2017). An adaptive cross-cut statistic is discussed in Rosenbaum and Small (2017); it cuts at several quantiles and picks the best. See Section 19.4 of "Design of Observational Studies"", second edition.

Value

quantiles

Quantiles that define the crosscut

table

A 2x2 table

output

Output from mh or mhLS when applied to table. The functions mh and mhLS are from the sensitivity2x2xk package. The output includes a one-sided P-value.

Note

The 'crosscut' function makes use of 'mh' and 'mhLS' from the 'sensitivity2x2xk' package.

Author(s)

Paul R. Rosenbaum

References

Rosenbaum, P. R. (2016) <doi:10.1111/biom.12373> "The crosscut statistic and its sensitivity to bias in observational studies with ordered doses of treatment". Biometrics, 72(1), 175-183.

Rosenbaum, P. R. (2017) <doi:10.1214/17-STS621> "The general structure of evidence factors in observational studies". Statist Sci 32, 514-530.

Rosenbaum, P. R. and Small, D. S. (2017) <doi:10.1111/biom.12591> "An adaptive Mantel–Haenszel test for sensitivity analysis in observational studies". Biometrics, 73(2), 422-430.

Examples

data(periodontal)
attach(periodontal)
crosscut(cigsperday[z==1],pcteither[z==1]-pcteither[z==0],ct=.2)
crosscut(cigsperday[z==1],pcteither[z==1]-pcteither[z==0],ct=.2,gamma=1.25)
crosscut(cigsperday[z==1],pcteither[z==1]-pcteither[z==0],ct=.2,gamma=1.25,LS=TRUE)
crosscut(cigsperday[z==1],pcteither[z==1]-pcteither[z==0],ct=1/3)
detach(periodontal)