Hello! Sorry if question is stupid. I can’t understand example on page 343 of Curriculum. We have t-statistic 2.29 and critical t-value 1.97, so why we cannot reject the null hypothesis?
I guess they are trying to make a point (in this particular example) that errors should be first tested for Heteroscedasticity.
Agreed with above.
No point the t test if it is conditionally heteroskedastic.
Ok. Thank you guys.