pairw.fried: Multiple pairwise comparison procedure to accompany a Friedman test.

Description

Replaces now defunct FR.multi.comp. As with ANOVA we can examine multiple pairwise comparisons from a Friedman test after we have rejected othe overall null hypothesis. However we will need to account for family-wise type I error in these comparisons which will be non-orthogonal. A conservative multiple comparison method used here is based on the Bonferroni procedure.

Usage

pairw.fried(y, x, blocks, nblocks, conf = 0.95)

Arguments

A vector of responses, i.e. quantitative data.

A categorical vector of factor levels.

blocks

A categorical vector of blocks.

nblocks

The number of blocks.

conf

The level of confidence. 1 - P(type I error).

Value

Returns a list of class = "pairw". The utility print function returns a descriptive head and a six column summary dataframe containing: 1) the type of contrast (names are taken from levels in x), 2) the mean rank difference, 3) the lower confidence bound of the true mean rank difference, 4) the upper confidence bound of the true mean rank difference, 5) the hypothesis decision rule given the prescribed significance level, and 6) the adjusted p-value.

References

Fox, J. R., and Randall, J. E. (1970) Relationship between forearm tremor and the biceps electromyogram. Journal of Applied Physiology 29: 103-108. Kutner, M. H., Nachtsheim, C. J., Neter, J., and W. Li (2005) Applied linear statistical models, 5th edition. McGraw-Hill, Boston.

Examples

Run this code

#Data from Fox and Randall (1970)
tremors <- data.frame(freq = c(2.58, 2.63, 2.62, 2.85, 3.01, 2.7, 2.83, 3.15, 
3.43, 3.47, 2.78, 2.71, 3.02, 3.14, 3.35, 2.36, 2.49, 2.58, 2.86, 3.1, 2.67, 
2.96, 3.08, 3.32, 3.41, 2.43, 2.5, 2.85, 3.06, 3.07), weights = 
factor(rep(c(7.5, 5, 2.5, 1.25, 0), 6)), block = factor(rep (1 : 6, each = 5)))

with(tremors, pairw.fried(y = freq, x = weights, blocks = block, nblocks = 6, conf = .95))