On Tue, 18 Sep 2001 21:15:25 +0200, Robert Chung [EMAIL PROTECTED]
wrote:
[ snip, some of mine, and his comments]
... My main point was and
still is that the Slate author used RTM in a sloppy way. That's
what I meant by cavalier.
I never read the Slate article
Robert Chung wrote
Snip
Yike, Rich. Are you still sore that Bonds left the Pirates? Go
back and check the entire thread. This thing started because on
July 13, Eugene Gall quoted an article in Slate that invoked
regression to the mean to prove that Bonds wouldn't hit 70.
I did post the original
the
cavalier way that people toss around the phrase, regression to
the mean, as if it were an immutable law that trumped all other
differences in conditions.
You know, I have never seen that. To the best I recollect,
I have never seen people toss around regression to the
mean in a cavalier way
, Eugene Gall quoted an article in Slate that invoked
regression to the mean to prove that Bonds wouldn't hit 70. I
only entered the thread a month later when you said that Bonds
must be on steroids, and pointed out that looking at Bonds' past
history wasn't much of a guide because he was hitting
Candlestick was, sorta
like the short right front porch at Yankee stadium. Matter of
fact, if Hank Aaron had played half his games in Candlestick
while Willie Mays had played half his in Atlanta...
Discussions about regression to the mean (and comments that Bonds
has never before hit 50) might
My main point was not about baseball or Bonds. It was about the
cavalier way that people toss around the phrase, regression to
the mean, as if it were an immutable law that trumped all other
differences in conditions.
--Robert Chung
right ... reg. to the mean is not a cause of anything
the short right front porch at Yankee stadium. Matter of
fact, if Hank Aaron had played half his games in Candlestick
while Willie Mays had played half his in Atlanta...
Discussions about regression to the mean (and comments that Bonds
has never before hit 50) might be more relevant if all other
SO, when bonds hits 73 ... what will people say vis a vis regression to the
mean?
At 11:40 PM 8/27/01 -0400, Stan Brown wrote:
Rich Ulrich [EMAIL PROTECTED] wrote in sci.stat.edu:
This was a topic a month ago. Just to bring things up to date
Rich Ulrich wrote:
On 28 Aug 2001 06:38:49 -0700, [EMAIL PROTECTED] (Dennis Roberts) wrote:
SO, when bonds hits 73 ... what will people say vis a vis regression to the
mean?
... steroids ... ?
(have to guess that for the 56 he already has.)
Hmmm. I would have suggested that Pac
This was a topic a month ago. Just to bring things up to date
Barry Bonds hit 38 homers in the first half of the season (81 games),
a record pace. Should we expect his performance to regress to the
mean sufficiently that he would not break the season record of 70?
BB had never hit 50 in
Rich Ulrich [EMAIL PROTECTED] wrote in sci.stat.edu:
This was a topic a month ago. Just to bring things up to date
Barry Bonds hit 38 homers in the first half of the season (81 games),
a record pace. Should we expect his performance to regress to the
mean sufficiently that he would not
- I am taking a second try at this question from dmr -
On 17 Jul 2001 15:23:29 -0700, [EMAIL PROTECTED] (dennis roberts) wrote:
At 04:08 PM 7/17/01 -0400, Rich Ulrich wrote:
But, so far as I have heard, the league MEANS stay the same.
The SDs are the same. There is no preference, that I
On 17 Jul 2001 15:23:29 -0700, [EMAIL PROTECTED] (dennis roberts) wrote:
At 04:08 PM 7/17/01 -0400, Rich Ulrich wrote:
But, so far as I have heard, the league MEANS stay the same.
The SDs are the same. There is no preference, that I have ever
heard, for records to be set by half-season,
On 16 Jul 2001 09:31:08 -0700, [EMAIL PROTECTED] (dennis roberts) wrote:
[ snip, RTTM is about 'relative' values ... ]
the issue that has to be raised with respect to the baseball example is ...
are the two halves PARALLEL HALVES? ... like, parallel tests given at
essentially the same
At 04:08 PM 7/17/01 -0400, Rich Ulrich wrote:
But, so far as I have heard, the league MEANS stay the same.
The SDs are the same. There is no preference, that I have ever
heard, for records to be set by half-season, early or late, team
or individual. My guess is that association between talent
regression to the mean has NOTHING to do with raw numbers ... it ONLY has
to do with relative location withIN a distribution
example: i give a course final exam the first day ... and get scores (on
100 item test) from 10 to 40 ... and an alternate form of the final on the
last day ... and get
that regression to
the mean doesn't imply a loss in diversity.
=
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=
Paige Miller wrote:
EugeneGall wrote:
This hardly PROVES anything. It is more a statement about what has
happened in the past.
