public library usage statistics

[Peterson, Glenn]

While Larry's points are valid, I don't know that all hope is lost. If
you and I both use web servers that log using the common format and we
use the same software program to evaluate our web logs we can compare
apples and apples. Even if we use different programs, we may be able to
configure them so we are comparing Cortands with Macintoshes. HCL uses
Webtrends, a popular web log analysis software, for example. Our weekly
reports are available on the Internet, contact me for details.

I would have sworn that the National Center for Education Statistics
(http://nces.ed.gov/), which tracks PL statistics and allows
comparisons, used to track web site visits. The numbers were all over
the board, for the reasons Larry cites. Maybe they gave up?

-----------------------------------------------------------------
Glenn Peterson                gpeterson at hclib.org
Web Administrator             http://www.hclib.org
Hennepin County Library     voice: 952-847-8659
12601 Ridgedale Drive           fax: 952-847-8642
Minnetonka, MN 55305         
-----------------------------------------------------------------


Date: Fri, 4 Feb 2005 07:46:41 -0800 
From: "Scritchfield, Larry" <LScritch at MAIL.co.washoe.nv.us>
To: Multiple recipients of list <web4lib at webjunction.org>
Subject: RE: public library usage statistics
Message-ID:
<52D1E8A877040744B1AA562F4AD9192B085B93F9 at mail.co.washoe.nv.us>

Jim >The public
Jim >library where I work is becoming increasingly interested in
measuring
Jim >and evaluating the public's use of its branch library web pages.
Are
Jim >there any forums or resources available for sharing and comparing
usage
Jim >statistics for public library web sites?

Any comparison of (say) my statistics with yours would be apples and
oranges.
The reason for that is, there is no accepted definition of "hit" or
"visit."
Some sites count hits as every .gif, .jpg, .html/.php/whatever, .css,
and
script. Others, only html.

And if you go for visits not hits, there is a time factor. When do you
start
counting hits from the same IP as coming from a new visitor? (Here, we
guess
that the average visit lasts 10 minutes.)

Proxies and caches throughout the Internet also muddy the picture.

So all you can practically do is track trends at your own location.

Larry Scritchfield                       lscritch at mail.co.washoe.nv.us
Internet Services Librarian                             (775) 327-8349
Washoe County Library System                      www.washoe.lib.nv.us


------------------------------

End of WEB4LIB Digest 3628
**************************