[829] in Pthreads mailing list archive
New Version Release
daemon@ATHENA.MIT.EDU (RCKnodt@aol.com)
Sat May 8 13:03:45 1999
From: RCKnodt@aol.com
Date: Sat, 8 May 1999 12:48:17 EDT
To: pthreads@MIT.EDU
Announcement: Our apology in advance for any cross-posting.
To be removed from our mailing list, see the end of this e-mail.
MODSTAT Ver. 11 now does over 230 statistical tests and routines with
NO PROGRAMMING on the part of the user. Completely menu drive, the
software handles descriptive statistics (over 35 pieces of information are
calculated).
The software does moving averages, various tests to check for randomness in a
set of data. It does Geometric and Harmonic Mean calculations, Permutations
and Combinations and moving average.
It does ten different ANOVA's including nested and mixed designs, and
three types of Latin Squares. Three types of covariance and three types
of Bayes' problems. Friedman's test and Kruskal-Wallis test are both
included.
It handles Linear, nonlinear, power, log, exponential, polynomial (up to 13th
power), multiple linear (up to 8 predictor variables), curvilinear and
multiple curvilinear regression. There is a complete analysis of residuals.
Kendall's and Spearman's tests, Biserial and Point-Biserial tests.
Various forms of the t-test and Z-tests are included. It does comparison of
samples to populations (mean and proportion) It allows comparison of
different proportions. There is no need to look up probability values
since the software calculates and reports them.
The software can handle all non-parametric tests as cited by Siegel as well
a number of other non-parametric tests. 1-, 2- and 3-way Chi Square designs
are included, McNemar and repeated measures Chi Square. Various methods to
determine sample size are provided along with a random number generator to
select samples.
The software also does many less-often used tests some of which are the
Change Point test, Odds ratio, Yules, Walsh, Lambda, Somer's, Omega Squared,
Robinson's 'A', Siegel-Tukey, Jonckheere, Coefficient of Alienation,
Lilliefor's test, and D'Agostino's test.
Produces charts of residuals, box-plots of data, group comparison charts.
Confidence intervals are determined where needed and various distributions,
(t-, F-.Chi-Square, Normal, Binomial, Exponential, Poisson, Geometric,
Hypergeometric, and Multinomial) are included. Too many other tests to
mention.
A FREE two-month trial is available. Over 230 tests and routines.
Registration fee is only $22.00 (about a dime a test) if you decide to
continue using the software after the trial period.
The learning curve is almost flat. Within 10 minutes after first setting up
the software you can produce complete summary charts and comparison of groups
using 1-Way ANOVA, or run a Pearson Linear Regression analysis, or completely
analyze a set of data. There are also two routines to help the user
determine the proper test to use. Suggestions are made when a test is run
regarding the conditions of the test and alternative tests are suggested if
the conditions are not met.
http://members.aol.com/rcknodt/pubpage.htm
<A HREF="http://members.aol.com/rcknodt/pubpage.htm">MODSTAT</A>
For a FREE two-month trial of the software, e-mail your request along with
your complete name and complete postal mailing address to:
statware@technologist.com <A HREF="mailto:
statware@technologist.com">statware at tech
</A>
If you would like a list of all the tests and routines, e-mail your request
to:
statistics@schoolmail.com
NOTICE: To be removed from our mailing list, send an e-mail including your
e-mail address to: statware@aol.com <A HREF="mailto:
statware@aol.com">statware at AOL
</A>
NOTE: SOME tests, but not all will accept data from Excel and Quattro
with slight modification.
Runs on 386, 486, Pentium, K6, Windows95, Windows98 and MacIntosh (using
VirtualPC). Also runs on older HP 100 and 200 hand-held PalmTops.
"Can you afford not to have MODSTAT on your PC? For a registration fee of
$22.00 you have software that will do over 230 Statistical Tests and
Routines. About a dime a test."
The software was produced by Dr. Robert C. Knodt, (U.S.A.)
statistical.help@workmail.com
<A HREF="mailto: statistical.help@workmail.com">Dr. R. Knodt</A>
