Subject: Population
From: =?iso-8859-1?Q?H=E5kon?= Finne
Date: Mon, 20 Jan 1997 22:51:37 +0100
A population is frequently defined as 'the totality of all possible
observations on measurements or outcomes' (Kmenta). This leads to the
assertions that most populations of interest are infinite, so that even
if we think we have data for all units, we should still apply
significance tests etc. in order to be able to infer from our sample of
all realized observations to all potential observations. Even more
frequently, however, we find that writers define their population in
terms of certain characteristics of the units, such as 'all small firms
in the West Coast electronics industry in the 1980s' (construed
example), as a reflection of their sampling limitations. What is the
relation between these two concepts of a population?
H=E5kon Finne
SINTEF IFIM (Institute of social research in industry)
N-7034 Trondheim
Subject: Re: This is impossible
From: rdadams@access1.digex.net (Dick Adams)
Date: 20 Jan 1997 17:57:20 -0500
Macarthur Drake wrote:
> This messege is to provoke a serious scientific debate.
Then it should have been written more seriously and without
a plethora of spelling and grammatical errors!!
> I am an engineer, no biologist, astronomer or statictician or anything,
> but something puzzles me. I am sure you are aware of the Late Dr.
> Sagan's quote " extraordinary claims require extraordinary proof "
> with regards to extraterrestrial life, UFOs etc. I have also heard
> people say that the discovery of life on another world would be the
> greatest discovery in human history.
> I beg to differ with both of these ridiculus statments.
> [snip]
> I would appriciate any math or stats expert to comment on the
> chances that we are alone in the entire universe. I bet that s/he'd
> say that it is statistically impossible for us to be alone, so what's
> the big deal we know that life is there, just a matter of time 'til
> we find it....or them us!
> [snip - again]
Extraordinary claims DO REQUIRE extraordinary proof. Statistics can
be used to measure the dimensions of reality, but not to verify the
moments of reality.
Let me introduce you to Dick Adams' Three Laws of Statistical
Analysis (from my unpublished paper "The Statistical Demise of
the Moss Klein Pitcher):
First Law: Every non-uniform distribution can be expected to
have a tail.
Second Law: Fairy tales come true in the tails of a distribution.
Third Law: Anyone who proposes an argument based on the tails
of a distribution without rigorous proof of a link
to reality should expect their argument to be viewed
as being without merit and is deserving of the
ridicule received.
When I apply Occam's Razor, Dr. Sagan's presentation is superior
to mine. But then he was being scientific while I was being
sarcastic.
Life elsewhere in the universe is probable; finding it may well
be the greatest discovery ever made up to that time.
Dick -- There are no Statisticians. We are all just students of
Statistics striving to learn more each day.
Subject: Re: This is impossible
From: jejanes@uclink4.berkeley.edu
Date: Mon, 20 Jan 1997 18:04:55 -0700
Macarthur Drake wrote:
>
> This messege is to provoke a serious scientific debate.
>
> I am an engineer, no biologist, astronomer or statictician or
> anything, but something puzzles me. I am sure you are aware of the Late Dr.
> Sagan's quote " extraordinary claims require extraordinary proof " with
> regards to extraterrestrial life, UFOs etc. I have also heard people say
> that the discovery of life on another world would be the greatest discovery
> in human history.
> I beg to differ with both of these ridiculus statments. From my
> understanding of biochemistry and the number of stars in the universe....to
> not find life would be the greatest discovery of all times. There is nothing
snip
> I would appriciate any math or stats expert to comment on the
> chances that we are alone in the entire universe. I bet that s/he'd say that
> it is statistically impossible for us to be alone, so what's the big deal we
> know that life is there, just a matter of time 'til we find it....or them
> us!
> By the way there are an estimated 100 million million million stars
> or as Dr. Sagan put it more stars than the number of grains of sand on all
> the beaches/deserts on the entire Earth!
>
> Logical and insightfully comments welcomed!
