Interpretation of the Breusch Pagan Godfrey test for heteroskedasticity

For econometric discussions not necessarily related to EViews.

Moderators: EViews Gareth, EViews Moderator

anni0308
Posts: 3
Joined: Sat Jun 02, 2018 3:04 am

Interpretation of the Breusch Pagan Godfrey test for heteroskedasticity

Dear Community

I have problems interpreting the Breusch Pagan Godfrey test for heteroskedasticity. I read the user guide but do not understand their interpretation:

http://www.eviews.com/help/helpintro.html#page/content/testing-Residual_Diagnostics.html

The Null Hypothesis states that no heteroskedasticity exists (= homoskedasticity exists).

On the bottom of the user guide they have an example with the following result:
testing.033.4.67.jpg (88.9 KiB) Viewed 3475 times

The user guide states:
This output contains both the set of test statistics, and the results of the auxiliary regression on which they are based. All three statistics reject the null hypothesis of homoskedasticity.

However, the Prob (F-Statistic) and Prob (Chi-Square) are both 0 and the Null Hypothesis should therefore not be rejected (=no heteroskedasticity exists). Why do they reject the null hypothesis?

Reviewing my data, I received the following results:

test for heteroskedasticity.PNG (56.76 KiB) Viewed 3475 times

Usually, I would reject the null hypothesis (as all values above 0,05) but is this correct? Where is the mistake?

I am grateful for any help!

Thank you very much,
Annika

startz
Non-normality and collinearity are NOT problems!
Posts: 3485
Joined: Wed Sep 17, 2008 2:25 pm

Re: Interpretation of the Breusch Pagan Godfrey test for heteroskedasticity

You are interpreting p-values backwards. A low p-value means reject the null.