![]() |
![]() |
Back |
There has been recent discussion on STAT-L regarding Simpson's Paradox, including worthy contributions by John Vokey and others. By chance, I have just completed and issued a program which undertakes to provide numerical demonstrations of Simpson's Paradox. The program also provides options for a filtering criterion based upon the value of the Pearson Chi-Squared statistic, and for dealing with a number of subtables/strata which may be greater than 2. The program is offered as standard (ISO/ANSI) conformant Pascal source (ASCII text), and although the source is large it should complie on any Pascal compiler system. It is located as part of the resources associated with the discussion list exact-stats at the UK Mailbase site. Set your browser URL as : http://www.mailbase.ac.uk/lists/exact-stats/files/genxtabspas.txt or send a 1-line e-mail message as follows : to: mailbase@mailbase.ac.uk message send exact-stats genxtabspas.txt (this causes the text to be sent to you as an e-mail message). I should be interested to hear of any feedback from users of this new resource/toy - messages direct to author please : jw34@liverpool.ac.uk Norman Marsh University of Liverpool jw34@liverpool.ac.ukReturn to Top
Hi I would like to estimate the entropy of a random variable from a sample without assumptions about the density function, or in other words a nonparametric estimator of the entropy. Does someonehow to do this? Thanks for any help. -- Peter Parzer Peter.Parzer@urz.uni-heidelberg.deReturn to Top
Is there any known stochastic process that can related mean or standard deviation of different averaging time? Please email me if you have information. -- Liang Lu, President Lu Scientific Consulting Tele: 817-320-3974 2806 Country Club Fax: 817-484-5282 Denton, TX 76201 Email: llu1@airmail.netReturn to Top