From: Automatic digest processor
STAT-L Digest - 13 Apr 1998
To: Recipients of STAT-L digests
Status: RO
There are 7 messages totalling 308 lines in this issue.
Topics of the day:
1. Test for Difference among Several Proportions (2)
2. Certification Programs for Forecasting
3. multi-curve fitting
4. repeated measures regression?
5. spatial correlation and sampling (2)
----------------------------------------------------------------------
Thu, 9 Apr :18 -0400
david rothman
Subject: Re: Test for Difference among Several Proportions
seems like a natural fit for a bootstrap ..dave
Raymond Tsang wrote in message ...
>May I ask what statistical test should I use in testing the difference
>among several proportions? The background for the question is that we
>have surveyed 4 different types of old age homes on several parameters.
>Because the size of these ole age homes vary a lot, the frequency of,
>e.g., residents with certain disease, would also vary a lot. We think
>that we should count the percentage of disease per number of beds in
>order to have standardization before the data are subject to test of
>difference. So we need some advice on the statistical test to be used.
>Thanks in advance.
------------------------------
Mon, 13 Apr :44 -0600
Subject: Certification Programs for Forecasting
I am a forecast analyst for a consulting company.
In an attempt to boost the
credibility of the department, the vice president is encouraging us to become
certified by a professional organization.
For example, he recommended APICS
certification.
However, a number of us believe that while APICS
certification may be a worthwhile goal for some, the material covered is only
tangentially related to our function.
Certification by the American
Statistical Association wo however, the ASA does not
certify people.
Is anyone aware of any such programs administered by a professional society
or university?
Please respond by e-mail to : .
Thank you.
-----== Posted via Deja News, The Leader in Internet Discussion ==-----
Now offering spam-free web-based newsreading
------------------------------
Thu, 9 Apr :18 GMT
Richard F Ulrich
Subject: Re: multi-curve fitting
This note is cross-posted to Newsgroups: sci.stat.consult,
sci.stat.edu,sci.stat.math, where the original message was
separate posted.
Because the original was sent in separate
posts, Responses could appear in each of those other groups.
1) Please cross-post if you can, rather than send out separate
2) One topic you might read about is SUR, Seemingly Unrelated
Regressions.
I have found the topic in texts on Econometrics.
Greene wrote a good one.
- separate criteria, same predictors.
3) "Linear (in the criterion)"
"Least-squares"
go together.
Non-linear equations usually imply that the
maximum likelihood solution is going to be appropriate, where
LS is not.
4) Once you have defined the problem, and have some keywords,
then maybe you can find software in the usual, on-line sites.
Rich Ulrich, biostatistician
wpilib+@pitt.edu
http://www.pitt.edu/~wpilib/index.html
Univ. of Pittsburgh
=====================> note
Lestaw K. Bieniasz (lestaw.k.bieniasz@uni-tuebingen.de) wrote:
: I am new to this list, so please excuse me if this message duplicates
: possible similar earlier messages. I have a difficulty with finding
: adequate literature related to the nonlinear regression problem I describe
: below, and I would appreciate very much your advice, as well as your help
: in finding a public domain source-code software (FORTRAN, C, C++, Pascal)
: that I could use or adapt for solving my problem. Here is the description:
: 1) I want to perform a SIMULTANEOUS least-square-type fitting of N
: DIFFERENT "theoretical" curves, to corresponding sets of "experimental"
: points. The curves are given by equations (all variables are real):
: y1=f1(x1,P)
: y2=f2(x2,P)
<< snip, about different
: Please send answers directly to my current address
: lestaw.k.bieniasz@uni-tuebingen.de
-- Is it too much trouble to login to look for an answer?
Even if you have a poor quality of local service,
there is Deja News....
------------------------------
Mon, 13 Apr :28 GMT
Richard F Ulrich
Subject: Re: repeated measures regression?
Bruce L. Lambert (lambertb@uic.edu) wrote:
: Hi everyone,
: Is there such thing as 'repeated measures regression'?
: I have data from a fairly standard short term memory experiment in humans.
