176 Chapter 6 Categorical Data and Chi-Square
Death Penalty
Seriousness Race Victim Yes No
1 White 2 60
Black 1 181
2 White 2 15
Black 1 21
3 White 6 7
Black 2 9
4 White 9 3
Black 2 4
5 White 9 0
Black 4 3
6 White 17 0
Black 4 0
Calculate the odds ratio of a death sentence with white versus black victims.
6.42 Fidalgo (2005) presented data on the relationship between bullying in the work force
(Yes/No) and gender (Male/Female) of the bully. He further broke the data down by job
level. The data are given below.
Bullying
Gender Job Category No Yes
Male Manual 148 28
Female 98 22
Male Clerical 68 13
Female 144 32
Male Technician 121 18
Female 43 10
Male Middle Manager 95 7
Female 38 7
Male Manager/ 29 2
Female Executive 8 1
a. Do we have evidence that there is a relationship between bullying on the job and gen-
der if we collapse across job categories?
b. What is the odds ratio for the analysis in part a?
c. When we condition on job category is there evidence of gender differences in bullying?
d. What is the odds ratio for the analysis in part c?
e. You probably do not have the software to extend the Mantel-Haenszel test to strata con-
taining more than a 2 3 2 contingency table. However using standard Pearson chi-
square, examine the relationship between bullying and Job Category separately by
gender. Explain the results of this analysis.