Project
Stat 461
This project will fulfill the "written requirement" of this course. The reports should be typed and computer output included in an appendix at the end. The reports should not contain more than five pages double spaced typed pages and ten pages of output. Clear references to the output should be made whenever needed. The reports are due the last day of classes, no late reports will be accepted.
The following is a description of the data set which can be accessed by clicking here.
The data set was compiled by Brant Deppa from Winona State University and consists of several variables recorded for 77 cities in the U.S. whose population in 1990 exceeded 200,000. Unless noted the information is from the year 1990. The variables included are:1. City name
2. Population
3. The growth rate from 1980 to 1990
4. Area in square miles
5. Black population
6. Hispanic population
7. Asian population
8. Percent of students enrolled in public schools
9. Percent of adults with bachelor's degree or higher
10. Median household income
11. City government taxes per capita
12. Unemployment rate
13. Cyclical threat indicator.
Indicator of cities vulnerability to recession based on too many blue-collar and military jobs. -1 low threat, 0 average threat, 1 high threat.
14. Median value of owner-occupied housing units
15. Infant mortality rate per 1,000 live births
16. Average annual precipitation (in)
17. Sulfur dioxide average (ppm)
18. Carbon monoxide average (ppm)
19. Ozone concentration maximum (ppm)
20. Crime rate per 100,000
21. Crime trend. Indicator of crime trend (1987-1992). 1 = increase, 0 = no change, -1 decrease
22. Number of days with minimum temperature below 32 degrees F
23. Number of days with maximum temperature above 90 degrees F
The tasks to be completed are:
1) Construct a model for the rate of unemployment.
a. Explain how each variable affects the response.
b. What measures would you suggest to reduce unemployment?
c. Do a complete analysis of the model chosen (residual/influence, collinearity)
2) Do the same for crime rate.