: We're looking at the effect of similarity on short term memory for
: medication names. We have 15 subjects in the experiment. Each subject does
: a total of 16 trials---4 trials at 4 different similarity levels. The
: dependent variable is mean percent correct at each similarity level. We've
: analyzed this with a standard repeated measures ANOVA, and we are somewhat
: content with those results, but the four levels of similarity are not
: equally spaced. For this and other reasons, we'd like to look at this data
: within a regression framework (i.e., regress error rate on similarity), but
: I'm not sure this is kosher to do when each subject contributes multiple
: observations.
: How might one proceed in such a situation?
- In BMDP-2V, which is a procedure for repeated measures,
the POINT(i) specification provides for unequal spacing
for factor (i).
I have not used it, but the documentation seems to
describe what you want.
Rich Ulrich, biostatistician
wpilib+@pitt.edu
http://www.pitt.edu/~wpilib/index.html
Univ. of Pittsburgh
------------------------------
Thu, 9 Apr :14 GMT
Richard F Ulrich
Subject: Re: spatial correlation and sampling
Ed Callahan () wrote:
: I need to select a random sample of ponds that are scattered throughout a
: large landscape.
I expect there to be spatial correlations in the dependent
: variables that will be measured in that sites physically close to each other
: will be more similar.
: Are there better alternatives than simple random sampling in the presence of
: spatial correlation?
Stratifying is one improvement that comes to mind.
- This sounds to me like a question for a specialist, and you have
to be more specific about the purpose, before you can do a PROPER job
of randomizing or stratifying.
For instance, to what extent is a "pond" analogous to a person?
SOME possibilities:
Should each pond have an equal chance to be chosen?
- number them, chose random numbers.
Or, should each pond have a probability proportionate to its area?
- throw darts at a map, take the ones that are hit.
Or, should each pond have a probability proportionate to its isolation?
- throw darts at a map, take the pond whose center is nearest to each.
Then, if you are analyzing ecology, it will matter whether you
look at materials that may be air-borne or carried by water -
constitutes "equal distances" depends on what you are measuring,
and how it might migrate, actively or passively.
To repeat: this sounds like something where you can't get a single,
generic answer.
Rich Ulrich, biostatistician
wpilib+@pitt.edu
http://www.pitt.edu/~wpilib/index.html
Univ. of Pittsburgh
------------------------------
Mon, 13 Apr :47 GMT
Undetermined origin c/o Postmaster
Subject: Re: Test for Difference among Several Proportions
In , wpilib+@pitt.edu (Richard F Ulrich)
>: Raymond Tsang wrote in message ...
>: >May I ask what statistical test should I use in testing the difference
>: >among several proportions? The background for the question is that we
>: >have surveyed 4 different types of old age homes on several parameters.
>: >Because the size of these ole age homes vary a lot, the frequency of,
>: >e.g., residents with certain disease, would also vary a lot. We think
>: >that we should count the percentage of disease per number of beds in
>: >order to have standardization before the data are subject to test of
>: >difference. So we need some advice on the statistical test to be used.
>: >Thanks in advance.
> - Raymond, it looks to me like reporting each disease as a
>PERCENTAGE at a facility is the natural starting point.
>percents are not near 0 or 100, you could do a simple ANOVA among
>the four types of facilities.
If the percentages are near 0 or 100 then you could fit a Poisson
regression model with an offset term representing the size of the old
For percentages near 0 the response would be the number
of residents WITH the disease.
For percentages near 100, the response
would have to be the number of residents WITHOUT the disease.
the Poisson distribution does model the number of occurrences of the
response, inclusion of the offset term in essence standardizes the
count so that you can compare places of different size.
In the SAS
procedure GENMOD, the appropriate offset term would be log(size).
Dale McLerran
dmclerra@fhcrc.org
Fred Hutchinson Cancer Research Center
(206) 667-2926
1124 Columbia Street
fax: (206) 667-5977
Seattle, WA 98104
------------------------------
Mon, 13 Apr :55 -0500
Ed Callahan
Subject: Re: spatial correlation and sampling
I should have been more specific about what I am measuring.
Each pond will
be categorized and the percent of ponds falling into each category will be
estimated and CIs formed.
After some more thought I've come to think that spatial correlation is a red
herring here.
A simplistic model of the problem is this:
I could line up
10 coins in a row, the first 5 heads up and the last five heads down.
is strong spatial correlation in this data set.
