LR or Wald test coefficient restrictions

[Naujokas]

Hi there,

I came across rather interesting problem( I went through this forum and yet couldn't find a suitable answer). Is it possible to employ 'not equal to' sign when using Wald test on coefficient restrictions? I have tried using <> in the little text bar that appears, however, unsuccessfully.

Also, is there any way I can restrict my coefficients using Likelihood ratio? Since I have seen only Likelihood ratio tests on omitted or redundant variables.

Looking forward to your replies.

Thanks

[startz]

Hi there,

I came across rather interesting problem( I went through this forum and yet couldn't find a suitable answer). Is it possible to employ 'not equal to' sign when using Wald test on coefficient restrictions? I have tried using <> in the little text bar that appears, however, unsuccessfully.

Also, is there any way I can restrict my coefficients using Likelihood ratio? Since I have seen only Likelihood ratio tests on omitted or redundant variables.

Looking forward to your replies.

Thanks

Most two-sided hypothesis tests are designed to test a point hypothesis against a general alternative so describing the null as "not equal" doesn't make any sense.

[Naujokas]

Hi there,

I came across rather interesting problem( I went through this forum and yet couldn't find a suitable answer). Is it possible to employ 'not equal to' sign when using Wald test on coefficient restrictions? I have tried using <> in the little text bar that appears, however, unsuccessfully.

Also, is there any way I can restrict my coefficients using Likelihood ratio? Since I have seen only Likelihood ratio tests on omitted or redundant variables.

Looking forward to your replies.

Thanks

Most two-sided hypothesis tests are designed to test a point hypothesis against a general alternative so describing the null as "not equal" doesn't make any sense.

Alright, so that would you suggest to perform me on Eviews in order to achieve it? Is there any way I can overcome this problem using different techniques?

[startz]

This has nothing to do with EViews, it's the way hypothesis tests are designed.

Maybe someone can help out if you can be more specific about exactly what you're trying to test.

[Naujokas]

Thanks, I was able to figure the issue. By the way how should I specify n-lag in Eviews, for instance, βX(t-n-1) when estimating an equation?

[startz]

x(-8) means x_(t-8)

[Naujokas]

x(-8) means x_(t-8)

I really appreciate your assistance, however, I am still struggling to estimate my model. I know it should be really simplistic to do it,but without in depth Eviews knowledge it can be rather tricky. For instance, looking at the highlighted bit in the file , how would you estimate it in Eviews? I will be really grateful if you could help on that as it is basically the only thing holding me back.
For some reason I could not upload the picture , however, here is the link: http://s8.postimg.org/o4z6ln8f9/Untitled.png.
Thanks

[startz]

I'm afraid I have no idea. Perhaps someone else can chime in.

[Naujokas]

I'm afraid I have no idea. Perhaps someone else can chime in.

Thank you anyway, hopefully someone else can help me to specify this equation in Eviews.

[Naujokas]

Does anybody now the good tutorials pages then with respect to the specification of an equations using different formulas?