 |
StatXact Example 4
Minority Discrimination: An Analysis of Ill-Conditioned Data
The court case of Hogan v. Pierce (Gastwirth, 1984) involved the following promotion data, by race.
|
Whites |
Blacks |
| Date of Promotion |
Promoted |
Not |
Promoted |
Not |
| 7/74 |
4 |
16 |
0 |
7 |
| 8/74 |
4 |
13 |
0 |
7 |
| 9/74 |
2 |
13 |
0 |
8 |
| 4/75 |
1 |
17 |
0 |
8 |
| 5/75 |
1 |
17 |
0 |
8 |
| 10/75 |
1 |
29 |
0 |
10 |
| 11/75 |
2 |
29 |
0 |
10 |
| 2/76 |
1 |
30 |
0 |
10 |
| 3/76 |
1 |
30 |
0 |
10 |
| 11/77 |
1 |
33 |
0 |
13 |
The most notable feature of these data is that at each promotion opportunity not a single black was promoted, whereas small numbers of whites were promoted. Thus the data support an infinite upper bound on the odds ratio of being promoted for whites relative to blacks. But it is fair to ask how low the odds ratio could be. StatXact produces the following exact and Mantel-Haenszel output.
 
The Mantel-Haenszel method cannot estimate the lower bound on the odds ratio because the point and variance estimators are undefined when the observed statistic is at its maximum. However both the exact and mid-p methods can. The mid-p method claims with 95% confidence that the odds of promotion for whites is no less than 2.298 times the odds of promotion for blacks. The more conservative exact method places a lower bound of 1.819 on the odds ratio.
|
|
|
|
| |
Example Submissions
Over the years, most of the examples have come from in-house statisticians, but we would like our examples to better reflect the actual uses customers have for our products.
Therefore, we would like to ask for data from our customers for use as examples. Any data we use would be properly cited, and with your permission, could be used for examples in our manuals, in marketing material, or even for one of our live webcasts.
This is a chance to have your hard work published and made known to many other statisticians in your field and in other disciplines.
 Get Acrobat Reader
|
|
