1. Re-read Chapter #6, especially pp. 239-246 (A copy
of Ch. 6 can be located in the Ereserves folder under Business Forecasting
Chapter 6.)
2. Read my Nonlinear Functional Forms
notes located under 'Course Documents' and my 'Additional Comments to
Chapter 6' found in your Discussion Board.
a. Nonlinear Functional Forms notes---some
of this particular lecture series notes relate to multiple regression but the
beginning of the document is in terms of a bi-variate or two variable
regression model
3. Review the Panopto Recording document on '
S&P--Income--Double Logarithmic Regression'. Also, see pp.
274-277 in the textbook for a discussion of the MINITAB regression procedure.
4. Do Prob_6_11 using the above Hanke data
Excel attachment.
5. Using a combination of MINITAB and ECONOMAGIC,
select the two data sets attached above with the Standard and Poor’s (i.e.,
Index Number) as the dependent variable (Y) and Real Disposable
Income ($B) as the independent variable (X). Notice the
periodicity of the data are monthly observations and the data begin
January 1959.
(a) Estimate and interpret the un-logged
regression equation with S&P as the dependent variable. Also compute
a point and 95% prediction interval forecast for the S&P assuming
Disposable Income is $15,500B ($15.5T)
(b) Estimate a doubled logged regression
equation.
1. First transform both dependent variable (S&P) and the
independent variable (Real Disposable Income) into their natural logarithmic
counterparts and then re-estimate and interpret the regression equation.
(the attached Excel document has done the transformations and can be located in
columns F and G)
2. Report your results comparing the
two regression equations (i.e., un-logged and logged) in a
brief one or two paragraph summary.
a. Make sure you interpret the regression
coefficients, the t-tests, the F-statistic, and the R-Square for both the
unlogged and logged regressions.
3. No forecasts are necessary for this logarithmic
portion of the assignment
6. Submit 4 & 5 to this link as a single
WORD document and clearly indicate in BOLD type the
question number with the appropriate an
Get Free Quote!
265 Experts Online