Re: Statistics in Psychology?


Herman Rubin wrote:
In article <lv2dnSXdt9zqogrZnZ2dnUVZ_v6dnZ2d@xxxxxxxxxxx>,
Jerry Dallal <gdallal@xxxxxxxxxxxxxxxxxxxxxxxx> wrote:
Brett Magill wrote:
On Mon, 19 Jun 2006 01:42:48 -0700, Reef Fish wrote:

.......................

There are more data--important data--than fully trained statisticians to
analyze them.

The problem in a nutshell is that the statistics profession as a whole
has not done a good job of coming up with a way to train people
analogous to the training of a PA or NP. Some masters biostat programs
do a good job, but many in the profession take the position that nothing
less than a complete statistical education can produce someone capable
of serious analysis. Maybe they are right. While there are
limits--there's little one can do in a single semester to turn someone
into a serious analyst--I think there are things that could be done to
train a statistical NP with perhaps four semester courses of hard work
spread out over two years. The real trick--as many of the posts in
sci.stat.* demonstrate--is teaching people to recognize when they are
out of their depths, the equivalent of an NP knowing when to refer to an MD.

My answer is that the most important part is missed; we
cannot TRAIN people to carry out the analyses, especially
the ones which should be done and are not being done.

The entire miseducational systems trains instead of
educating.

Herman and I agree on our "miseducational system", but we
also disagree on the HOW and WHY. The quibble about the
word "training" and "educating" seems to me at best rhetorical.

I have heard Herman's old refrain and his commandments
before, and I am sure he had heard some of mine.

But it's worth repeating -- just for the COUNTERPOINT that
Herman is speaking (and indicting himself) as a MATHEMATICAL
statistician, in the same manner he seriously ERRED in arguing
that the Measure Theory (of the Halmos and Loeve level) is
essential to applied statistics where I consider it completely
unnecessary (having taken several measure theory courses)
for a top-notch applied statistician and data analyst!

THAT is the crux of our essential disagreement. In that respect,
I consider the departments in which Herman taught (heavily
mathematical statistics oriented) to be a MISEDUCATIONAL
SYSTEM for an applied statistician.


For example, we emphasize how to solve
problems in algebra, and also in calculus, rather than how
to formulate the real problems, which may or may not be
solvable with the devices available in the elementary
courses. As I have stated many times, it does no good to
know how to add if one does not know when, and if one
knows when and how to get the computer to carry out the
how, fine.

We agree, this much.


To use statistics properly, the user needs to FORMULATE
the probability model. This requires a more basic
understanding of probability than merely the limit of
relative frequency, but something inherent. So at
whatever level, probability is needed. The probability
model, and whatever assumptions need to be made, are not
a matter of statistics, but of the field, and here we have
a lot of garbage going on, such as transformations to
normality (a permanent non-no), and even normalizations
to get correlations instead of covariances; they mess up
the conclusions which can be drawn. If statistics is going
to be used for decision making for action on individuals,
I see no way of avoiding a Bayesian approach, although
robustness (definable by the user, but unlikely to be
evaluated except by a mathematical statistician) can be
used to reduce the prior assumptions.

Here, Herman is tripping all over place on his Mathematical
Statistician's slip! He is confirming his ANTI-DATA-ANALYSIS
and anti-applied statistics position.

To say that transformation to normality is garbage says it all.


I am often requested to repost my five commandments.

Deja vu all over again. More relevant to the role of a client vs a
consultant than what APPLIED STATISTICS is all about.


The consultant is obligated to point out how their assumptions affect
their views of their domain;

The JASA article I have cited at least a dozen times, by George Box, on
Science and Statistics articulates the point MUCH better than Herman.

That the applied statistician (analyst) is both the SPONSOR and the
CRITIC of his tentative models. THAT is why if the normality
assumption is violated in a model that requires it, transformations
are NECESSARY. That is why I said about Herman's heresy,

RF> To say that transformation to normality is garbage says it all.

Box is also very eloquent about Mathematistry -- something Herman
and his students would be good at. (Read about it from the archives
or from the "Science and Statistics" paper).

Data Analysts and competent Applied Statisticians know ALL about
the caveats of what's real and what's approximate and what the
real problems are and what the practical solutions are.

All of those concepts are within the TRAINING and EDUCATION
in the courses in Data Analysis I've taught over a period of 30 years.


There are, unfortunately, many fields in which much of the activity
consists of using statistical procedures without regard for any assumptions.

Much worse than that. But that's the strawman that needs no
introduction, especially in the sci.stat.* groups. You have to be
BLIND
not to see all those strawmen running around everyday.

But what we are discussing, or supposed to be discussing, is HOW
we may improve the education of statisticians in general, and those
in special applied areas.

In that regard, I strongly disagree with your stance against Data
Analysts and what I take to be competent Applied Statisticians.

-- Bob.

.

Relevant Pages

  • Re: Medical Research-Evidence
    ... Dale is correct in that radical prostectomy has been shown ... "Radical Prostatectomy versus Watchful Waiting in Early ... What Herman was trying to tell you is that this is not so. ... his comments on statistics are valid. ...
    (talk.politics.medicine)
  • Re: Experienced Statistician to help decide whether a regression is legitim
    ... offered to me, I would probably pass them to Herman Rubin, ... of Herman and Jack as "mathematical statisticians". ... Statistics has its role in the world, ... In mathematical statistics, m00es might be ...
    (sci.stat.math)
  • Re: Highest Posterior Density
    ... illywhacker wrote: ... NOT have any basic knowledge about Bayesian methods. ... ignorance about Bayesian statistics and Bayesian inference. ... with 6 posts each by Herman and myself, ...
    (sci.stat.math)
  • Re: Preaching or Teaching?
    ... for learning statistics be included in high school. ... Herman hasn't changed his views in the nearly ten years I know of him. ... don't recall that he mentioned anything with any frequency regarding his ...
    (soc.retirement)
  • Re: Statistics in Psychology?
    ... While there are limits--there's little one can do in a single semester to turn someone into a serious analyst--I think there are things that could be done to train a statistical NP with perhaps four semester courses of hard work spread out over two years. ... To use statistics properly, ... the probability model. ...
    (sci.stat.edu)