Page 1 of 1

how to deal with missing values in data set in eviews

Posted: Tue Aug 23, 2011 7:40 am
by shazeen
Hi,

I have a dataset of 4 variables for one year (2005). Need to perform OLS for this cross section data. However, there are some vales that are missing. When i input the data in eviews its treats the missing cells as NA. On performing the regression eviews seems to remove the country with even one missing value for any variable and this reduces the sample size. Also, i read in the manual that with NA eviews mixes up the values in the columns which will not result in the correct output.

A sample of the data is in the excel file attached.
If anyone knows what i can do about this please let me know.thanx

Re: how to deal with missing values in data set in eviews

Posted: Tue Aug 23, 2011 8:32 am
by startz
Deleting an observation with any missing value is called listwise deletion. It is the standard procedure in regression analysis. Anything else requires some kind of imputation for missing data.

I'm not sure what you read in the manual. EViews does not mix up values.