[WEB4LIB] RE: cataloging web sites

[Karen Harker]

I wonder if there have been any studies on the effectiveness and usability of these two methods of cataloging?

We will be re-considering this issue in the very near future. We currently use a moderately-simple SQL Server database to "catalog" our electronic resources.  We will be implementing Encompass and will likely be moving these "links" into Encompass. We now need to reconsider to what level to catalog them.  

It would be nice to have some evidence on which to base our decisions.



>>> Harvey Brenneise <HBrenne at MPHI.org> 6/21/02 1:45:24 PM >>>
I've done it both ways (see www.mchel.org).  I see problems with both.  With
a "mulver" (multiple version) MARC record you possibly keep the catalog
simpler for the user who can then see all versions on the same record.  On
the other hand, you also create complications (fixed fields that aren't
entirely accurate, problems with the GMD in |h of certain title fields, 5xx
fields that only really apply to one "version" of it.  Even worse, some of
the titles may also be deceptive when applied to all versions (titles with
CD-ROM in them, Online in them, and so forth).  And is one absolutely sure
that the print and the online are EXACTLY equivalent?  Sometimes (for
example National Academy Press), probably so.  Other times, doubtful.

And then there are serials.  What if you have access to the same serial from
more than one source?  You can certainly put in more than one 856, but then
how do you deal with different holdings, different access information, and
so forth.  And on and on.  I personally tend toward separate records myself,
at least in most cases.

Harvey Brenneise
Michigan Public Health Institute.
hbrenne at mphi.org