Chapter 5: Mosaic Displays for n-Way Tables


Summary

Mosaic displays help to visualize the pattern of associations among variables in two-way and larger tables. Extensions of this technique can reveal partial associations, marginal associations, and shed light on the structure of loglinear models themselves.

Contents

5.1. Introduction
5.2. Two-way tables
5.3. The strucplot framework
5.4. Three-way and larger tables
5.5. Model and plot collections
5.6. Mosaic matrices for categorical data
5.7. 3D mosaics
5.8. Visualizing the structure of loglinear models
5.9. Related visualization methods
5.10. Chapter summary
5.11. Lab exercises

Selected figures

view R code
  • Figure 5.1

    Basic mosaic display for hair color and eye color data. The area of each rectangle is proportional to the observed frequency in that cell, shown as numbers.
  • Figure 5.4

    Two-way mosaic for hair color and eye color, reordered. The eye colors were reordered from dark to light, enhancing the interpretation.
  • Figure 5.13

    Mosaic displays for other models fit to the data on hair color, eye color, and sex. Left: Mutual independence model; right: Conditional independence of hair color and eye color given sex.
  • Figure 5.16

    Four-way mosaics for the PreSex data. The left panel fits the model [GPE][M]. The pattern of residuals suggests other associations with marital status. The right panel fits the model [GPE][PEM].
  • Figure 5.19

    Mosaic displays for the employment status data, with separate panels for cause of layoff.
  • Figure 5.24

    Mosaic matrix of the UCBAdmissions data showing bivariate marginal relations.