>>>>> "DR" == Dennis Roberts <[EMAIL PROTECTED]> writes:
DR> i don't know the answer to this but ... i have a general
DR> question with regards to using spreadsheets for stat analysis
DR> why? ... why do we not help our students and encourage our
DR> students to use tools designed for a task ... rather than
DR> substituting something that may just barely get us by?
DR> we don't ask stat packages to do what spreadsheets were
DR> designed to do ... why the reverse?
Let's think about this in another way. Spreadsheets do form a nice
means for thinking about datasets. In fact, a good number of "decent"
statistical packages can interact with spreadsheets, or provide a
spreadsheet-style interface.
There are a number of rather useful EDA tools that spreadsheets like
Excel provide (easily programmable in R or XLispStat, to name my two
favorite statistics languages).
So, what's missing in spreadsheets? the ability to designate dataset
structures for statistical modeling with the same flexibility as a
decent statistical language (I'm thinking generalized linear models
(not general linear models!), correlated regression, event-time
regression analysis; robust numerical computations at the same
level, and a large set of tools.
However, more critical is, "what's missing in statistical languages?",
and I think that the answer are tools to help novices ramp up
"properly".
I'll leave the definition of "properly" open for debate.
--
A.J. Rossini Rsrch. Asst. Prof. of Biostatistics
U. of Washington Biostatistics [EMAIL PROTECTED]
FHCRC/SCHARP/HIV Vaccine Trials Net [EMAIL PROTECTED]
-------------- http://software.biostat.washington.edu/ --------------
FHCRC: M-W: 206-667-7025 (fax=4812)|Voicemail is pretty sketchy/use Email
UW: T-Th: 206-543-1044 (fax=3286)|Change last 4 digits of phone to FAX
Rosen: (Mullins' Lab) Fridays, and I'm unreachable except by email.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at
http://jse.stat.ncsu.edu/
=================================================================
