to report the p-value in a table

[nesheus]

I need to report the p-value of a regression in a table in addition to the following statistics. I cannot find or know how to produce the p-value for each variable in the regression, could you help please?
I am using Eviews 6.
Thanks
best

equation eq2_2.ls dlcopp c ar(1) ar(2) ma(1) ma(2)
scalar aic2_2 = eq2_2.@aic
scalar schwarz2_2 = eq2_2.@schwarz
scalar coef2_2_2 = eq2_2.@COEFS(2)
scalar coef2_2_3 = eq2_2.@COEFS(3)
scalar coef2_2_4 = eq2_2.@COEFS(4)
scalar coef2_2_5 = eq2_2.@COEFS(5)

[startz]

EViews shows the p-value for each coefficient in the rightmost column of the regression output.

[nesheus]

I know. I want to put it in a customized table along with other statistics. So I need to program it.

[EViews Gareth]

http://forums.eviews.com/viewtopic.php?f=5&t=2549

[nesheus]

How can I calculate the p-values directly from t-statistics.

I am running an ARIMA model,

equation eq2.ls dlcopp c ar(1) ar(2) ma(1) ma(2)
I have tried the following for the second term ar(1) but it didn't work

scalar pval2= @tdist(eq2.@tstat(2),@obs(ar(1)) or scalar pval2= @tdist(eq2.@tstat(2),@obs)

how can I get the number of observations for each AR or MA series? please help.
thanks

[EViews Gareth]

you probably want to use eq.@regobs (although you might want to adjust by eq.@ncoef

[nesheus]

Should I write it like scalar pval2= @tdist(eq2.@tstat(2),@regobs(2) for the regression eq2.ls dlcopp c ar(1) ar(2) ma(1) ma(2)

Could you be more specific about the use of @regobs or @ncoef?

thanks a lot for your help.

[EViews Gareth]

Code: Select all

scalar pval2 = @tdist(eq2.@tstat(2), eq2.@regobs - eq2.@ncoefs)

[nesheus]

thanks a lot, this was very helpful.

Just a quick question. can I write @regobs or @ncoefs after the regression without attaching it to the eq2? would I get the same result?

scalar pval2 = @tdist(eq2.@tstat(2), @regobs - @ncoefs)

[EViews Gareth]

Without putting the eq2 in there, @regobs will return the number of observations used in the last regression run. Without knowing the rest of your code, I don't know if the last regression run was eq2 or not.

I think it is always best to be very specific as to which equation you are referring.