---------- Forwarded message ----------
Date: Fri, 12 Jul 1996 15:17:51 EDT
From: Albert Henderson <70244.1532@COMPUSERVE.COM>
Subject: The end of "Access not Ownership"

 Heinrich C. Kuhn <hck@IPP-GARCHING.MPG.DE> writes

> In my response ((Thu, 11 Jul 1996 11:39:47 -0400) to Albert
> Henderson's kind reply to a request for further information
> by me) I expressed some doubts as to the reliability of statistical
> information drawn from bibliographical databases as a basis for
> conclusions on the number of publications of a certain type in a
> certain period of time. Yesterday I used data from INSPEC and
> INSPHYS. Today I tried the more or less same thing in MEDLINE
> (clinical medicine etc.) and what I found is far more apt to
> fundament my pledge for caution than what I mailed yesterday.

[snip]

> Thus: Be cautious: Bibliographical databases are valuable tools
> for quite a number of things, but they do not necessarily provide
> reliable data for conclusions about the development of the numbers
> of a certain type of publications in a certain field!

This is absolutely correct. As I noted in my reply, I had double checked
with the author at CAS whether any change in editorial policy might have
affected the statistical representations because our main interest was the
extent to which the CAS statistics represented the history of publishing
in chemistry. The author took the time to check with CAS's editors and
came back with the assurance that there had been no change in coverage.

MEDLINE has had quite a different history. In 1879 Index medicus covered
21,169 articles estimated by NLM as the totality of biomedical literature.
One hundred years later, the director of NLM advised that coverage was
estimated at about 12 per cent. (Corning, M.E., and M M Cummings.
"Biomedical Communication, in Advances in American Medicine. New York.
Macy Foundation. 1976. vol. 2 p. 722ff).

Richard Kaser, director of National Federation of Indexing and Abstracting
Services, describes the slowing of coverage by major databases reflected
in their production statistics (PRQ 11,3:10-24 1995). He attributes
reduced coverage to the impact of subscription cancellations.

In the case of NLM, the reduced coverage, of course, is more likely to
spring from the unique quality of government funding and management. The
indication that no monographs are covered means that they must be
cataloged, classified, and on the shelf in order for them to be any use to
researchers and practitioners. As a side note, the items covered by
MEDLINE 1976-81 were not "monographs" in the sense of a book by a single
author. They were collections of articles that were not published as a
part of a serial. MEDLINE covers serials and has not covered "monographs."
More important, you have demonstrated how MEDLINE expanded then dropped
its coverage of conference papers.

More important we could see that without adequate coverage by A&I
services, the foundation of "access not ownership" is doomed.

Thanks for calling this to our attention.

Albert Henderson, Editor, PUBLISHING RESEARCH QUARTERLY
70244.1532@compuserve.com