<- file stat 97www.html -> Here are more Sites that may be of interest. - I will add to these or correct them, or remove them if they are defunct, when corrections are offered, but I am just starting to post whatever is offered. RFU. Bayes linear updating [B/D], Wilkerson; Random number generation, Miller; Actuarial exams (plus commentary); Journal of Applied Statistical Reasoning (with sites); SPSS archive, JSE (SPSSX-L); SPSS Help archive, UC, Boulder, Colo.; Data Analysis Briefbook, CERN; IRT - NOHARM, McArdle; Meta analysis, Solomon; ROC sensitivity/specificity, Metz Software Metrics Research Lab Kappa macros, SPSS Spacially Oriented Research in Ecology, WSU * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

"Bayes linear updating" [B/D]

D.J. Wilkinson, 16 Jun 1997

[B/D] home page href=http://www.ncl.ac.uk/~ndjw1/ Subject: Re: Q: Tool for Bayesian Dr Darren Wilkinson mailto:d.j.wilkinson@newcastle.ac.uk [B/D] is a computing environment for Bayes linear updating. Bayes linear updating is formally equivalent to Gaussian updating. See the Bayes linear home page: http://fourier.dur.ac.uk:8000/stats/bayeslin/ and the [B/D] home page: http://fourier.dur.ac.uk:8000/stats/bd/ for more info. A Windows version is available free over the net. *--------

Normal random number generation

Alan Miller, 19 Jul 1997 alan@vic.cmis.csiro.au, milleraj@ozemail.com.au http://www.ozemail.com.au/~milleraj The most efficient algorithm I know is that in TOMS algorithm 712 which can be downloaded from the toms directory of netlib. There is a Fortran 90 version at my web site. *--------

Actuarial exams

SOA site

CAS site http://www.soa.org http://www.casact.org Mark Menzie, 14 May 1997 From: mamenzie@aol.com (Mamenzie) Subject: Re: ACTUARIAL Exam #110 Tricks about 110 revolve around solving for all the different parameters of distributions. Sample size required to acheive a given power, alpha etc...Solving a simple linear regression problem given one of the parameters. Lots of maximum likliehood and Neyman-Pearson type problems. And there are always a few abstract problems. They do give you some tables, so you don't need all the pdf's memorized. You should have the binomial, hypergeometric,eponential and Poisson down pat. You should also understand the relationships between: t and normal; normal and chi-square; chi-square and F; and binomial, normal and Poisson. I'm not familiar with the Freund book, but I imagine that it covers most if not all the bases. Try getting a study guide with old tests in it and see if you can work the problems. If you can't from memory, look up the details in your book of choice. The SOA syllabus of examinations lists page references for particular topics of interest. If you can find solution methods for the problems in your book and the topics on the referenced pages are well covered, I would say it's fine. Reading several different discussions on topics you are having problems with can be helpful, too. For 100, the major thing I remember is that there is a lot of multivariate on it. Integration, directional derivatives, optimization and constrained optimization. There is also some set theory, linear algebra and a little pre-calculus geometry. For motivation, try finding some applied problems relevant to the field. I have not found an actuarial job to date, so I can't give you too much insight. From what I've seen, transformations and conditional probability are very important. Like using data about seat belt use, driving speeds, epidemics, alcohol and drug use etc to infer effects on future claims. The actuary also needs a detailed understanding of regulation and certain aspects of Economics. The SOA and CAS have web sites with some more detailed information. *--------

Journal of Applied Statistical Reasoning'

Zaine, 13 Jun 1997 From: zaine@aol.com (Zaine) Check out the new 'Journal of Applied Statistical Reasoning' site at http://www.o2.net/~jasr.html They have a 'related statistical sites' section that is the best I've seen. *--------

SPSS archive Date: Mon, 24 Mar 1997 From: "Tverdek, Edward" <etverdek@SPSS.COM> The search engine is located at: gopher://jse.stat.ncsu.edu:70/7waissrc%3A/othergroups/spssxl/spssxl *--------