>
> drake.79@osu.edu
Experts in math and stats could provide no meaningful incite unless they
were also expert in Biochemistry. More expert, in fact, than anyone
currently is. Statistics does not provide one with a magical elixir to
make meaningful predictions from a point of ignorance. As far as my
biological knowledge goes, this is the bare essentials for life:
liquid water
20 amino acids (and their biosynthetic pathways, unless the primordial
soup is extremely rich)
20 tRNA'a
20 tRNA synthetases
functional ribosomes
An RNA-dependent RNA-polymerase (I skip DNA to make it easier)
RNA which codes for the last 4 requirements
Note that all of these must be present simultaneously. RNA that codes
for ribosomes will do no good unless their is already a ribosome to
implement it. RNA polymerase will do no good unless there is already
RNA (coding for the polymerase) to duplicate, and etc.
To be more generous, let's say we only need 5 amino acids, tRNA's, and
synthetases. Still looks pretty darn unlikely, even for billions of
stars and billions of years.
Of course life doesn't have to start this complicated. It could have
consisted only of catalyic RNA at first. But we have little to no
evidence that RNA can catalyze such a range of transformations.
Until (or unless) we learn more about the beginnings of life on this
planet, we cannot even speculate to within a power of 10, or for that
matter a power of 1000000000, the likelihood of life elsewhere.
Jeff
Subject: Re: This is impossible
From: "Eric Lucas"
Date: 21 Jan 1997 02:50:50 GMT
As I see it, there is one big problem with the whole debate. The most
likely number of stars in the universe that have life is equal to the
product of the number of stars and the probability that any one star had
the right conditions to foster the formation of life. We know the number
of stars, roughly speaking. If the probability of life forming around any
one of these stars is much less than 1/(the number of stars), then the
product is small and it is unlikely that there is other life out there. On
the other hand, if the probability that any one star has the right
conditions to give rise to life is of the order of 1/(the number of stars)
or greater, then there is a great likelihood that some other star has life
circling it.
The problem is that we have *absolutely* no way to rationally assess the
probability that any given star has the right conditions to have given rise
to life or intelligent life. All we have is the two emotional "arguments",
a) "There are *so* many stars, there *must* be other intelligent life out
there."
and
b) "The formation of life is *so* improbable that it *can't* have happened
more than once in the universe."
Unfortunately, both of these arguments have exactly equal validity and
probability until we have some means of assessing how probable (or
improbable) the origin of intelligent life at any spot in the universe
might have been.
Eric Lucas
Macarthur Drake wrote in article
<5bsc70$f1d@csu-b.csuohio.edu>...
> This messege is to provoke a serious scientific debate.
>
> I am an engineer, no biologist, astronomer or statictician or
> anything, but something puzzles me. I am sure you are aware of the Late
Dr.
> Sagan's quote " extraordinary claims require extraordinary proof " with
> regards to extraterrestrial life, UFOs etc. I have also heard people say
> that the discovery of life on another world would be the greatest
discovery
> in human history.
> I beg to differ with both of these ridiculus statments. From my
> understanding of biochemistry and the number of stars in the
universe....to
> not find life would be the greatest discovery of all times. There is
nothing
> extraordinary about looking the biochemistry of life and looking at the
> billions of stars (in this galaxy alone) and concluding that there MUST
be
> life out there. From a scientific view, there is nothing unique about
amino
> acids.....the elements that make them up are not located only on this
planet
> for sure. Now if there were only 10 stars in the entire universe the Dr.
> Sagan's stament maybe more logical. But we can say, based upon all our
> scientific theories, that LIFE MUST exist elsewhare in the universe. If
> not, then everything we understand about the universe is false. I even
> suspect that life is rather common, say every 20,000 stars or so.
> Now I am not saying that UFO are here or anything, just that life
> must exist. Maybe they are just prokaryots or something, but that is
life.
> Why all the hype? We know that alien life MUST exist in some form or
> another, so why get so excited? Now the debate over UFOs and aliens
visiting
> is another story, although I think that can be debated scientifically
also.
> I'd be glad to do that with anyone who likes debating as much as I do.
> I would appriciate any math or stats expert to comment on the
> chances that we are alone in the entire universe. I bet that s/he'd say
that
> it is statistically impossible for us to be alone, so what's the big deal
we
> know that life is there, just a matter of time 'til we find it....or them
> us!
> By the way there are an estimated 100 million million million stars
> or as Dr. Sagan put it more stars than the number of grains of sand on
all
> the beaches/deserts on the entire Earth!
>
> Logical and insightfully comments welcomed!