However, if I take a simple
random sample the ordering of the objects will have no effect on the
estimate of the percent heads in the population, or on the estimate's
So I think I don't need to worry about spatial correlation and I'll take a
simple random sample.
Please, let me know if it sounds like I'm missing the boat to anyone.
-------------------------------------------------------------------
Edward Callahan
Environmental Statistics
PO Box 563
Fountain City, WI
(608) 687-3205
(608) 687-3409 (fax)
-------------------------------------------------------------------
-----Original Message-----
From: Statistics and statistical discussion list: STAT-L
[mailto:STAT-L@VM1.MCGILL.CA]On Behalf Of Richard F Ulrich
Sent: Thursday, April 09,
To: Multiple recipients of list STAT-L
Subject: Re: spatial correlation and sampling
Ed Callahan () wrote:
: I need to select a random sample of ponds that are scattered throughout a
: large landscape.
I expect there to be spatial correlations in the
: variables that will be measured in that sites physically close to each
: will be more similar.
: Are there better alternatives than simple random sampling in the presence
: spatial correlation?
Stratifying is one improvement that comes to mind.
- This sounds to me like a question for a specialist, and you have
to be more specific about the purpose, before you can do a PROPER job
of randomizing or stratifying.
For instance, to what extent is a "pond" analogous to a person?
SOME possibilities:
Should each pond have an equal chance to be chosen?
- number them, chose random numbers.
Or, should each pond have a probability proportionate to its area?
- throw darts at a map, take the ones that are hit.
Or, should each pond have a probability proportionate to its isolation?
- throw darts at a map, take the pond whose center is nearest to each.
Then, if you are analyzing ecology, it will matter whether you
look at materials that may be air-borne or carried by water -
constitutes "equal distances" depends on what you are measuring,
and how it might migrate, actively or passively.
To repeat: this sounds like something where you can't get a single,
generic answer.
Rich Ulrich, biostatistician
wpilib+@pitt.edu
http://www.pitt.edu/~wpilib/index.html
Univ. of Pittsburgh
------------------------------
Status: RO
There are 6 messages totalling 2027 lines in this issue.
Topics in this special issue:
1. Simulated Annealing
2. AutoBox Financial Forecasting
3. Multivariate Analysis Tool (New features)
4. multi-curve fitting
5. estimating a nonlinear regression function
6. SPSS Consultants
----------------------------------------------------------------------
Sun, 12 Apr :26 GMT
Lester Ingber
Subject: Re: Simulated Annealing
Here is an update of the reference below to ASA:
/asa_retrieve
/asa_retrieve
========================================================================
Adaptive Simulated Annealing ASA-16.17
________________________________________________________________________
ASA is one of the most powerful optimization algorithms for nonlinear
and stochastic systems.
The latest code can be retrieved using these
instructions.
Interactively Via WWW
The ASA archive can be accessed via WWW path
http://www.alumni.caltech.edu/~ingber/ [mirror homepage]
Interactively Via Anonymous FTP
Code and reprints can be retrieved via anonymous FTP from
Interactively [brackets signify machine prompts]:
[your_machine%] ftp
[Name (...):] anonymous
[Password:] your_e-mail_address
[ftp>] binary
[ftp>] get file_of_interest
[ftp>] quit
The 00index file contains an index of the other files.
Files have the
same WWW and FTP paths under the main / e.g.,
/MISC.DIR/00index_misc and
/MISC.DIR/00index_misc reference the same file.
Electronic Mail
If you do not have WWW or FTP access, get the Guide to Offline Internet
Access, returned by sending an e-mail to mail-server@rtfm.mit.edu with
only the words
send usenet/news.answers/internet-services/access-via-email
in the body of the message.
The guide gives information on using
e-mail to access just about all InterNet information and documents.
To get on or off the ASA_list send your request to either
asa-request@alumni.caltech.edu
Update notices are sent to the ASA_list about every month or two, more
frequently if warranted, e.g., in cases of these
notices are the only e-mail sent to the ASA_list.
Additional Information
Lester Ingber Research (LIR) develops projects in areas of expertise
documented in
InterNet archive.
Limited help assisting people with queries on my codes and papers is
available only by electronic mail correspondence.
