🚀 go-pugleaf

In article <4a4rqm$l70@hecate.umd.edu>, Rod Martens <martens@eng.umd.edu> wrote:

> Greetings,
> Can anyone recommend relatively inexpensive statistical software packages for 
> windows?  (I'm on a student's budget.)
> Thanks for the help,
> Rod Martens.

Last week I posted a long note about stat s/w evaluation for NT - which I
am now completing.  There are many good packages, but I would strongly
recommend for someone on a shoe-string budget that if you already have
Excel, you supplement it with Ken Berk and Patrick Carey's book, "Data
Analysis with Microsoft Excel 5.0 for Windows," publ. by Course
Technology, Inc., ISBN 1-56527-525-X
AND buy their Excel Add-ins, which accompany the book on floppy.

Having made this recommendation, I must caution that there are many things
available in a real stat package that you will STILL not be able to do. 
Our list of such items ran like this (different people will have different
versions):

analysis of covariance
ARIMA (auto-regressive-integrated-moving average) model identification /
estimation
boxplots
canonical correlation
categorical data analysis (MLE for log-linear models, generalized logits)
cluster analysis - hierarchical, exclusive, fuzzy, etc.
contingency table analysis
contour plots
definition & testing of contrasts, estimable functions
design of experiments aide / catalog - 2-level, response surface, mixture,
D-optimal
discriminant analysis
error term identification or specification
filter development
fourier / spectral analysis
general linear modeling
leverage plots
logistic regression
multiple comparisons
neural net analysis (limited)
non-linear fitting
non-parametric analysis
principal components analysis, factor analysis
random effects models
response surface models
robust fits / statistics
screening models & analysis
statistical tolerance limits / intervals
stepwise regression
state space modeling
survival analysis
3D spinning plots
time series analysis (see ARIMA, fourier, transfer-function)
transfer-function modeling
variance components analysis - method of moments, (restricted) maximum
likelihood
wavelet analysis (limited)


Happy analysis!

P.S. We recommended JMP� and SAS�, but think that SPSS, S-plus (not yet on
NT), Statgraphics, Systat, and many other packages deserve serious
consideration.

-- 
The views expressed are not necessarily shared by Alcoa.

David Coleman
Sr. Technical Specialist - Statistics
Applied Mathematics and Computer Technology-ATC-D
Alcoa Technical Center
100 Technical Dr.
Alcoa Center, PA 15069