>
>
>
> drake.79@osu.edu
>
>
>
Subject: Re: STAR OF CANOPUS
From: america2@ix.netcom.com (Brad)
Date: 21 Jan 1997 09:24:03 GMT
In <5c1kqi$cdj@csu-b.csuohio.edu> drake.79@osu.edu (Macarthur Drake)
writes:
>
>In article <5c0t8g$j5g@access1.digex.net>, rdadams@access1.digex.net
says...
>>
>>Macarthur Drake wrote:
>>> This messege is to provoke a serious scientific debate.
>>
>>Then it should have been written more seriously and without
>>a plethora of spelling and grammatical errors!!
>>
>
>
> Also very cute, but I thought I missed
>'alt.correct.my.english.please'...if not then I'll do so next time.
Remember
>it is the thought that counts....words are just a specific pattern of
>varying density of air...and typed words on computers are just a bunch
on
>transmitted electrons...so lighten up buddy....
>
I'm interested in knowing about the star of CANOPUS? What the name
means, it's history, how far away it is, it's size, it's spectral
class, can NASA ever send a spaceship there, what constellation it's
in. Any information would be greatly appreciated.
COHEN
america2@ix.netcom.com
Subject: Re: This is impossible
From: rwhite@rideau.carleton.ca (Robert White)
Date: 21 Jan 1997 07:28:52 GMT
The probability of life elsewhere other than planet earth is exactly
1.0! This is a true statement and I have evidence! So does NASA in so
far as they keep blasting it into space. Suffice it to say that there
is life in space at this very moment. Moreover, what goes up must come
down!
It's all relative, but there still ain't no absolutes. If in theory
one wanted to assert that 'life must exist on another planet' one
would also have to assert that life did not exist on another planet
if one was to make these assertions within frames of
empiricism/science. To assert that 'alien life' existed one would need
empirical evidence to back the assertion and no one has
any. Theoretically, if one says that 'alien life' does indeed exist,
but they have no empirical proof to back the assertion, it is not
theory but more aptly dogmatic posturing.
--
----------------------------------------- Carleton University ----------
Robert G. White Dept. of Psychology
Ottawa, Ontario. CANADA
INTERNET ADDRESS ----- rwhite@ccs.carleton.ca ------------------- E-MAIL
------------------------------------------------------------------------
Subject: Re: Statistical significance of covariance values
From: Greg Hancock
Date: Tue, 21 Jan 1997 07:33:15 -0500
Karl Merkley wrote:
> I have a covariance matrix that should be banded
> (values around the diagonal, zeros elsewhere) based
> on the nature of the problem. I end up with small
> values far from the diagonal. I would like to when
> I can safely ignore these covariance terms. Is there
> a test to indicate the statistical significance of these
> terms. I have not found one in the multivariate texts and
> the multivariate prof here at BYU didn't know of one.
> Anybody out there have any ideas?
I can think of a couple of options, both of which involve structural
equation modeling (SEM). To get at individual covariances, you could
impose a path model on your variables in which all variables are allowed
to covary freely. Using EQS, LISREL, AMOS, or any of a number of SEM
software packages, you will get standard errors associated with each
covariance thereby allowing you to conduct tests of each one.
If you wish to test how well a specific model fits your data (i.e., a
model in which all covariances away from the band around the diagonal are
zero), impose a path model upon your data in which only those variables in
the band are allowed to covary. The overall chi-square value for this
model, with it's appropriate degrees of freedom, will tell you if this
model should be rejected as a whole on statistical grounds.
I hope this is a useful direction for you.
Good luck,
Greg Hancock
Subject: Re: This is impossible
From: Frank Harrison
Date: 21 Jan 1997 14:34:44 GMT
"\"Uncle Al\" Schwartz" <#UncleAl0@ix.netcom.com> wrote:
> The basic chemistry of star tar is default carbon-based life. Start
> with a large plop which, for whatever reason, contains liquid water and
> you are in business. Look at what infests Antarctic sandstone, deep sea
> ridges, and wet basalt two miles underground. The Miller and
> Ponamperuma experiments generate abiotic fatty acids, sugars, amino
> acids, nucleic acid bases... and ordered coacervate droplets within a
> week. Give it a billion years to organize in a diversity of physical
> and chemical environments.
