Multiple Regression Help

[sonam00]

Hi there,
I am new to E Views. Currently, I am using EViews 7 to do my research. I was wondering if anyone could help me to create the Multiple Regression as below:
LnFDI = α + β1LnGDP + β2LnGDPG + β3LnGNI + β4LnNR + β5LnEXR + β6LnINF + β7LnEXP +β8LnIMP + β9LnID + β10LnOPEN +ε
All variables are in logarithm form where:
FDI = dependent variable
GDP, GDPG, GNI, natural resource (NR), exchange (EXR), inflation (INF), imports (IMP), infrastructure development (ID) and trade openness of host country (OPEN) = independent variable
Using ordinary least square analysis (OLS). Because variables are calculated in different units such as current US$ (FDI, GDP, GNI, natural resource, exchange rate and imports), percentage (GDP growth, inflation, trade openness) and kWh (infrastructure development), therefore variables will be transferred into logarithm form as a standard process.
Note: Can you please also tell me what I can do with the missing data if I don’t want to delete them from my research because I got some missing data for GNI and ID.

Kind regards,
Sonam

[EViews Gareth]

You should read through the first three chapters of EViews Illustrated:
http://eviews.com/illustrated/illustrated.html

[sonam00]

I have read chapter 1,2,3 and be able to estimate the regression. Unfortunately, the coefficients seem wrong like -17. or 8. (is it true that coefficient must be between +1 and -1) :(
Please advice me how to correct it.
The attached file is my regression result
Thanks a lot

[startz]

You are trying to estimate 10 parameters with 21 observations. That's not likely to work well.

[sonam00]

I got 25 observations but then 2 variables got some missing data and it adjusted to 21 observations

[startz]

You need to look at your data and think about what kind of model makes sense rather than mechanically running a regression.

[sonam00]

thanks for your advice. by the way, is it true that correlation coefficient must be between +1 and -1? I have run correlation matrix but it seems fine

[startz]

Yes, correlation coefficients must be between -1 and +1. But regression coefficients are unrestricted.