jarque bera sample size used

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

rbriceno
Posts: 14
Joined: Wed Dec 08, 2010 1:46 pm

jarque bera sample size used

Postby rbriceno » Wed Jan 12, 2011 10:48 am

Hello friends of the forum. I'm using Eviews 7 for estimation of some time series regression. When I run Jarque Bera tests I noticed that the sample size is reduced in one observation. I don't know if this is particular to me, since my data file always have a ZERO as the final observation. It is this a possible cause of my problem?

Example: my first data series start at 1 and ends at observation 133 (with a zero value). Jarque bera is estimated with only 132 observations, instead of 133.

Thanks for your help!

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13315
Joined: Tue Sep 16, 2008 5:38 pm

Re: jarque bera sample size used

Postby EViews Gareth » Wed Jan 12, 2011 11:00 am

Could you describe how you know that only 132 observations are being used?

For example, if I run the following simple program, it is clear that all observations are being used:

Code: Select all

create u 133
series y=nrnd
series x=nrnd
equation e1.ls y c x
scalar jb =133/6*(@skew(resid)*@skew(resid) + (@kurt(resid)-3)^2/4)
show jb
show e1.hist
Follow us on Twitter @IHSEViews

rbriceno
Posts: 14
Joined: Wed Dec 08, 2010 1:46 pm

Re: jarque bera sample size used

Postby rbriceno » Wed Jan 12, 2011 11:17 am

Thanks for your mail Gareth. I have data numbered from 1 to 133 (for my dependent variable). Then I ran the regression and ask for JB (sample there is from 1 to 132). Maybe I'm not considering something more simple about the way regressions are estimated in Eviews:

Series: Residuals
Sample 1 132
Observations 132

Mean 1.18e-16
Median -0.064637
Maximum 1.142320
Minimum -0.443211
Std. Dev. 0.248322
Skewness 2.250969
Kurtosis 9.861139

Jarque-Bera 370.3847
Probability 0.000000

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13315
Joined: Tue Sep 16, 2008 5:38 pm

Re: jarque bera sample size used

Postby EViews Gareth » Wed Jan 12, 2011 11:19 am

It isn't that the Jarque-Bera is being calculated using 132 observations, rather it is that the entire regression only used 132 observations. I bet if you look at your equation output, it will say at the top that only 132 observations were used.

EViews probably dropped one observation from the regression since data wasn't available (because of lags, or missing numbers etc...)
Follow us on Twitter @IHSEViews

rbriceno
Posts: 14
Joined: Wed Dec 08, 2010 1:46 pm

Re: jarque bera sample size used

Postby rbriceno » Wed Jan 12, 2011 12:26 pm

You're right Gareth. The estimation of JB allowed me to detect one error that I made when converting excel files to eveiws file. By mistake the last observation was skipped. That was the source of one missing observation and then Eviews dropped. Thanks again for your help.


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 22 guests