Log-linear regression equation also estimation of variables
Suppose which Elisabeth wants to test whether or not more hazardous jobs pay a higher salary. She goes to the bureau of labour statistics website also gets data on average annual pay by occupation also the probably of dying in work related accidents. she uses ordinary least squares to regress the log of annual pay on the probability of worker death also obtains the following results with standard error in parentheses:
log salary= 0.50+0.02 death R-squared=0.23
(0.05)
A. according to Elisabeth's estimated coefficient, explain how much does annual salary increases if the chance of dying on the job rises by 10 percentage points?
B. Nancy claims which this model is too simple also which one must also consider the cost of living, since more hazardous jobs may simply be more remote also thus require grester pay in order to attract workers. Nancy adds the distance to the nearest major metropolitan area in hundreds of miles, Distance also she finds:
Log salary= 0.40+0.01death+0.03distance
(0.004) (0.01)
R-squared= 0.56
Elisabeth claims which Nancy's findings are not as supportive as hers in finding a link between worker death also salary because Nancy's coefficient is half as large. Elucidate why, despite Elisabeth's protests, Nancy's findings are more supportive of a link between worker death also salary.