bug in Excel 2003; need to install Excel 2002

[Guest]

I have Office 2003, and I found a bug in Excel 2003. The linear regression
results using Excel 2003 differs with that of Excel 2002 for some cases.

How do I can uninstall Excel 2003 (and install Excel 2002) without affecting
the other Office apps (like Word 2003 and Powerpt2003)?

 

[Frank Kabel]

Hi
Excel 2003 has solved some bugs/errors with statsitical functions. So I
would guess the 'new' result are more correct than the old ones

 

[Jerry W. Lewis]

In general, Excel 2003 algorithms are far superior to previous versions
for statistical and probability calculations
http://support.microsoft.com/default.aspx?scid=kb;en-us;828888

The one exception is that when LINEST returns a coefficient that is
exactly zero, that zero value may be in error. I have suggested
work-arounds
http://groups.google.com/groups?selm=412980D4.5040305@no_e-mail.com

If you think you have a different situation where LINEST is worse (not
just different) in 2003, please provide more information.

As to your specific question, Multiple versions of Office may coexist
peacefully on your computer.
http://www.microsoft.com/office/ork/022/022.htm
I believe that the last installed version is the one that will be used
by default when you double click on a file (though I have only tried it
with the newer version installed last).

Jerry