ASSIGNMENT #1
Purpose: Familiarity with Economic Data sources and using Simple Linear Regression.
Task:
(1) Go to BEA.gov and get data on U.S. Nominal GDP, copy them (download) to your excel spreadsheet, name it GDP. Also get data on Aggregate Consumption (Personal Consumption Expenditure) – then name the column PCE. Put them along side the GDP data. Use at least 50 pair of datasets (Annually), you can use Table 1.1.5 that you can find by the following steps:
2) BEA.gov> Tool > Interactive Data > National Data (click on GDP)> “Begin using the Data”>GDP & Personal Income > Section 1 > Table 1.1.5> look at GDP and PCE ( Line 1 and 2 only); “Modify” (on the right middle space, right above the last column of the table), then change the years (Annual) from 1970 until the most current.
(3) Make sure each pair of data are lined up on the same year accordingly. Since the data is lined up in rows, it will be easier to “Transpose” it into two columns (using a feature in Excel that you can transpose the Rows into Columns using Copy and Paste: highlight the data, Ctrl+C, it will give you option for “transpose”, then Ctrl+V). Delete unnecessary rows and columns. Save it on your computer.
(4) Using the software, estimate OLS, with Equation: PCE = f { GDP}; ~ always include the “constant” (intercept). Note: GDP is total Income of a country, and PCE is total consumption/spending. Your regression output will show the US trend of Spending with regards to the Income.
(5) Using the output of your estimation, conduct t-test on your estimated parameters using 95% confidence interval (the t-values are there on the output, or you can use the *** signs, or the p-values), and the overall measure of goodness of fit (using R-sq), then make your conclusion.
(6) Interpret the meaning of the Intercept and the slope.
(7) Copy this output into your report, call this Model A; take note of the Model Selection Criteria (R-squared, Adjusted R-squared) of that regression output.
(9) Now, I would like you to do another estimation by making both measurements in percent.
You can do it in two ways: a) in Excel, you can transform your data into percent; b) transform into log-log model. Note that using Log-Log model will give you similar (slightly different due to roundups) results to using Percentage forms, which entails that in the future, do not use apply Log into dataset that is already in percentage form.
(10) Then do your OLS estimation:. Call this Model B.
(11) compare the results of Model A and B. Make your conclusion: choose the best model and give some reasonings. Write up a cover page for your assessment and conclusion, and attach your summary of outputs (NO need to include the Raw Data). Make it organized and presentable.
The post ASSIGNMENT #1 Purpose: Familiarity with Economic Data sources and using Simple Linear appeared first on essayfab.