|
|
Re: MEDIAN and MODE as AVERAGES is NOT to be CONDONED or TOLERATED, EVER.
Posted:
Aug 9, 2006 5:40 PM
|
|
On 4 Aug 2006 13:52:15 -0700, "Reef Fish"
<large_nassua_grouper@yahoo.com> wrote:
>
> Richard Ulrich wrote:
> > On Wed, 02 Aug 2006 22:39:42 -0400, Jerry Dallal
> > <gdallal@SPAM_BLOCK.world.std.com> wrote:
RF > >
> I already shot down Jerry's argument once. Doing it twice would be
> a gross overkill, for the simple statistical definition of a mode.
The rational substance of "shooting down Jerry's argument"
seems to be, quoting from Bob's post to Jerry's --
RF>
> These men are all DEAD, so are their sins of improper usage of the
> term "average". This is the 21st century now, Jerry.
Thus, Bob retreats from defending the historical use
of the mode as an average. Now he focuses on the 21st
century. Jerry, reasonably enough, walked away from the
debate before this, after showing that Bob was thoroughly
wrong in the 20th century context that everyone had assumed.
Apparently, it is the 21st century use that Bob is concerned
with, since the 20th and 19th don't support him.
RU >
> > Jerry,
> > Thanks for the excellent set of posts.
> >
> > "Linguistic prescriptivism",
RF >
> Richard Ulrich immediately dived into his "tangents" (which had
> nothing to do with calculus or trigonometry BTW).
>
My tangent has to do with Bob's "argument".
Anyone arguing at length, for years, about words ought
to have a passing familiarity with linguistics. The same
arguments tend to recur. Bob is entering strange territory
for prescriptivism, when he claims that a usage is new, 21st century,
and yet "right" -- and not that he proposes this for the future,
but that, mystically, already, everyone should agree with him.
I say Bob should have familiarity with linguistics arguments, which
Bob (apparently) doesn't have.
That's kind of parallel to my saying that anyone who wants to
argue about smoking/ cancer ought to have a passing familiarity
with the principles that guide epidemiologists today, and
the data that they rely on; and Bob doesn't.
Or that anyone who wants to discuss social and behavioral
sciences ought to have a passing familiarity with innovators and
influential writers like Jacob Cohen; and Bob doesn't.
In my first Reply concerning the "REVISTED" thread on July 17,
I posted,
======= start quote of myself from July 17.
What's the upside or downside of Replying to this?
The downside is that Bob is that I don't "dialog" with Bob,
because he can't. There is practically no chance that he
will have any relevant reply, and less chance that he will
admit he is wrong about either argument -- whether the mode
is a mean, and whether it matters much what someone asserts.
The upside is that Bob Ling has presented his argument.
I will show how little there is to it.
======= end
I think the threads lived up to that. I did not mention that
Bob would throw in a lot of verbal abuse, though perhaps
I should have. Further, I did not really consider that Bob would
take up a hobbly of malicious prevarication against me in
unrelated threads, but that is not wholly unprecedented, either.
Here are some highlights from several threads.
Bob presented his scorn for dictionaries and other references.
He offered baffling advice about only looking up one term
(mode, not average); omitting further citations; and discarding all
definitions he does not like since they are "wrong" [according
to his intuition or early learning, apparently -- not explicated
further by Bob].
Bob accepted an expanded "intuitive" meaning for an average
by accepting geometric means, etc.; that should cost support
from the people whose intuition only includes the arithmetic mean.
The proposition was an extreme fashion of its statement, with
Bob heaping scorn on anyone who diverges from him,
"The statistical mode is NOT an average".
- Refuting his extreme opinion only requires that some mode, in
some context, be considered an average, by some reasonable people.
There have been plenty of examples of that for noted statisticians,
if the dictionaries did not settle the commonplace view. (U-shaped
beta curve is not relevant in any apparent way.)
If Bob had maintained, "The statistical mode is NOT an average"
"for *statisticians* who want to be precise in the future", he
would not have waged a holy war on the subject across multiple
usenet groups, over a couple of years. That's a different and more
boring question. Eventually, Bob seemed to try to change the
proposition to something like the longer version.
A reader who agreed with Bob's proposition in the first place
might be unmoved by all the words, and still agree at the end.
Or they might take to heart (especially) Jerry's historical evidence,
and feel forced to accept that, and probably accept
Jerry's earlier recommendation that "averages" should always
be treated with suspicion if they aren't explicit.
Anyone (especially) who had doubts about the proposition
should be impressed, I think, with Bob's lack of cogency,
in addition to his lack of civility.
[snip]
--
Rich Ulrich, wpilib@pitt.edu
http://www.pitt.edu/~wpilib/index.html
|
|
