salbm (version 1.0)

PrepInferenceAssumption: Sensitivity Analysis for Binary Missing Data

Description

Given an object returned from salbmCombine this function prepares results for presentation in a table.

Usage

PrepInferenceAssumption(M, CItp=2, qt = c(0.025, 0.975), SampCIType=1, carm=2)

Arguments

M

is an object returned by the salbm function salbmCombine. It should contain confidence bands ECI and SCI and means and standard deviations from bootstraps (Main1wts, Main2wts, Samp1wts and Samp2wts).

CItp

a number ranging from 1 to 4, indicating which confidence interval to use.

qt

two values passed to quantile in producing confidence intervals.

SampCIType

type of CI for samples

carm

comparison arm

Value

a matrix which inferenceAssumption can display.

Details

PrepInferenceAssumption is used in conjunction with influenceAssumption to produce a summary of an salmb results object,

It produces summeries for each arm of the study and the difference in estimates between arms under various assumptions:

MCAR when the mean value is substituted for missing at each timepoint

missing=0 when 0 is substituted for missing at each timepoint

missing=1 when 1 is substituted for missing at each timepoint

benchmark results returned from salbm or salbmM

See Also

salbm, salbmM, influenceAssumption

Examples

# NOT RUN {
  # M is an object returned from salbm's salbmCombine function

  prep <- PrepInferenceAssumption( M )
  inferenceAssumption(prep)
# }