Thanks to all those, too numerous to mention, who sent me their opinions of
various packages. I haven't avoided the need to do my own evaluation, but I
now have a list of 12 potential packages to look at instead if 6. I feel the
opinions expressed are too subjective to summarise in detail publicly, and
too few to form a useful survey of useage. But useful none-the-less, so
thank-you all.
Maybe I should have defined "simple statistics" a bit more clearly - what I
had in mind is what my users will have been taught on their various service
courses - "Statistics for experimental Biologists" or whatever.
There were a couple of replies along the "don't do it" lines, with which (on
bad days) I have some sympathy. But we live in the real world, and they are
going to do it whether we like it or not, and our job is to try and persuade
them to at least talk to us first, and to try and make sure they are using
good tools. Anyway I have some intelligent clients who are quite capable of
doing the "simple" (variously defined) things themselves and know when they
are out of their depth and to seek advice. Part of our job is to train them
to get it right.
Finally the Excel question. Opinions here are widely divergent, as are mine
from day-to-day. There are clearly problems with speed and reliability of
macro-based approaches. My feeling today is that it is better to give the
users the discipline of a "proper" statistical package as it at least makes
them think about their data, but that that package might itself link to
Excel and appear on the Excel menus.
A subjective response to subjective replies to a subjective question:-)
Thanks to you all,
Steve.
Steve Roberts
Biostatistics
Paterson Institute, Christie Hospital.
Manchester, UK
[log in to unmask]
|