Amazon SearchBox

Sunday, 8 May 2011

REGRESSION ANALYSIS OF PANEL DATA

Panel data is a combination of data cross (cross section) with time series data (time series). Panel data introduced by Howles in 1950. Time series data typically includes an object (eg stock prices, foreign exchange rates, or inflation rates), but includes several periods (can be daily, monthly, quarterly, yearly, etc.). Cross data consists of several or many objects, often called the respondent, (eg a company) with several types of data (eg employment, investment, economic growth and government spending).

The use of Eviews
To do the processing with Evies, as usual the data has been stored in Microsoft Excel. Especially for this panel data to consider how to give names to individual variables. We must put the sign "_" before the name of the individual variables. For example we want to give names to the west of Lombok district with "lobar" only, then the variable must be written with "_LOBAR".
After the requirements are met, now we'll do retrieval of data from Microsoft Excel. It should also be mentioned here that the steps that must be very different from the data import both time series data and panel data as it has been studied.
The first step to import data is not large, namely:
• Click File, New and Workfile
Furthermore, although our panel data, we must define the time (not to his cross section). For this data we choose 'Annual' and enter the Start date and End date to 2001 with 2005. then as usual click OK.
After the windows 'Workfile' open, then we have the procedures are slightly different from what we learned earlier. As for the steps:
• Click on Object
• Select New Object
After that will come some objects that can be selected as can be seen in the picture below. Given our goal is to import the data panel, then we select the 'Pool'. Next we click OK.
After that itâ will find the windows 'Pool'. Here we must fill in all the name of the individual variables that we created in Microsoft Excel.
The next step, from the windows 'Pool' are we select 'Procs', then select 'Import Pool Data (ASCII, XLS, WK?)' And look where we store the data panel
Fill in the name of the variable bound to the box 'Dependent Variable'. Once again we see how the writing, in which end with the symbol "?" Therefore, in this matter which became the dependent variable is employment (KK). While on the 'Common Coefficints' included all independent variables. Writing the same way with the writing of the dependent variable, so for this problem is written with "INV? PE? PP? ".
Dependent Variable: KK?

Dependent Variable: KK?
Method: Pooled Least Squares
Date: 09/20/08 Time: 14:05
Sample: 2001 2005
Included observations: 5
Number of cross-sections used: 7
Total panel (balanced) observations: 35
Variable Coefficient Std. Error t-Statistic Prob.
C 2001.376 0.844902 2368.766 0.0000
INV? -0.008389 0.004058 -2.067087 0.0472
PE? 2.55E-06 1.06E-06 2.420115 0.0216
PP? 1.28E-09 2.02E-09 0.634973 0.5301
R-squared 0.160629 Mean dependent var 2003.000
Adjusted R-squared 0.079400 S.D. dependent var 1.434860
S.E. of regression 1.376718 Sum squared resid 58.75596
F-statistic 1.977474 Durbin-Watson stat 0.426048
Prob(F-statistic) 0.137845


Dependent Variable: KK?
Method: Pooled Least Squares
Date: 09/20/08 Time: 14:07
Sample: 2001 2005
Included observations: 5
Number of cross-sections used: 7
Total panel (balanced) observations: 35
Variable Coefficient Std. Error t-Statistic Prob.
INV? -0.006940 0.007582 -0.915408 0.3687
PE? 1.23E-05 1.33E-06 9.209857 0.0000
PP? 9.89E-09 1.07E-08 0.921493 0.3656
Fixed Effects
_LOBAR--C 1980.388
_LOTIM--C 1984.231
_LOTENG--C 1979.788
_DOMPU--C 1987.197
_SUMBAWA--C 1994.851
_BIMA--C 1990.053
_MATARAM--C 1988.238
R-squared 0.828470 Mean dependent var 2003.000
Adjusted R-squared 0.766720 S.D. dependent var 1.434860
S.E. of regression 0.693025 Sum squared resid 12.00708
F-statistic 13.41637 Durbin-Watson stat 1.151196
Prob(F-statistic) 0.000000

No comments:

Post a Comment