Detecting Multicollinearity - What's The R^2?

CFA Level II
#21

Paraguay Wrote: ------------------------------------------------------- > If there are only 2 independent variables pairwise > correlation is all you need. > > When adding variables past two you need to look > for significant f, non-significant t-values. +1 I answered C in the Mock and went back to cfai book and had to read that part how to detect multico. several times to find the little sentence about two independent variables and ok to use correl

#22

Hi Everybody, I need a little help on this one,

I understand the answer about the correlation on the two only independent variables to be a rational for multico on the regression (for a 2 independent variables model).

But the use of a significant F test coupled to significant variables is widely used to discard multico in the model… I would have go with C.

And to finish it all I have a Schweser’s that states the SPURIOUS correlation between 2 variables in a 3 VARIABLES model is an EVIDENCE of multico in the model, giving even more weight to coupled correlation in a more than 2 variables model… O_O, which seems absurd given that there will always be a little amount of correlation between variables in a model. And also contradicts the general discussion you all had.

************ THE SCHWESER’S*******************************

An analyst is testing to see whether a dependent variable is related to three independent variables. He finds that two of the independent variables are correlated with each other, but that the correlation is spurious. Which of the following is most accurate? There is:

A) no evidence of multicollinearity and serial correlation. B) evidence of multicollinearity but not serial correlation. C) evidence of multicollinearity and serial correlation.

Your answer: A was incorrect. The correct answer was B) evidence of multicollinearity but not serial correlation.

Just because the correlation is spurious, does not mean the problem of multicollinearity will go away. However, there is no evidence of serial correlation.

**********************************************************************************************************

Again I would stick with the F-test and Variables significance, what do you think??

Thanks again,

#23

Any up on this one?

Have a good one,