[Joao Santos Silva]

The distance data that you have here does not correspond to the distance data for which you presented summary statistics above

[Raisa Yuliana]

something wrong with my data, my model, or the way I run STATA? *_*

[Joao Santos Silva]

Something wrong with how you created the logs...

[Raisa Yuliana]

so, the problem with my model?

[Joao Santos Silva]

No, with the way you transformed the data.

[Raisa Yuliana]

The formula to change the number to their natural logarithm is ln(number), right? >,<

[Joao Santos Silva]

That is right. If you want us to help you, you need to show us the do file where you read the data, transform it, and run the regressions.

Best wishes,

Joao

[Eric de Souza]

The distant variable in your dataset is constant and so is the logarithm of distance (as it should be). It is equal to 9.864766994, and not what you have in your Stata output. You've done something wrong.

[Raisa Yuliana]

Dear Joao and Eric,
Thank you for both of you. I will check my data and re-change it to the log form. I hope this time my way to change the data is right.
-Raisa-

[Raisa Yuliana]

I re-did my data and like both of you said something wrong with my way change the data to the log form. After I re-did it, I get the result not like in my statistic result before. *_* This is my data and after i change it to log form. It is indeed different with the statistic result before. -Raisa-

[Joao Santos Silva]

This looks much better :-)

Joao

[Raisa Yuliana]

can I still input the distance in my model after OLS regression? lnE=27.85232 + 8.610551lnPOPi + 0.8550523lnGDPi - 0.1475834lnGDPc - 10.40047lnPOPc + 0lnDic + 0.255134ACFTA the equation will be like this right?
-Raisa-

[Eric de Souza]

No, you cannot. The Stata output is telling you that the distance variable is omitted (because it is a multiple of the constant). It is not telling you that the coefficient is equal to zero.
Another point: João told you above that the RESET was not a test of omitted variables. This is correct. It is Stata that leads you astray by calling the test ovtest and adding that the null hypothesis is "no omitted variables".

[Raisa Yuliana]

Is the coefficient value for distance is 0 in the result? So, it is not mean that the coefficient equal to 0 in the real? About the RESET test, Joao told me that test for the fitted model that we have.

[Eric de Souza]

The coefficient is zero in the sense that it is omitted, not in the sense that it is estimated to be zero. It would be better just to state that it is omitted because of collinearity.