See girl, I doubt there is anything major to explain in q# 2 and #4.Its plain theory.

I'll write heteroskedasticity = HSK ; serial correlation = SC & multicollinearity = MCN

#2 - Breusch- Pagan test is used for HSK and Durbin Watson test is used for SC.

[and if u ask me to explain this , I doubt you've read Quant

and if you have not, you surely dont read further. Surely I'm not being rude. Just doubtful.]

#4 - I am simply copying the content of the question.

"To correct

**HSK**, we use White corrected standard error and replace the standard error of independent variable with that error and again conduct a t-test using the original regression coefficient."

"

**SC** can be correcting by adjusting the specification of the model by incorporating a seasonal term.

For correcting

**MCN**, one or two independent variables can be omitted from the original regression equation and the regression can be conducted again."

q# 6 - In both HSK & positive SC, standard errors terms tends to be small (not always, but general case in economic and financial data) which makes computed t-stats to be larger than usual and therefore we reject null hypothesis very often incorrectly, even when null is actually true. And this is what we call type 1 error - Rejecting ture null incorrectly.

In MCN, the case is almost opposite. The standard errors of the affected coefficients tend to be large. In that case, the test of the hypothesis that the coefficient is equal to zero may lead to a failure to reject null hypothesis, which is actually false. That what we call a type 2 error.

Q#3 -

HSK Check - n*R^2 from the regression of residuals = 50*.032 = 1.6 which is less than Critical Chi-square value 3.841. So we dont reject the null that there is no HSK.

SC Check - DW stat is between dl & du , that means its inconclusive. So cant answer in favor of SC.

MCN Check - None of the individual coefficients significantly differ from zero due to higher p-value than .05 . But R-square is 74.67% , that means all coefficients together explain much of variation but not individually. Ideal condition for MCN.

p.s. - If u still need more explanation, drop me a PM.

## Bookmarks