Data (ranks): Salad Dressings (Critchlow and Fligner)

Description

The dataset contains the rankings of four salad dressings concerning tartness by 32 judges, with values ranging from 1 (most tart) to 4 (least tart).

Usage

salad

Format

A data frame with 32 observations on 4 variables (A, B, C, D) each representing a different salad dressing.

References

Douglas E. Crichtlow and Michael A. Fligner (1991), Paired Comparison, Triple Comparison, and Ranking Expermients As Generalized Linear Models, and Their Implementation on GLIM. Psychometrika Vol. 56, No. 3

Examples

# Example for object covariates
# fit object covariates:
# salads A - D have varying concentrations of acetic and gluconic acid.
# The four pairs of concentrations are
# A = (.5, 0), B = (.5, 10.0), C = (1.0, 0), and D = (0, 10.0),

conc <- matrix(c(.5,0, .5,10, 1,0, 0,10), ncol = 2, byrow = TRUE)
sal  <- patt.design(salad, nitems = 4, resptype = "ranking")
X    <- as.matrix(sal[, 2:5]) 

glm(y ~ X, data = sal, family = poisson)