Sorry, I cannot mail
out hardcopies of code or papers.
========================================================================
In article ,
IQTech-P.Reilly
:In article , umbonner@cc.umanitoba.ca
:> Can anyone tell me where to find an "idiot's guide" to simulated
:> annealing that details
:> the general approach and the Metropolis algorithm?
I've looked at the
:> travelling salesman application in "Numerical Recipes in C" and I've
:> seen an application where a C++ implementation of simulated annealing
:> was used to determine the "optimal" way to cut a sheet of glass, given
:> certain size requirements for the "sub-pieces".
The former went
:> screaming over my head like a 747 and the latter was too specific to the
:> glass problem to provide me with the general understanding that I
:> desire.
:> I have m.(Hons) with a major in OR (completed in antiquity) and I
:> am currently working on a Computer Science degree.
I am fluent in C and
If you have come across anything that would enable a humble
:> undergrad. to understand simulated annealing, I would genuinely
:> appreciate you passing it on.
:> Thank you and regards,
:> Murray Bonner
:Found in my archives and may no longer be up to date...
:Patrick L. Reilly, Gen. Mgr., IsoQuantic Technologies, LLC
1343 N. Alma School Rd.,Ste. 125, Chandler, AZ 85224
:----------------------------------------------------------
Communications Network Architecture, Design & Analysis
GSM Cellular, Mobile Satellite Systems, IMT-2000
:Lester Ingber: Adaptive Simulated Annealing (ASA) version 3.201 Oct
:========================================================================
Adaptive Simulated Annealing (ASA) Version 3.20
:________________________________________________________________________
General Information
:The latest Adaptive Simulated Annealing (ASA) code and some related
:(p)reprints can be retrieved via anonymous ftp from
:ftp.alumni.caltech.edu [131.215.139.234] in the /pub/ingber directory.
:This archive also can be accessed via WWW path
:http://alumni.caltech.edu/~ingber/ or
:ftp://ftp.alumni.caltech.edu/pub/ingber/
:Interactively [brackets signify machine prompts]:
[your_machine%] ftp ftp.alumni.caltech.edu
[Name (...):] anonymous
[Password:] your_e-mail_address
[ftp>] cd pub/ingber
[ftp>] binary
[ftp>] get file_of_interest
[ftp>] quit
[ftp>] get file_of_interest
[ftp>] quit
:The 00index file contains an index of the other files and information
:on getting gzip and unshar for DOS, MAC, UNIX, and VMS systems.
:The latest version of ASA, ASA-x.y (x and y are version numbers), can
:be obtained in several formats.
ASA-shar.Z is a compress'd shar'd file
:of the current code.
For the convenience of users who do not have any
:uncompress/gunzip utility, there is a file ASA-shar which is an
:uncompress'd copy of ASA-shar.Z; if you do not have sh or shar, you
:still can delete the first-column X's and separate the files at the
:END_OF_FILE locations.
There are tar'd versions in compress and gzip
:format, ASA.tar.Z and ASA.tar.gz, respectively.
There also is a
:current zip'd version, ASA.zip, in which all files have been processed
:through unix2dos.
Directory /pub/ingber/0lower.dir contains links to
:these files for some PC users who may have difficulty with upper case.
:If you do not have ftp access, get information on the FTPmail service
:by: mail ftpmail@, and send only the word "help" in the
:body of the message.
:If any of the above are not possible, and if your mailer can handle
:If any of the above are not possible, and if your mailer can handle
:large files (please test this first), the code or papers you require
:can be sent as uuencoded compressed files via electronic mail.
:have gzip, resulting in smaller files, please state this.
:Sorry, I cannot assume the task of mailing out hardcopies of code or
My volunteer time assisting people with their queries on my
:codes and papers must be limited to electronic mail correspondence.
:To get on or off the ASA_list e-mailings, just send an e-mail to
:asa-request@alumni.caltech.edu with your request.
Update notices are
:sent to the ASA_list about every month or two, more frequently if
:warranted, e.g., in cases of these notices are the
:only e-mail sent to the ASA_list.
