In article <anetqm$[EMAIL PROTECTED]>,
 [EMAIL PROTECTED] (Herman Rubin) writes:
> In article <anb426$9r9$[EMAIL PROTECTED]>,
> Mark <[EMAIL PROTECTED]> wrote:
> > I'm been told that I should normalise all my variables.
> > I'm using Tobin's Q as my dependent variable which simplified is the market
> > value over the total assets.
> > It has been suggested that therefore all my independent variables should be
> > normalised by being over assets.
> 
> > Is this correct?
> 
> I would state otherwise; NEVER do any processing to your
> data not determined by the problem.  Never normalize;
> never let the computer do your thinking for you.

Herman, read the question again. Mark is not asking about
transforming variables to have normal distributions. He's
asking about financial entities that vary in size (total
assets), and whether the problem requires that the variables
be expressed as relative to total assets. That's a perfectly
reasonable question indicating that he's doing some thinking.

It would help if Mark told us what the other variables _are_,
but it is quite plausible that variables that would be
expected to scale with total assets should in fact be
normalised by total assets.

-- 

Warren S. Sarle       SAS Institute Inc.   The opinions expressed here
[EMAIL PROTECTED]    SAS Campus Drive     are mine and not necessarily
(919) 677-8000        Cary, NC 27513, USA  those of SAS Institute.
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to