Re: multicollinearity in regression

Reef Fish Bob

Only two very short comments on your mail:.
1) Comments you use to do without substance and not indicating WHERE and WHY are wrong in your *professorial* opinion I do not take in account. Nobody do.
2) Towards an insistent, narcissistic and maniac attitude, using inadequate and offensive terms, I react, be sure. I have the concern to do it proportionally to the attack by respect to standards you seem not to have.

Remarks that are not (never are, in your <case> in medical means) objectively directed to the actual matter that is in discussion, them, do not interest. Bob forgets very often that we are all, of us, rational beings.

In principle, in a free Forum the readers have not to justify his/hers intentions. The texts they posted, themselves, in a implicit way, do it. (for no dyslectic people, of course )
The following note is almost superfluous therefore.
1) I try to show, throughout numerical examples, that probability and statistics can be faced as an experimental science (let me say).
2) I check, every time I can do it, (always empirically) what is true and what is not about the *recipes* that the text books are full (example: the computer intensive methods).
3) My only *tool* is a home computer and my skills in programming are restricted to BASIC. It seems to be sufficient but is quite adequate to my goals: to justify what I achieve and allow the Readers to check its soundness. In the almost totality of cases I had not to ask pardon because I was wrong, but, I am always receptive to critics without prejudice.
By nature I am not dogmatic. Remember the old * René*: I know that I know nothing.

________licas (Luis A. Afonso)
.