RE: eJournal interface can influence usage statistics

["MUKHERJEE, Mithu" <mithu.mukherjee@oupjournals.org>]

The following response was posted to Liblicence on Wednesday 27th July.
Please accept my apologies: the post actually relates to Phil Davis' post
"eJournal interface can influence usage statistics", not Calculating the
"Cost per Article in the Current Subscription Model".

Mithu Mukherjee 
Communications Executive 
Oxford Journals 
Oxford University Press 
Great Clarendon Street 
Oxford 
OX2 6DP 

#01865 354471 
mithu.mukherjee@oupjournals.org 

A response posted on behalf of Richard Gedye, Sales and Marketing
Director, Oxford Journals

As Oxford University Press is mentioned several times within Davis' and
Price's article on how the interface of an online journal can influence
usage statistics, I felt it would be useful to comment on two specific
issues from Oxford Journals' perspective:

1. Links from Tables of Contents:

On page 6 of the article, the authors state "A researcher starting at
the Table of Contents page on the HighWire interface is required to
first download the article in HTML". Oxford Journals has now changed
this linking strategy within the TOC interface so that all our journals
offer links to Abstracts (where they exist), Full Text HTML (if it
exists), and Full Text PDF (if it exists). For example:

http://hmg.oxfordjournals.org/current.dtl

This new interface applies to all content in current and subsequent
journal issues, and has also been implemented for the back archive of
every journal in our online collection.

It is worth dwelling for a moment on why the previous interface design
was in place (that of linking to HTML articles, where they existed, in
the first instance). Until relatively recently PDF functionality was
much less satisfactory than it is now - PDFs were not well integrated
into the average browser experience, they were slow to download, and
often impossible to print on available institutional printers. The
previous interface was designed, therefore, to support a user friendly
preference for speed in scanning the literature (in HTML form) before
potentially choosing to download or print the article (either in HTML or
as a PDF).
This interface design preceded the introduction of COUNTER compliant
usage statistics, so it would be anachronistic to surmise that it was
designed with intent to artificially boost usage statistics. The authors
are correct in their surmise that "it is entirely possible for a
publisher to optimise its interface to maximise the number of article
downloads" (page 7), but both Oxford Journals and HighWire (who host the
entire Oxford Journals collection) would like to stress that this was
never the intention of our previous interface design.

2. Links from Crossref

On page 6 of the article, the authors also state "it was discovered that
HighWire and Oxford (which uses the HighWire platform) direct all
external Crossref links to the HTML version, while Blackwell, Nature and
Wiley direct external links to the abstract page". This statement is
incorrect as it relates to Oxford Journals. Crossref links are based on
DOIs, and our DOIs resolve to:

* The abstract (if there is one) (e.g.
http://dx.doi.org/10.1093/chinesejil/jmi001)

* The 'extract' if it's a full text html article with no abstract (e.g.
http://dx.doi.org/10.1093/frebul/kti014)

* The registration/pdf page if the article has no abstract and is not in
full text HTML (e.g. http://dx.doi.org/10.1093/fs/kni063)  This matter
has been discussed with the authors, and I understand that they have
amendedtheir article, which has now been accepted for publication by
Journal of the American Society for Information Science and Technology
(JASIST)

If readers of this listserv have any further questions about Oxford
Journals and HighWire usage statistics and linking strategies, please
don't hesitate to contact me directly.

Richard Gedye,
Sales and Marketing Director
Oxford Journals
Oxford University Press
Great Clarendon Street
Oxford
OX2 6DP
richard.gedye@oupjournals.org