Econometrics help please!

[dafatha00]

I'm on the last problem of my homework and its got me stumped. It doesn't seem that bad but I just can't figure it out. Here's the problem.


Please show in detail how you would perform a White Test on the following regression model:

Y=alpha + Beta(1)X(i) + Beta(2)X^2(i) + Beta(3)D(i) + U(i)

where D(i) is a dummy variable. Be sure to show your auxilary regression to test for heteroscedasticity.


I'm tired and I can't think straight...been up all night trying to finish my homework. Help would be greatly appreciated! Thanks :beer:

 

[dafatha00]

up !

 

[notfred]

I don't know.

 

[wyvrn]

I think I do that in my statistics class, but not until next week. 🙁

 

[HomeBrewerDude]

what is a White Test?


i think heteroscedasity has something to do with showing that you data isn't skewed (its evenly distributed).

 

[dafatha00]

Originally posted by: yamahaXS
what is a White Test?


i think heteroscedasity has something to do with showing that you data isn't skewed (its evenly distributed).

Well not exactly. Heteroscedasticity represents unequal scatter and for each observation, there is a different error variance. For instance, say you wanted to plot the relationship between income and electricity usage. At low levels of income, there is no choice but for those people to use low levels of electricity, so the variance is low. On the other hand, at high levels of income there are those people who can use a lot of electricity because they can afford it and there are those who choose to conserve and use less. The variance here is much higher. That's what heteroscedasticity is.