In: Statistics and Probability
18-E3.
Results of multiple regression for expend
Summary measures
Adj R-Square |
71.5% |
|||||
Model Error |
4.877 |
|||||
Regression coefficients
Coefficient |
Std Err |
t-value |
P-value |
|||
Constant |
-7.53 |
4.27 |
-1.76 |
0.08 |
||
Age |
3.786 |
0.260 |
14.56 |
0.000 |
||
Age 2 |
-0.041 |
0.004 |
-10.72 |
0.000 |
a) The regression equation is:
Salary = -7.53 + 3.786*Age - 0.041*(Age)^2
Predicted Salary (Age = 30) = -7.53 + 3.786*30 -
0.041*30*30
Pred. Salary = 69.15
b) Age associated with highest salary:
Differentiating Salary wrt Age,
dS/dA = 3.786 - 2*0.041*age
Age = 3.786/0.082 = 46.17
Double differentiation is negative, hence it is a maxima
Therefore, the age associated with highest salary is 46
c) The model explains 71.5% variation in the salary. Hence, the model is pretty good in predicting salary from someone's age alone.
d) If this data was used in an age discrimination case, the arguments one could provide are that the model is not an exact fit for predicting salary by age. Hence, we cannot say that age predicts an individual's salary at all times. There are exceptions when an individual has higher than expected salary irrespective of their age. Therefore, there is no age discrimination of salary, this purely depends on the individual