Proves was in the original article. I'm assuming Ellenberg, a mathematics
prof, was using 'proves' in a tongue-in-cheek fashion. However, he was serious
in
EugeneGall wrote:
Jordan Ellenberg, in today's Slate, PROVES that Bonds won't break the
HR record because of regression to the mean. The argument is a
little sloppy, but there is definitely some RTM involved:
If our discussion above is correct, then hitters who
lead the major leagues
On 14 Jul 2001 00:26:03 GMT, [EMAIL PROTECTED] (EugeneGall)
wrote:
[ snip, about Bonds and home runs, and Regression to the Mean ]
I'd be curious if reduction in the 1st half leaders was comparable to the
improvement in the 2nd half leaders.
Huh?
If you are asking what I think you
Jordan Ellenberg, in today's Slate, PROVES that Bonds won't break the
HR record because of regression to the mean. The argument is a
little sloppy, but there is definitely some RTM involved:
If our discussion above is correct, then hitters who
lead the major leagues in home runs at the All
Dennis,
I agree with all your points; I thought the news report was a nice
example related to regression to the mean and of attempting to prop
up a (nearly certainly true) statement using invalid statistics -
note that the report appears to refer to the *current* top bottom
10% and what
In article 94qi92$[EMAIL PROTECTED],
[EMAIL PROTECTED] (Herman Rubin) writes:
This has nothing to do with regression to the mean.
The people in the top 10% and the bottom 10% have changed.
I see "regression to the mean" and "the people in the top 10% and
the bottom 1
... that
is ... some % value is that amount that increments their salaries ...
snip, other stuff based on this misreading
And Dennis quotes the material --
At 05:07 PM 1/25/01 +, wrote:
Avid regression-to-the-mean watchers may be interested to know that,
according to yesterday's summary
Avid regression-to-the-mean watchers may be interested to know that,
according to yesterday's summary of the growing rich-poor divide
(on teletext news), the current top 10% of earners have had
a higher percentage increase in income over the past x years
(for some x that I've forgotten) than
this will be quite high) ... thus, there still will be
SOME regression to the mean that is ... if we isolate the top 50 ... and
the bottom 50 ... and look at their percentile ranks (or mean z scores)
from this year to next ... you will still see that the lower 50 have
relatively higher percentile ranks than
In article 94pmgo$rn7$[EMAIL PROTECTED],
[EMAIL PROTECTED] wrote:
Avid regression-to-the-mean watchers may be interested to know that,
according to yesterday's summary of the growing rich-poor divide
(on teletext news), the current top 10% of earners have had
a higher percentage increase
My response is about regression to the mean generally, which got done
over a little over a week ago.
It occurred to me recently that you could reduce the
regression-to-the-mean effect by using the subjects' least-squares
means to divide them (the subjects) up into quantiles for separate
e two sets of "test" measures ... would qualify for
being a context in which to illustrate RTM?
Not at all. If there's a perfect r you _won't_ see regression to the
mean! What it means
is that not everything which expands or compresses the ends of a
distribution is RTM.
My understa
I've heard this before -- probably read it in stat
books. It isn't true. Galton worried over the
problem until he understood the statistical
mechanism. He even designed a device (the
Quincunx) to demonstrate how it works. Steve
Stigler has a section on it in his new book.
Thus Galton found
31
26 2142
27 2032
28 2024
29 1439
30 1143
if you are thinking about regression to the mean in the typical way ... how
come this "regression reversal" s
On Wed, 17 Jan 2001, Bob Wheeler wrote:
I've heard this before -- probably read it in stat
books. It isn't true. Galton worried over the
problem until he understood the statistical
mechanism.
you may abe right; that's why I said apparrently
J. Williams [EMAIL PROTECTED] wrote:
Would this not
: be the same as the offspring of either the very tall or the very short
: among us moving toward an arithmetic average? Is it inconceivable
: that a pair of dullards could produce a Beethoven or a Fermi for
: example? Frankly, I believe old
here are some of the actual reported galton data ... scatterplots between
fathers' and sons' heights ... interesting tidbit about these data ...
clearly, some fathers sired not only sons ... but also daughters ...
S ... for the case of daughters ... the value that was imputed was
a
you could clarify what is and what is not ... RTM?
Thus it
hides the effect of regression to the mean; however, we may guess that
the size of the improvement is somewhat reduced.
-Robert Dawson
=
Instructions for joining a
Paul R Swank forwarded Dennis' scattterplot:
- *
post - * *
- *
- 2 *
80+ * 2 *
- 2 * * *
- * * *
- * *
- * *
60+
- * *
- * * *
-
- *
40+ *
-
-
+-+-+-+-+-+--pre
10.0 15.0 20.0 25.0 30.0 35.0
Aha! So
divide at the midpoint of the pretest to form two equal size groups.
At 01:37 PM 1/17/01 -0500, you wrote:
>At 12:28 PM 1/17/01 -0600, Paul R Swank wrote:
>>But if you group the subjects on the basis of their pretest scores, the
>>lowest group gains 23.1 points while the highest group only gains
37 matches
Mail list logo
