I don't think being able to integrate the formula (which
isn't possible anyways) is necessary.

Better to teach non Statistics degree practitioners
(like social scientis) the more general theory.

Think of Statistics as medicine (not my original idea).

Although learning particle physics (integrating the PDF)
may not make a med. student a worse doctor, it really won't
help them.  Better that they get a good general overview
of all the subjects relevant to their field (chemistry, anatomy,
physic, etc.)

I agree a lot of stuff being pushed out is sketchy.
And, excep for maybe the PhD in Dance or Music Performance, what
academic folks
don't use Statistics? So, the social scientist, for example, SHOULD
(though they certainly don') learn general methods (General linear
models, Discrete models, etc.).

Problem is, they only learn a few narrow areas and apply them TO
DEATH.
So, it's like going to Med school and then treating every disease
with antibiotics.  And unfortunately, they are unaware of the other
possible anlaysis options.

Maybe non-statistics-degree practitioners are too proud to ask
for help (not that there aren't good practitioner who aren't
formally trained)?

So, Gordon, IMHO, you should (if you haven't already) make sure you
teach yourself (or take courses in) General Linear Models, Generalized
Linear Models,
Discret Model and Multivariate Analysis. Then, you'd realize many of
the commonly used methods are just special cases of these (except
Nonparamteric methods).

Hard core probability stuff, while certainly not detrimental,
won't really help you as a practicitioner.  That stuff (I found) 
is useful only if your doing a PhD in Statistics.  I have yet to
dust-off the Radon-Nikodym theorem at work!

Sorry, rant over.


[EMAIL PROTECTED] (Gordon Kenyon) wrote in message 
news:<[EMAIL PROTECTED]>...
> In quantitative methods text-books for the social and behavioral

>>>> Snip <<<<

>   Anyway, integrating the normal distribution function sure is a bear,
> huh?
.
.
=================================================================
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