---------- Forwarded message ----------
Date: Fri, 13 Apr 2001 10:37:58 -0400
From: "DuBose, Stefanie" <DUBOSES@MAIL.ECU.EDU>
Subject: Re: E-J statistics (David Goodman)

Hi David,

This is an issue we are struggling with as well.  As mentioned by many of
our colleagues, I am also simply working with what I'm given by the
vendors/publishers.  One option we have discussed, and I certainly make no
claim to understand the technical aspects, is to place a redirect script in
each ejournal URL in our ejournal database that will help us track which
journals our patrons are at least going to.  Again, there will be gaps in
our data collection: it doesn't account for URLs in our catalog and the
titles our users have bookmarked.   A second method we're looking into is to
bring the issue of uniform vendor statistics to the state level, where the
bigger negotiations take place.

Stefanie DuBose
Collection Development/Serials Librarian
Joyner Library
East Carolina University
Greenville, NC 27858-4353
(p)252-328-2598
(f)252-328-4834
duboses@mail.ecu.edu
The views expressed are my own and do not necessarily represent those of
either Joyner Library or East Carolina University.

---------- Forwarded message ----------
Date: Thu, 12 Apr 2001 17:40:30 -0400
From: David Goodman <dgoodman@Princeton.EDU>
Subject: Re: E-J statistics (Mitch Turitz)

Mitch, all of your comments are certainly true, and I am sure you, I,
and most of us could add a great many other sources of uncertainty. But
I do not think that this means we must go blindly ahead spending ever
increasing sums of money without knowing how the material we buy is
being used. Up to now, like many, I have relied on scattered data,
anecdotes, and complaints. Even the best counts of physical item use,
like Wisconsin's, acknowledge that the actual values might range from
100% to 200% of that reported, and that is dealing with well understood
behavior and techniques.  Most library statistics involve very
inaccurate counts--the best we can do is have some indication of where
the errors are most likely to be, and in what direction. All this is a
reason to start working, not a reason to abandon measurement.