Experience with micro-computer statistics packages

[Laurie Moseley]

MICRO-COMPUTER STATISTICS PACKAGES

I am looking for a relatively simple statistics/ data analysis package
for use on an IBM-PC or AT. The statistics do not need to be sophisticated.
What we need de minimis are:

1. Data-manipulation:

Facilities for grouping (like RECODE in SPSS)
Facilities for constructing variables (like COMPUTE in SPSS)
Preferable (but not essential) data selection (like SELECT IF)
Preferable (but not essential) logarithmic transformation
Preferable - some simple graphical display e.g. scattergrams,
or possible linking to a graphics package.

2. Data presentation

Cross-tabulation with percentages
Preferable - 3/4 way cross-tabulation (but not essential)
Mean, median, SD
Equivalent of SPSS BREAKDOWN
Some form of labelling

3. Stats (All desirable but not essential)

Chi-square /Fisher exact
Spearman or Kendall non-parametric correlation
Kendall coefficient of concordance
Some simple linear regression

There are plenty of packages on the market. However, I have no idea how well
they work in practice. I would be grateful if anyone who has had either very
positive or very negative experiences with any such packages could share that
experience with me. Perhaps I should note that the package is intended for
desk-top analysis and data-exploration by some social scientists (and I mean
social scientists not woolly theoreticians !).

Of course, if there is any public domain material, we would be overjoyed to
hear about it.

With thanks

Laurie Moseley
University College, Swansea

--
If you have skill, you have power.
If you have power, you need responsibility.
Never demonstrate one without the other.