INTRODUCTION
Eviews is a statistics software program which can be used to analyze, plot, and print data.
For example, an economist interested in the time-series relationship between consumption, income and the interest rate might collect data on these three variables. Once the data is collected the economist could use Eviews to plot each of the three variables and then use econometric techniques to model the relationship between these variables.
Eviews is a windows based program and is very user friendly. There is also a Help menu at the top right of the Eviews Screen.
EVIEWS BASICS
1. Access to Eviews.
To access Eviews in the 486 UCSB computer lab (Phelps 1508) in a Windows 95 environment, simply click on the "classes" folder on the desktop. Then click on the Econ 101 file. An Eviews workfile will then automatically open up and all of the variables for this class will also be loaded automatically.
2. Data Analysis.
Most of your operations can be found in the QUICK option on the top toolbar. The options are:
SAMPLE - This permits you to choose the beginning and end dates of the regression. Leave blanks between dates. Ex. 1984:1 1994:4
GENERATE SERIES - You use your code names to create new variables. You have to pick a name for any new variable. Usual order of operation applies. Be sure to use sufficient paratheses to separate operations. Ex. NEWV = GDPB/GDPD < > NEWV will be added to your data set workfile.
SHOW - Lets you look at any of your variables. Highlighting any variable in the workfile will be included on the SHOW screen. For multiple variables use the CTRL key. This is an easy way to check your calculations for new variables.
GRAPH - First include your time series variable then click on OK to get a menu of graphs available. We will use Line Graphs. Choose an appropriate SCALE measure if the variables are a different order of magnitude, e.g. billions of dollars and percent. Click on SHOW OPTIONS for other graph options.
ESTIMATE EQUATION - This runs a least squares regression for you. Make sure your sample period is what you want, then follow directions.
The order of entry is important for this option. First type in the name of the dependent variable you wish to explain followed by the names of the independent variables you believe explain your dependent variable. For example if you are trying to run a regression to explain aggregate Consumption ( CONS ) patterns over time and you believe that Gross Domestic Product (GDP) and the interest rate ( R ) are the two key explanatory variables, then you would type the following:
CONS GDP R. If you wish to include an intercept (or constant term) in your regression (as you will) then you would type:
CONS C GDP R. <>
If you do not specify a data sample range in the window provided, Eviews will set it automatically to include all of the data. If you wish to only use data from say the first quarter of 1960 to the last quarter of 1990, then you would change the sample range to "1960:1 1990:4:"
Eviews will then calculate the coefficients (slopes) on these variables, the t-statistics on these coefficients, and the R2 (goodness of fit of the regression). See section 4.2 for definitions of these terms.
PRINT - Occurs several places. for printing regression or graph results click on print opitons within these sub-routines.