SPSS Help Archive From: brumbaug@sobek.Colorado.EDU (Susan Brumbaugh) University of Colorado, Boulder. http://socsci.colorado.edu/LAB/STATS/SPSS/ - list of links to other helpful Internet resources, including: SPSS Web Site SPSS FTP Archive Keywords (SPSS Publication with helpful articles) Articles by David Nichols (from Keywords) SPSS Macros (syntax files written by David Nichols to perform tasks unavailable through windows) Searchable Gopher Archive of all comp.soft-sys.spss and SPSSX-L posts gopher://jse.stat.ncsu.edu/11/othergroups/spssxl *--------

Data Analysis Briefbook Date: Fri, 31 Jan 1997 From: rkb@afsmail.cern.ch The Data Analysis Briefbook: an extended glossary (or condensed handbook), covering subjects in statistics, computing, analysis. Gives an overview and many formulae, along with solid references to literature for in-depth reading. Available on the web at URL: http://www.cern.ch/Physics/DataAnalysis/BriefBook/ *--------

John McArdle's Home page, NOHARM http://kiptron.psyc.virginia.edu:80/jack_mcardle/ Date: 18 Mar 1997, answer to the question: >I'm looking for a software - NOHARM - that does analyses in the Item >Response Theory field. It was written by Dr. Colin Fraser from the >University of New England, Armidale, Australia. Can anyone direct me >to this person via Internet. *--------

Meta-analysis Date: Sun, 13 Apr 1997 From: Paul Bernhardt <Paul.Bernhardt@m.cc.utah.edu> Here is a very good web site for meta-analysis. It may lead you to an answer to your question. http://www.mnsinc.com/solomon/MetaAnalysis.html *--------

ROC sensitivity vs specificity From: Charles Metz <c-metz@uchicago.edu> 30 Jan 1997 Subject: Re: ROCFIT and CORROC2 Any or all of our current University of Chicago software for ROC analysis (including ROCFIT and CORROC2) can be downloaded at http://www-radiology.uchicago.edu/sections/roc/software.cgi *--------

Software Metrics Research Lab From: agray@commerce.otago.ac.nz (Andrew Gray) Subject: Papers from the Software Metrics Research Laboratory Date: Tue, 18 Mar 1997 This is just a reminder that the SMRL has available a number of online papers (on topics including neural network and statistical modelling techniques applied to software metrics, multimedia metrics, GIS metrics, authorship analysis, etc.). These can be found from the SMRL homepage whose URL is in my signature below. http://divcom.otago.ac.nz:800/COM/INFOSCI/SMRL/home.htm *--------

SPSS Kappa macros available From: nichols@spss.com (David Nichols) Subject: Kappa macros available ... a macro that will compute Kappa for the non-square case. It's actually been up on the SPSS ftp site since last week, at ftp://ftp.spss.com/pub/spss/statistics/nichols/macros in nskappa.sps. The read me file is nskappa.rm. .... The macro computes Kappa, both ASE measures given in CROSSTABS, the asymptotic significance for testing population Kappa=0, and also adds 95% confidence interval bounds under the alternative (using ASE1). A second macro is also there, providing the multi-rater Kappa statistic and standard error discussed by Siegel and Castellan. This is mkappasc.sps, with read me file mkappasc.rm. I gave it the name mkappsc rather than just mkappa, because there are multiple versions for multiple rater problems, .... *--------

Spatially Oriented Research in Ecology From: Jason Thomas <jthomas@wsu.edu> Subject: spatial research web site I am maintaining a web site called SpORE (Spatially-Oriented Research in Ecology) which is dedicated to providing a guide to some of the resources on the web which relate to spatial aspects of ecology, natural resource, and statistics issues. The URL is: http://www.wsu.edu:8080/~jthomas/spore.html *--------*

tin newsreader Subject: tin newsreader developments Date: 8 May 1997 > Unfortunately the original author [of tin] does not seem to support > it any more. Does anybody know if the maintenance has been taken > over by somebody else ? See <URL:ftp://ftp.akk.uni-karlsruhe.de/pub/news/clients/tin-unoff/>. * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Document by Rich Ulrich. E-mail to wpilib+@pitt.edu

FAQ top.

Ulrich home page.

Ulrich FAQ. http://www.pitt.edu/~wpilib/stats99.html