pairwiseOrdinalIndependence function

Pairwise tests of independence for tables with one ordered nominal variable

Conducts pairwise tests for a 2-dimensional table, in which one variable is ordered nominal and one variable is non-ordered nominal. The function relies on the coin package.

pairwiseOrdinalIndependence(
  x,
  compare = "row",
  scores = NULL,
  method = "fdr",
  digits = 3,
  ...
)

Arguments

• x: A two-way contingency table. One dimension is ordered and one is non-ordered nominal.
• compare: If "row", treats the rows as the grouping variable. If "column", treats the columns as the grouping variable.
• scores: Optional vector to specify the spacing of the ordered variable.
• method: The method to adjust multiple p-values. See stats::p.adjust.
• digits: The number of significant digits in the output.
• ...: Additional arguments, passed to stats::chisq_test.

Returns

A data frame of comparisons, p-values, and adjusted p-values.

Examples

### Independence test for table with one ordered variable
data(Breakfast)
require(coin)
chisq_test(Breakfast,
           scores = list("Breakfast" = c(-2, -1, 0, 1, 2)))
PT = pairwiseOrdinalIndependence(Breakfast, compare = "row")
PT
cldList(comparison = PT$Comparison, 
        p.value    = PT$p.value, 
        threshold  = 0.05)
        
### Similar to Kruskal-Wallis test for Likert data
data(PoohPiglet)
XT = xtabs(~ Speaker + Likert, data = PoohPiglet)
XT
require(coin)
chisq_test(XT,
           scores = list("Likert" = c(1, 2, 3, 4, 5)))
PT=pairwiseOrdinalIndependence(XT, compare = "row")
PT
cldList(comparison = PT$Comparison, 
        p.value    = PT$p.value, 
        threshold  = 0.05)

References

https://rcompanion.org/handbook/H_09.html

See Also

pairwiseNominalIndependence

Author(s)

Salvatore Mangiafico, mangiafico@njaes.rutgers.edu