>
> The problem is that billion years. The Earth is unique in that 2/3 of
> its planetary crust is orbiting 240,000 miles overhead - the Moon.
>
Rare perhaps, but almost certainly not unique. Presumably the event
(collision?) that put so much of earth's crust into orbit could happen
elsewhere, and at any rate there could be other possible scenarios that
might bring all the right materials together under the conditions needed
for life to evolve and remain undisturbed enough to where civilization
could develope.
Frank
Subject: Re: This is impossible
From: Erik Max Francis
Date: Tue, 21 Jan 1997 07:51:14 -0800
Macarthur Drake wrote:
> I am an engineer, no biologist, astronomer or statictician or
> anything, but something puzzles me. I am sure you are aware of the Late
> Dr.
> Sagan's quote " extraordinary claims require extraordinary proof " with
> regards to extraterrestrial life, UFOs etc.
Specifically, he was talking about UFOs being aliens spacecraft. It's not
the claim that extraterrestrial life exists that Sagan was calling an
extraordinary claim, it's that the aliens were regularly visiting Earth
surrpetitiously (but very poorly since they always seem to be seen).
> But we can say, based upon all our
> scientific theories, that LIFE MUST exist elsewhare in the universe. If
> not, then everything we understand about the universe is false.
Yes. The question is not whether or life is possible or not, but how
common it is -- and, furthermore, how common _intelligent_ life is. For
that science has no answers.
Besides, don't knock confirming theories, even the ones that we're sure
about. That's how science works.
--
Erik Max Francis, &tSftDotIotE; / email: max@alcyone.com
Alcyone Systems / web: http://www.alcyone.com/max/
San Jose, California, United States / icbm: 37 20 07 N 121 53 38 W
\
"Gods are born and die, / but the atom endures."
/ (Alexander Chase)
Subject: SPlus 3.3 Windows: changing graph line widths fails
From: Tom Hilinski
Date: Tue, 21 Jan 1997 09:07:59 +0700
Using the default graphics library for SPlus 3.3 for Windows, I am
trying to change the width of non-solid lines using the "lwd" parameter
in plot functions.
e.g.
lines(x, y, lty=2, lwd=3)
If the line type (lty) = 1 (solid) then the width will change both on
the screen and on HP and PostScript printers. For any other line type,
the width will not change. This results in dashed lines being much too
thin for publication.
I've sent this question to StatSci support email address over two weeks
ago and have had no response.
Any suggestions please to get the line widths thicker?
--
Tom Hilinski
Dept. Soil & Crop Sciences, Colorado State University, Fort Collins, CO
80523
hilinski@argid.agsci.colostate.edu
Subject: Re: This is impossible
From: rdrost@globaldialog.com (RF Drost)
Date: Tue, 21 Jan 1997 14:52:05 -0500
In article <32E41634.2178@uclink4.berkeley.edu>, jejanes@uclink4.berkeley.edu wrote:
> Experts in math and stats could provide no meaningful incite unless they
> were also expert in Biochemistry. More expert, in fact, than anyone
> currently is. Statistics does not provide one with a magical elixir to
> make meaningful predictions from a point of ignorance. As far as my
> biological knowledge goes, this is the bare essentials for life:
>
> liquid water
> 20 amino acids (and their biosynthetic pathways, unless the primordial
> soup is extremely rich)
>
> 20 tRNA'a
> 20 tRNA synthetases
> functional ribosomes
> An RNA-dependent RNA-polymerase (I skip DNA to make it easier)
> RNA which codes for the last 4 requirements
>
Add a pinch of salt, bake at 350 degrees for 2 hours and voila
... Pauley Shore.
Subject: Re: Probability and Wheels: Connections and Closing the Gap
From: nveilleu@NRCan.gc.ca (Normand Veilleux)
Date: Tue, 21 Jan 1997 18:09:45 GMT
>From: bm373592@muenchen.org (Uenal Mutlu)
>
>We have to seperate the things:
>
> - probability of 54 fixed tickets in 1 drawing
> - probability of 1 fixed ticket in 54 drawings
> - probability of 54 fixed tickets in 54 drawings
>
> - probability of 54 random tickets in 1 drawing
> - probability of 1 random ticket in 54 drawings
> - probability of 54 random tickets in 54 drawings
>
>(there are no duplicate tickets in each case)
>
>Can someone calculate some or all of them?
