Kruskal-Wallis test in Eviews6

[jwcy45]

Hello,

I would like to ask how can i compute the non-parametric KW test in Eviews?

I am wondering whether the 'Chi-squared statistic' in Wald Test (statistic results below F-statistic) representing the KW test?

Cheers

[EViews Glenn]

Not sure the context in which you are asking this question...

The KW statistic can be performed an any series in EViews using the median by-test option, or (I believe) the group testing option. But I'm not certain what you mean by the Wald Chi-Square...

[jwcy45]

Not sure the context in which you are asking this question...

The KW statistic can be performed an any series in EViews using the median by-test option, or (I believe) the group testing option. But I'm not certain what you mean by the Wald Chi-Square...

ok. I am confused with my question as well. alright its like this, i am trying to test the F-statistic of the null hypothesis (all mean being the same) against the alternative hypothesis (at least one of them are not the same). so i preform a Wald test of C(1)=C(2)=C(3)=C(4)=C(5)=C(6)=C(7)=C(8)=C(9)=C(10)=C(11)=C(12)=C(13)=C(14) (lets say 14 variables).
then part of the output results are:

Wald Test:
Equation: GJR_MOTY_WNY_SUB1

Test Statistic Value df Probability

F-statistic 1.249957 (13, 2555) 0.2368
Chi-square 16.24944 13 0.2359

so i want to ask whether the 'Chi-square' results represent the results of non-parametric KW test?

cheers

[EViews Glenn]

No, it's the asymptotic form of the Wald test.