In: Statistics and Probability
After ten years on the job, Drew is trying to decide whether to go back to school and get a masters degree. He performs a cost-benefit analysis to determine whether the cost of attending school will be covered by the increase in salary he will received after he attains his degree. He does research and compiles data on annual salaries in his industry, health care, along with the years of experience for each employee and whether the employee has a master’s degree. Earning his master’s degree will require him to take out about $20,000 in student loans. He has decided that if his analysis show that the degree will increase his salary by at least $10,000, he will enroll in the program.
The data used in his analysis are the following:
Salary ($) |
Years of Experience |
Master's Degree |
Master’s Dummy |
37620 |
22 |
No |
0 |
67080 |
27 |
Yes |
1 |
31280 |
15 |
No |
0 |
21500 |
2 |
No |
0 |
75120 |
28 |
Yes |
1 |
59820 |
25 |
Yes |
1 |
40180 |
15 |
Yes |
1 |
81360 |
32 |
Yes |
1 |
35080 |
19 |
No |
0 |
36080 |
12 |
Yes |
1 |
36680 |
22 |
No |
0 |
29200 |
1 |
Yes |
1 |
33040 |
18 |
No |
0 |
30060 |
14 |
No |
0 |
53300 |
21 |
Yes |
1 |
22820 |
7 |
No |
0 |
72900 |
31 |
Yes |
1 |
55920 |
22 |
Yes |
1 |
19280 |
0 |
No |
0 |
26000 |
7 |
No |
0 |
Drew added a dummy variable to indicate whether the employee had a master’s degree. The variable “Master’s dummy” took the value 1 if the employee had a master’s degree, 0 if not. He then regressed Salary on Years of Experience and the Master’s degree dummy. What results does he find? Show in JMP or Excel.
1. Conduct an F-test for the overall significance of the model.
2. Conduct a t-test to determine whether there is a statistically significant relationship between years of experience and salary, holding master’s degree constant.
3. Conduct a t-test to determine whether there is strong evidence that a Master’s degree increases salary by $10,000 or more, holding years of experience constant.
4. Based on the regression results, would you recommend that Drew pursue a Master’s degree? Why or why not?
Using Excel, go to Data, select Data Analysis, choose Regression. Put Salary in Y input range and Years of Experience and Master's Dummy in X input range.
Regression Statistics | |||||
Multiple R | 0.961266541 | ||||
R Square | 0.924033362 | ||||
Adjusted R Square | 0.91509611 | ||||
Standard Error | 5606.872279 | ||||
Observations | 20 | ||||
ANOVA | |||||
df | SS | MS | F | Significance F | |
Regression | 2 | 6500623195 | 3.25E+09 | 103.3912 | 3.05699E-10 |
Residual | 17 | 534429284.9 | 31437017 | ||
Total | 19 | 7035052480 | |||
Coefficients | Standard Error | t Stat | P-value | ||
Intercept | 12172.34146 | 2577.149149 | 4.723181 | 0.000196 | |
Years of Experience | 1362.195122 | 148.4359557 | 9.176989 | 5.37E-08 | |
Master’s Dummy | 15772.68293 | 2827.305592 | 5.578698 | 3.33E-05 |
1. H0: β1 = β2 = 0
H1: At least one βi ≠ 0
p-value = 0.000
Since p-value is less than 0.05, we reject null hypothesis and conclude that at least one βi ≠ 0. The model is significant.
2. H0: β1 = 0
H1: β1 ≠ 0
p-value = 0.000 (5.37E-08)
Since p-value is less than 0.05, we reject null hypothesis and conclude that β1 ≠ 0.
There is a statistically significant relationship between years of experience and salary, holding master’s degree constant.
3. H0: β2 = 0
H1: β2 ≠ 0
p-value = 0.000 (3.33E-05)
Since p-value is less than 0.05, we reject null hypothesis and conclude that β2 ≠ 0.
There is a strong evidence that a Master’s degree increases salary by $10,000 or more, holding years of experience constant.
4. Slope of Master's dummy = 15772.683
This means that if employee has Master's degree, salary increases by 15722.683.
So, Drew should pursue a Master’s degree.