Hypothesis Testing Wald

[ssk322]

Hi All,

I am currently having an issue to determine the responsiveness between two variables within my model in terms of Hypothesis Testing.

For example:

I wish to determine C(1) is more responsive than C(2). However I cannot simply input C(1)>C(2) within the Wald test.

Please advise.

Help is greatly appreciated.

Kind regards

[startz]

Input c(1)=c(2). Then look up the one-tailed critical value in a table.

[ssk322]

Input c(1)=c(2). Then look up the one-tailed critical value in a table.

Thanks Startz for your swift reply. One more question regarding this. So if the result is statistically significant, can we determine that C(1) is not more responsive than c(2) given the equation?

Kind regards

[startz]

Well, you can test that one coefficient is larger than the other. I assume that's what you want.

[ssk322]

That is exactly what I am after. So does the equation c(1)=c(2) still apply?

[startz]

Yes. Of course you have to interpret the result. If the estimated c(1) is greater than the estimated c(2) and the t-statistic from c(1)=c(2) is greater in absolute value than the one-tailed critical value, then you can conclude variable one has a bigger coefficient than variable 2.

[ssk322]

Will this also apply to Chi-Square values? My apologies, I promise that is the last question.

Kind regards

[startz]

I don't think so, because I don't think you can get one-tailed chi-squared critical values.

[ssk322]

I think you're right. As my model is a SUR I will have to interpret the result if there is significance then H0: C(1) is more responsive than c(2) will ultimately lead to a rejection of the null.

Appreciate the help Startz!