You may use a calculator and MINITAB to conduct the necessary calculations for all questions.
Analysis of US GDP and GDP growth rate (1959-2004). The following variables can be retrieved from MINITAB worksheet (Exam2.MTW)
C1 year C10 trend (1,..,46) 1959=1
C2 3-mon TB rate (%) C11 VietNam_War (1965-1975=1)
C3 Unemployment rate (%) C12 Trend*VietNam
C4 Corporate Profit (billions $) C13 Post_clinton (1992-2004)=1
C5 GDP (Billions $) C14 Trend*Clinton
C6 CPI,(1982/84=100) C15 Logged-GDP
C7 Growth_rate_Corporate_profit(%) C16 Logged-Corporate-profit
C8 Growth_rate_GDP (%) C25 lag LGDP
C9 Inflation_rate (%) C26 lag_LCorportate-p
C28 lag_GDP_growthRate
Conduct the analysis using variables as suggested in each question. If you need clarification on variable definition, please ask.
(0) Use print command, PRINT C1 c10-c14 to verify the variable definitions.
(1) Draw time series plots of logged GDP (C15) and GDP growth rate (C8) with starting year =1959 separately.
10 pts Briefly describe the main features of logged GDP and GDP growth rate (%).
Logged GDP:
GDP Growth rate:(2) Model (2) Consider a regression model using Growth rateGDP(C8) as the response and 3-mon TB rate (C2), unemployment rate (C3), growth rate of Corporate profit (C7), inflation rate (C9), and trend (C10). Report the DW statistics.
5pts write down the regression equation
10 pts Interpret each of the estimated parameters 3-mon TB rate (C2),
unemployment rate (C3),
growth rate of Corporate profit (C7),
inflation rate (C9),
trend (C10)
Report the DW statistics and interpret the message of this statistics
DW =
(3) Model 3 Improve the above model by adding the variables indicating Vietnam War (1965-1975) and Clinton period (1992-2000), (C11 C12 C13 and C14). i.e. Response variable C15 and predictors include C2 C3 C7 C9 C10 C11-C14.
10 pts (3.1) Write down the regression model and mark the variables with insignificant effects.
Interpret the estimates of C11 c12 c13 and c14,
15pts (3.2) Testing the hypothesis that the join effects of trend (C10), Vietnam War(C11), Vietnam War*trend (C12), Post_Clinton (C13) and Trend*Clinton (C14) are not significant, using the following steps:
(3.2a) Calculate SS due to the above 5 variables after other variables are included in the model
(3.2b) Calculate the F-statistics
(3.2c) The critical value for degree of freedom (5,35) is 2.485. Draw your conclusion of the test based on your result in (3.2b)
10 pts (3.3) Do you see any improvements in model (3) versus model (2). Explain.
10pts (4) Add some variable(s) of your choice to model (3) to improve the status (the new model should have a DW closer to 2) indicated by the DW statistics (Currently DW=1.67). Report the regression equation where the DW statistics is satisfactory. Report the DW statistics as well.
Variable(s) Added:
DW =
Regression Equation:
10pts (5)Model 5: Use stepwise procedures to identify which variables are considered important for explaining GDP Growth Rate (C8). Use C8 as response, and predictors c2 c3 c7 c9 c10 c11 c12 c13 c14 c26 and c28.
Write down the regression equation selected in this procedure and report the adjusted R-square, s and C-P.
10pts (6) Model 6: Use stepwise procedure to identify which variables are considered important for explaining logged GDP (C15). USE C15 as response, and predictors c2 c3 c7 c9 c10 c11 c12 c13 c14 c26 and c25.
Write down the regression equation selected in this procedure and report the adjusted R-square, s and C-P.
10pts (7) Compare results in Model (5) for GDP growth rate and Model (6) for logged GDP. Write a short paragraph to summarize your analysis on US GDP growth rate and US logged GDP in the period of (1959-2004).