multinom.ci.test {lcd}R Documentation

Conditional independence test for multinomial data

Description

Performs a conditional independence test for variables u and v with conditioning set cond.

Usage

multinom.ci.test(tb, u, v, cond = c())

Arguments

tb a frequency table summarizing the data, an object of class freq.tb.
u name of one variable under investigation.
v name of the other variable under investigation.
cond a vector of variable names indicating the conditioning set in the test.

Details

The test is based on the deviance test in log-linear model for count data. The null hypothesis is that all interaction terms involving both u and v vanish.

Value

deviance the deviance of the saturated model from the model under null hypothesis.
df the degrees of freedom for the deviance.
p.value the approximate p-value for the test based on Chi-square approximation.

Author(s)

Zongming Ma and Xiangrui Meng

References

Agresti, A. (2002). Categorical Data Analysis. 2nd Ed. John Wiley and Sons. Hoboken, NJ.


[Package lcd version 0.7-2 Index]