Problem 4.Consider the following time series data representing quarterly sales of dishwashers at Big Boys Appliances over the past two years:
Time
|
Sales
|
2010 Quarter 1
|
25
|
2010 Quarter 2
|
85
|
2010 Quarter 3
|
64
|
2010 Quarter 4
|
30
|
2011 Quarter 1
|
70
|
2011 Quarter 2
|
125
|
2011 Quarter 3
|
105
|
2011 Quarter 4
|
90
|
The scatter plot of the data above shows seasonality with trend. Hence, a multiple regression model is run to forecast the demand. We denote sales (Y) as the depend variable, and denote time (t, with t=1 represent the first quarter of 2010) and dummy variables (S1, S2, and S3) as independent variables. Here, we choose quarter 4 as the baseline and adopt three seasonality dummy variables, such that S1=1 represents quarter 1, S2=1 represents quarter 2, and S3=1 represents quarter 3. Answer the following questions. Use α=0.05.
-
(5 points) Use the above variable definition to code the data above.
Time
|
Sales(Y)
|
Time (t)
|
S1
|
S2
|
S3
|
2010 Quarter 1
|
25
|
1
|
|
|
|
2010 Quarter 2
|
85
|
2
|
|
|
|
2010 Quarter 3
|
64
|
3
|
|
|
|
2010 Quarter 4
|
30
|
4
|
|
|
|
2011 Quarter 1
|
70
|
5
|
|
|
|
2011 Quarter 2
|
125
|
6
|
|
|
|
2011 Quarter 3
|
105
|
7
|
|
|
|
2011 Quarter 4
|
90
|
8
|
|
|
|
-
Copy all data into Excel and run the corresponding multiple regression. Write the estimated multiple regression equation use the variables defined above. (Hint: No need to conduct backwards elimination at this point.)
-
Explain the meaning of the slopes for seasonally dummy variable S1, S2, and S3. (Note: Use actual variable name and numbers to answer the questions. Put numbers in the right context.)
-
Test the overall fitness of the model. (State the hypotheses, and use Significant F to complete your test, use short version of 7 steps.)
Which variables in the current regression model are significant? And which are not significant? Justify your answer. (short version of 7 steps)