Weeks SP	Age	Years	Pay
13	37	16	46
13	53	19	48
11	36	8	35
14	44	16	33
3	28	4	40
10	43	9	31
4	29	3	33
7	31	2	43
12	45	15	40
7	44	15	32
8	42	13	42
11	41	10	38
9	32	5	25
10	45	13	36
18	48	19	40
17	52	20	34
13	42	11	33
14	42	19	38
5	27	2	25
11	50	15	36
10	46	14	36
8	28	6	22
15	44	16	32
7	40	6	27
9	37	8	37
11	44	12	35
10	33	13	32
8	41	14	42
5	33	7	37
6	27	4	35
14	39	12	36
12	50	17	30
10	43	11	29
14	49	14	29
12	48	17	36
12	41	17	37
8	39	8	36
12	49	16	28
10	37	10	35
11	37	13	37
15	44	19	33
5	31	6	37
8	42	9	36
11	40	11	32
15	35	15	30
11	46	13	40
6	25	5	33
6	40	7	33
13	40	14	48
9	38	10	37

Question 1

Determine the regression equation:

First:Choose the Dependent Variable

A. Number of weeks of severance pay

B. Number of years with the company C. Age of employee D. Annual pay (in thousands of dollars)

Question 2 Comment on how well themodelfits the data:

One way to see if the model fits thedata or not is to see the p-value of the ANOVA table. If the p-value is smaller than 5%, it fits the data well. Otherwise, it does not fits, meaning, the severance pay has little relations of all these independent variables.

What is the p-value of the model?Tip:If the answer is1.2345E-5,means:0.000012345.for4decimals,the answer is 0.0000or 0.

Question 3 Comment on how well themodelfits the data:

The second way to see if the model fits thedata or not is to see the Adjusted R-Square which explains how much the variation forseverance pays can be explained by these independent variables as a whole.If the adjusted R squareis higher than 60%, it fits the data well (as a rule of thumb).

What is the adjusted R square value?If it is 23.45%,enter as 0.2345.(Four decimals)

Question 4 Now,we are talking about the three independent variables.Are they significant?Meaning It does explain why the severance pay is X weeks.All we need is to look at the p-value of individual variables (not the intercept):

What is the p-value of the Age of employees?

Question 5 Now,we are talking about the three independent variables.Are they significant?Meaning It does explain why the severance pay is X weeks.All we need is to look at the p-value of individual variables(not the intercept):

What is the p-value of the Number of years with FDU?

Question 6 Now,we are talking about the three independent variables.Are they significant?Meaning It does explain why the severance pay is X weeks.All we need is to look at the p-value of individual variables(not the intercept):

What is the p-value of the Annual pay?

Question 7 Which of the three independent variables is the MOST significant?

Meaning:The smallest p-value (You may have to look carefully with the decimals)

A. Annual pay B. Number of years with FDU C. Age of employee D. All at the same significance

Question 8 Now is the time to perform Bob Stranger's dispute?His 5weeks severance pay is within the 95%of the regression?

Through the above analysis,we should realizethat only the "Number of years with FDU"is the relevant(age might be relevantbut it ishighlycorrelatedwiththe number of years at FDU, right? If Bob works at FDU for 10 years, he can't be 20 years old now, right?). Annual pay might be mattered only for the dollaramount, not how many weeks.

So now, we have only one independent variable left: Number of yearsat FDU. Rerun the regression with only one column (Column C in excel file for Number of years)

What is the lowest severance pay if use lower95%coefficients?(enterweeks with 2 decimals)

Tip:Bob's severance pay (Y)=Intercept +b1*number of years at FDU, where intercept and b1should be at the lower 95%column of the regression table.

Why lower 95%?If Bob gets more pay than ex-staff,of course, he is not going to complain!

Why 95%?Because the regression is like a normal distribution and the middle is just the average and not everyone will hit in the middle. But if Bob's pay is less than 95% of the ex-staff in term of the number of weeks, he will have a case!

Question 9 Is Bob's complaint valid?Or does he has a case against FDU?

A. Yes B. No