nestedRanksTest_weights

Treatments, coerced to factor.  Must contain two levels.

Groups, coerced to factor, with elements in the same order
               as for <code>x</code>.

groups


<code>ntestedRanksTest_weights</code> is used by <code>nestedRanksTest</code> to
calculate group weights based on group sizes.  The number of group members
in each of the two treatment levels is determined (<code>n1</code> and <code>n2</code>)
together with their product (<code>n1.n2</code>), and the group-specific weight
is calculated by dividing <code>n1.n2</code> by the sum of <code>n1.n2</code> for all
groups.


Calculate a Mann-Whitney-Wilcoxon test for a difference between treatment levels using nested ranks.  This test can be used when observations are structured into several groups and each group has received both treatment levels. The p-value is determined via bootstrapping.  The nested ranks test is intended to be one possible mixed-model extension of the Mann-Whitney-Wilcoxon test, for which treatment is a fixed effect and group membership is a random effect.

nestedRanksTest_weights: Calculates weights for `nestedRanksTest` based on group sizes.

Description

Usage

Arguments

Value

See Also