:+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Parallelizing ASA and PATHINT Project (PAPP)
:See the file parallel.txt in /pub/ingber/MISC.DIR
:+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
:\\ Prof. Lester Ingber
:\\ Lester Ingber Research
E-Mail: ingber@alumni.caltech.edu //
:\\ P.O. Box 857
WWW: http://alumni.caltech.edu/~ingber/ //
:\\ McLean, VA 22101
Archive: ftp.alumni.caltech.edu:/pub/ingber/ //
/* Lester Ingber
Lester Ingber Research *
* ingber@alumni.caltech.edu
http://www.alumni.caltech.edu/~ingber/ *
* PO Box 06440
Wacker Dr PO - Sears Tower
Chicago, IL
------------------------------
Sun, 12 Apr :10 -0700
Subject: AutoBox Financial Forecasting
I am interested in obtaining feedback from users of AutoBox who are
using the program for financial time series analysis.
Thank you. Dan.
tradeup@sj.bigger.net
------------------------------
Mon, 13 Apr :22 GMT
Subject: Multivariate Analysis Tool (New features)
Download a free multivariate analysis tool (Data Explorer For Windows
95/NT) from
The software offers you:
New to this release:
Virtual/derived variables (e.g. shift by 2 samples).
Control charts with specified limits.
And many enhancements...
The feature set includes, but is not limited to:
PCA/PLS/PCR modelling and prediction.
View your dataset and results in spreadsheet and charts.
Import dataset from local file system or
remote data sources on same or different platforms.
Point-and-click instant analysis and diagnosis.
Off-line analysis and on-line real-time prediction.
Virtual/derived variables (e.g. shift by 2 samples). [New]
Control charts with specified limits. [New]
Step-by-step demo.
And more for you to explore...
Applications include, but are not limited to, data correlation
analysis, grouping, pattern recognition, modeling, fault
detection, real-time system monitoring,
statistical process control (SPC), statistical quality control (SQC),
survey data analysis, finacial data analysis,
forecasting/predicting...
Send e-mail to
for more information.
------------------------------
Thu, 9 Apr :41 +0200
"Lestaw K. Bieniasz"
Subject: multi-curve fitting
Tuebingen, 9.04.98
Dear Statisticians,
I am new to this list, so please excuse me if this message duplicates
possible similar earlier messages. I have a difficulty with finding
adequate literature related to the nonlinear regression problem I describe
below, and I would appreciate very much your advice, as well as your help
in finding a public domain source-code software (FORTRAN, C, C++, Pascal)
that I could use or adapt for solving my problem. Here is the description:
1) I want to perform a SIMULTANEOUS least-square-type fitting of N
DIFFERENT "theoretical" curves, to corresponding sets of "experimental"
points. The curves are given by equations (all variables are real):
y1=f1(x1,P)
y2=f2(x2,P)
yN=fN(xN,P)
By saying "DIFFERENT" I mean that dependent
variables y1, y2... may all
be different (although some of them may perhaps be identical), and
independent variables x1, x2, ... may also be different for every curve,
and also functions f1(), f2(), ... may be different from each other.
The magnitude, or "physical units" of the dependent variables may be quite
different.
The only common element of all the curves is that they depend on the same
vector P of parameters. By saying "SIMULTANEOUS" I mean that I want
the parameters to be estimated in such a way so that all the curves
simultaneously fit their experimental points well in some sense, with the
same vector of parameter values.
2) The number of experimental points may be different for every curve, and
they may be located at values of independent variables (x1, x2, ...) that
are different for every curve, and not necessarily equally spaced.
3) Nothing is known about statistical errors of the experimental points,
with the exception that the errors are possibly all uncorrelated, and
subject to the normal distribution. However, the variance of the errors
may be different for every set of experimental points associated with
a single curve, and it can also possibly depend on x1, or x2, ...etc.
(i.e. the variance of the errors within every set of points is not
constant, although the functional dependence is not known). The variances
are not known. Let's assume that only values of y1, y2, ... etc.
coordinates of the experimental points are in error, and coordinates x1,
x2, ... etc. are precise. (However, I'd like to see a discussion of the
case of erroneous vales of x1, x2, ... too, if possible).
4) The regression problem is constrained, in the sense that
allowable parameter values may be limited to some finite domains in the
parameter space (for example some of the parameters must always be
positive).
