On 3/6/13 3/6/13 € 1:09 PM, "Stephen Hearn" <[log in to unmask]> wrote:

> Another interesting extension of this would be a set analysis of the
> records containing the field or subfield--e.g., for 583, a date of
> publication profile, a date of last update profile, proportion of
> identified government documents, etc.

An interesting idea that would mean creating a new set of procedures that
I'm not prepared to tackle just yet. But if a compelling case could be made
for a thorough profile of records with certain tags it would be something to
consider. But I would have to know that there was some action that would be
taken with the information, not just "oh, isn't that interesting".

> Also, would it be possible to mark the subfields which are found in
> WorldCat but not in the MARC standard, e.g., for 583, subfields t, w,
> and y? Either that, or to exclude fields which fail validation from
> the count?

I'm not set up presently to validate the data, but if I were I would mark
the subfields but not suppress them. My goal was (and so far remains) to
report on the existing situation. Having said this, I have suppressed
*fields* that do not exist in MARC, but those spurious fields (and their
associated OCLC record numbers) have been reported to WorldCat Quality
Control staff so they may be fixed (presumably in most cases the data moved
to fields that *do* exist). Thanks,
Roy