principle component model

[shahd]

Hi all, 
I have a salary dataset with 10 variables with no problems with multicollinearity. I am seeking for advanced regression technique. Can I use principal component regression, Could anyone please advise with a type of regression that works for such data

[PaigeMiller]

---

@shahd wrote:

Hi all, 
I have a salary dataset with 10 variables with no problems with multicollinearity. I am seeking for advanced regression technique. Can I use principal component regression, Could anyone please advise with a type of regression that works for such data

---

I'm skeptical, how do you know there is no problem with multicollinearity? What analysis did you do that shows no problems with multicollinearity? It is hard for me to imagine a salary data set with 10 variables which has no problems with multicollinearity

However, if there is no problem with multicollinearity, then Principal Components will not help. You can use ordinary least squares regression.

By the way, the proper spelling is "Principal".

[Ksharp]

PROC PLS  is superb regression model .

[PaigeMiller]

---

@Ksharp wrote:

PROC PLS  is superb regression model .

---

But it doesn't help if there is no multicollinearity.