We have already shown you that 1 ticket in 54 draws has about 63%
chance of winning. It's totally irrelevant how it's selected.
By the same token, when you say 'random' above you mean that the
set of tickets was SELECTED randomly. Because of this, you can
expect a variation in the percent of coverage. Like I stated before,
in the case of 27 tickets in 1 draw, you can expect the percentage to
vary from 4.7% to about 45%. So for 54 tickets it would be in the
range of 15% to 75%. So the probability for the 54 drawings will
also be a range.
What you have to understand is that the EXACT probability of a set of
tickets can only be calculated AFTER it is selected. And once it IS
selected, then it's a FIXED set!! A random set of tickets cannot have
an EXACT probability, only a range.
Subject: Re: New User
From: campbell@acs.ryerson.ca (Kent Campbell)
Date: 21 Jan 1997 18:02:56 GMT
Joseph McClernon (fjoseph1@saluki-mail.siu.edu) wrote:
: M. Stojanovic wrote:
: >
: > Thank you for your fast answer. I like to ask somebody what will be the
: > best software for data entry,cleanning the errors and coding, for not to
: > large survey. Right know I have SPSS 7.5 . I can do the data entry with
: > this software, but what I should use for clleaning the errors and for
: > the rest.
: >
: > Best regards,
: > Mirjana stojanovic
: Try using Microsoft Access. There is a bit of a learning curve but I
: found over the years that there is nothing better for data entry,
: cleaning, management etc. Access is especially useful if you datasets
: are large and/or complex. Another advantage is that it can be used to
: import data from a large variety of sources and can be used with SPSS
: with ODBC.
: Hope this helps.
: Joe McClernon
Hi Joe & Mirjana,
there is an SPSS newsgroup on Usenet. (I don't know its
name but try searching using "SPSS". If you can't find it, e-mail me and
I will look it up for you.) Some SPSS employees check that newsgroup
fairly often and answer questions from users.
Best wishes,
Kent.
Subject: Statistical Applications Programmer Job -- Houston
From: bwb@odin.mdacc.tmc.edu (Barry W. Brown)
Date: 21 Jan 1997 22:05:52 GMT
COMPUTER PROGRAMMER II: Section of Computer Science,
Department of Biomathematics,
The University of Texas M. D. Anderson Cancer Center
Houston, TX
OVERVIEW: Our group develops computer code to provide statistical
capabilities needed in cancer research but not available commercially.
The code so developed is posted for general use. We also test
competing methods for performing common statistical functions and
report on the comparisons. Examples of comparisons include methods to
correct for multiple testing and methods for confidence intervals in
binomial proportions.
REQUIREMENTS: College degree (M.S. preferred) in a science and at
least one year of relevant experience. A good understanding of
statistical philosophy and methods is necessary to understand tasks to
be performed. We strive for quality, hence knowledge and application
of good coding practices are required. A high level of oral and
written communications ability is necessary to refine computer program
specifications and to internally and externally document code. A
general knowledge of numeric methods is a definite plus.
JOB DESCRIPTION: The programmer will receive general, sometimes vague,
specifications from a faculty member. The programmer considers and
refines these specifications in collaboration with the faculty member.
The programmer designs code through successive refinement, a process
that may well involve further discussions. The programmer writes code
and drafts documentation. As all software developers know, this
process may well be iterated.
ENVIRONMENT: The primary compute servers are Unix machines with the
usual utilities (e.g., EMACS, LaTeX). Our standard procedural
language is Fortran 77; we are considering switching to Fortran 90.
We also write a good deal of code in Splus. We have most of the
standard scientific software (Maple, Mathematica, SAS, etc.) as well
as language translators: Fortran to C, f77 to f90.
NOTE: The successful applicant need not know Fortran or Splus, but
will be expected to rapidly learn them. Our experience has been that
people with excellent skills in one computer language easily transfer
them to other languages.
MORE INFORMATION: on the Section, Department, and Institution can be
obtained on our web page. Included are code and documentation
available for download. The URL is
http://odin.mdacc.tmc.edu/
SALARY RANGE: For a holder of a B.S. and one year's experience, the
salary is somewhat less than $30K. With a M.S. and several years
experience the salary could be $40K or slightly more.
