Panel Data: How to add NAs into the number of obs

[roderick]

Hello,

I have a big problem with my panel regression output. For the data I observed 54 countries from 2004-2011 (8 years).
In general I want to analyse if foreign aid cn reduce poverty so y is representing the headcount poverty in %. The explanatory
variables are x1 which shows the total amount of foreign aid the country received, x2 shows the amount of foreign aid which came from multilateral
organizations and x3 is showing the amount of foreign aid which is comming from NGOs. Each of the explanatory variables (x1, x2, x3) is divided by the country specific GDP per capita.
I took the data from the OECD data base but unfortunately I cannot get the information for every country within the whole time period. Nevertheless my prof said that I
should keep the countries even if there is not full information. So as a result I have NAs in my tables.
So now I actually should get 432 observations (54 countries over 8 yers) but because of the NAs I got an output where it was written that eviews took only 111 observations from the 432 obs.

Is there a way that eviews can take the NAs into account so that I will get 432 obs for y output????

attached I send a screenshot of my outut

thank you for every support

[EViews Gareth]

No. EViews will not make up numbers to use in your regression. If the data isn't there, EViews won't use it.

[roderick]

ok thanks a lot for solving this because I really thought there could be a way like in stata to take the NAs into account.