Date:         Wed, 3 Dec 2003 09:24:55 -0500
Reply-To:     "DePuy, Venita" <depuy001@DCRI.DUKE.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "DePuy, Venita" <depuy001@DCRI.DUKE.EDU>
Subject:      Re: off-topic question: tail index
Content-Type: text/plain

Further side note regarding Excel -
I use their Data Analysis add-in occasionally to do a quick regression, t
test etc. - handy if you're at home!
BUT Excel isn't always accurate  (Knusel 1998 and McCullough & Wilson 1999,
both in Computational Statistics and Data Analysis)

So I do still use it, and personally I haven't found any discrepancies, but
I'd always double-check my results in another package before publishing
anything based on it.

-Venita
> ----------
> From:         David L. Cassell[SMTP:cassell.david@EPAMAIL.EPA.GOV]
> Reply To:     cassell.david@EPAMAIL.EPA.GOV
> Sent:         Tuesday, December 02, 2003 6:25 PM
> To:   SAS-L@LISTSERV.UGA.EDU
> Subject:      Re: off-topic question: tail index
>
> Tom <tom-l@AON.AT> wrote:
> > An off-topic question:
> > I would appreciated some hints on how to calculate the tail index with
> MS Excel.
>
> One big hint:  don't.
>
> Never attempt to program serious statistical analyses in any
> spreadsheet.
> Use an appropriate piece of software.
>
> > What I am particularly looking for are statistical methods for
> > distinguishing non-trending time series from trending time series.
>
> If you don't want to do this using SAS/ETS , then take a look at
>
>     http://www.autobox.com/index.html
>
> HTH,
> David
> --
> David Cassell, CSC
> Cassell.David@epa.gov
> Senior computing specialist
> mathematical statistician
>