IF INTERESTED: contact Barry W. Brown.
Barry W. Brown
Department of Biomathematics,
Box 237
University of Texas M. D.
Anderson Cancer Center
1515 Holcombe Blvd
Houston, TX 77030
NEW EMAIL ADDRESS
bwb@odin.mdacc.tmc.edu
internet address is (143.111.62.32)
Subject: Re: Performance Index for Covering Designs, Wheels
From: nveilleu@NRCan.gc.ca (Normand Veilleux)
Date: Tue, 21 Jan 1997 17:57:28 GMT
>From: bm373592@muenchen.org (Uenal Mutlu)
>
>EXAMPLE DCPI CALCULATIONS:
>
>Input: | Processing: | Output:
> | |
>vAll vSub k b minWin NCC | C(v,k) pWin p1 | DCPI
>---------------------------------|------------------------------|-----------
> 49 6 6 27 3+ 6213118 |13983816 0.444308 0.018637545 | 88.2941 %
> 49 6 6 43 3+ 8510946 |13983816 0.608628 0.018637545 | 75.9442 %
> 49 6 6 54 3+ 9718896 |13983816 0.695010 0.018637545 | 69.0571 %
> 49 6 6 168 3+ 13983816 |13983816 1.0 0.018637545 | 31.9376 %
>
>Interpretation:
> ie. the above 27-ticket wheel covers only 88.2941% of the
> theoretically possible maximal case (ie. 100% :-), so there
> is theoretically still room for some improvements (in this case,
> either pWin upto 50.32137%, or equally DCPI upto 100%).
> And: this 27-ticket wheel performs better than for example the
> given 43-ticket wheel
Sorry Uenal, the DCPI calculation is directly related to the WCD
and hence is not better than simple % to indicate which cover is
best. WCD increases with the number of lines whereas DCPI decreases.
As long as the number of lines in a cover remains constant, then WCD
can be used to compare them. But it's all useless in a sense since
what we are comparing is a conditional probability, and not matter
what the percentage of cover is, all sets with the same number of
lines have the exact same average expected gain. Hence, on average
you will win (or lose) about the same amount of money no matter what
is the percentage of coverage.
Subject: Is it possible???
From: val
Date: Tue, 21 Jan 1997 14:07:50 -0800
Is it possible???
Base - theory of Teilhard de Shardin.
Creation of Hyper brain:
Increasing speed of net ~ 10000 times.
Direct contact from net to brain , virtual reality(?).
Self-organization of initial seed - new evolution structure.
Finally, new structure restricted only geometrically, by surface
of planet and by number of involved humans (sells of Hyper brain).
Power of Hyper Brain will excel power of human brain,
as human brain excel power of brain cell.
Power of Hyper Brain is INFINITE.
Evolution approaches to omega point - crown of evolution on
earth and , finally, in universe.
Earth is unique place in universe - if this
process was going somewhere else ,Omega would involve us
already.
================================================
WE ALONE IN UNIVERSE, BECAUSE WE ARE FIRST !??
================================================
It puzzle me long time, please, any comment.
Thank you.
Val.
Subject: Re: This is impossible
From: Ramsa
Date: Tue, 21 Jan 1997 20:01:49 -0500
RF Drost wrote:
>
> In article <32E41634.2178@uclink4.berkeley.edu>, jejanes@uclink4.berkeley.edu wrote:
>
> > Experts in math and stats could provide no meaningful incite unless they
> > were also expert in Biochemistry. More expert, in fact, than anyone
> > currently is. Statistics does not provide one with a magical elixir to
> > make meaningful predictions from a point of ignorance. As far as my
> > biological knowledge goes, this is the bare essentials for life:
> >
> > liquid water
> > 20 amino acids (and their biosynthetic pathways, unless the primordial
> > soup is extremely rich)
> >
> > 20 tRNA'a
> > 20 tRNA synthetases
> > functional ribosomes
> > An RNA-dependent RNA-polymerase (I skip DNA to make it easier)
> > RNA which codes for the last 4 requirements
> >
>
> Add a pinch of salt, bake at 350 degrees for 2 hours and voila
> ... Pauley Shore.
Damn! And there I was, thinking "sugar and spice and everything nice"
and "frogs and snails and puppy dog tails"!
Ramsa