5) I guess some kind of least-square fitting is adequate for this problem,
but I don't know how to sum up the squared deviations so that the problem
is properly formulated from the statistical point of view. From the
fitting procedure I would like to obtain:
a) estimates of the parameters,
b) estimates of the covariance matrix for the parameters,
c) confidence areas for the parameters, and
d) estimates of the error variances for every experimental point.
I would also like to know how to perform statistical or other tests for
the goodness of the fit in this case.
All the books that I have at hand describe only a fitting of a single
curve to a given set of points, often with additional simplifications
(constant variances of absolute or relative errors). But I am sure
someone must have studied this more general problem. Please HELP!
Please send answers directly to my current address
lestaw.k.bieniasz@uni-tuebingen.de
Best regards,
L. Bieniasz
*-------------------------------------------------------------------*
Dr. Leslaw Bieniasz
| temporary address: (3rd March 1998 - 31st July 1998)
Institut fuer Organische Chemie, Universitaet Tuebingen
Auf der Morgenstelle 18, 72076 Tuebingen, Germany.
E-mail: lestaw.k.bieniasz@uni-tuebingen.de
*-------------------------------------------------------------------*
| permanent address:
| Institute of Physical Chemistry of the Polish Academy of Sciences,|
| Molten Salts Laboratory, ul. Zagrody 13, 30-318 Cracow, Poland.
nbbienia@cyf-kr.edu.pl
http://www.cyf-kr.edu.pl/~nbbienia
*-------------------------------------------------------------------*
------------------------------
Mon, 13 Apr :37 +0300
Amir Hetsroni
Subject: estimating a nonlinear regression function
This is a multi-part message in MIME format.
--------------FDC2DE0E9B59C364DF4BDD3F
Content-Type: multipart/
boundary="------------51FF97C7E8AAF13"
--------------51FF97C7E8AAF13
Content-Type: text/ charset=us-ascii
Content-Transfer-Encoding: 7bit
I am working on a database that investigates complex relationship
between interest in an object
and the inclination to participate in activities in favor of the object.
Based on the "principal components of attitude" theory by Guttman I
expected to find a W shape relationship between the interest component
and the activities component.
It appears that at least as far as it concerns the object of national
history - mission accomplished (see the control chart below plotted by
However, when it comes to calculating a regression function I am having
a problem. Clearly, the most accurate function is not linear, looks more
like to parabolas, but how do I write this function? What value should
the parameter(s?) have?
I tried to draw a scatter plot to learn more about the function but it
appears that when the sample is large even fairly large linear
relationship (r>0.6), not to mention this case, look messy on a scatter
plot (at least in SPSS).
Can someone give a hand with the estimation of the function or at least
AMIR HETSRONI
The Department of Communication
The Hebrew University of Jerusalem
E-Mail: msamir@mscc.huji.ac.il
amir_h@barley.cteh.ac.il
--------------51FF97C7E8AAF13
Content-Type: multipart/ boundary="------------93F230C169A34EEB57D84B0C"
--------------93F230C169A34EEB57D84B0C
Content-Type: text/ charset=us-ascii
Content-Transfer-Encoding: 7bit
I am working on a database that investigates complex relationship between
interest in an object
and the inclination to participate in activities in favor of the object.
Based on the "principal components of attitude" theory by Guttman I
expected to find a W shape relationship between the interest component
and the activities component.
It appears that at least as far as it concerns the object of national
history - mission accomplished (see the control chart below plotted by
&However, when it comes to calculating a regression function I
am having a problem. Clearly, the most accurate function is not linear,
looks more like to parabolas, but how do I write this function? What value
should the parameter(s?) have?
I tried to draw a scatter plot to learn more about the function but
it appears that when the sample is large even fairly large linear relationship
(r>0.6), not to mention this case, look messy on a scatter plot (at least
Can someone give a hand with the estimation of the function or at least
AMIR HETSRONI
The Department of Communication
The Hebrew University of Jerusalem
E-Mail: msamir@mscc.huji.ac.il
&amir_h@barley.cteh.ac.il
--------------93F230C169A34EEB57D84B0C
Content-Type: image/jpeg
Content-ID:
Content-Transfer-Encoding: base64
Content-Disposition: filename="C:\OLDWIN\TEMP\nsmailGK.jpeg"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}