Dear all,

In November I sent out a somewhat lengthy questionnaire addressed to UK
HE institutions who use EPrints for their institutional repository.  I
had seven responses, which have provided us with some valuable pointers
for formulating our own policies and practices.  We are extremely
grateful for the time taken by respondents to help us in this matter.  I
provide a summary of the responses below:

What terminology do you use to refer to your institutional repository?
Three institutions use some sort of acronyms, and overall there is a
variety of terminology: repository; research repository; EPrints;
electronic collection; research online; central archive. 

Who is in overall charge of the IR?/Is this a dedicated post or extra
duty?
A wide variety of set-ups.  Three institutions have a dedicated post:
one has a background in bibliographic database design rather than
librarianship; one is funded by the Research School rather than the
Library, on a fixed-term basis; one is based in the Planning Support
Office.  Where not a dedicated post, post-holders have titles such as
E-Resources Manager; Research Support Specialist; Technical Development
Manager.  Report is variously to Technical Services Manager; Head of
Content Development and Head of Learning and Research Support;
Librarian; Deputy Director (Academic Services).

What grade of staff do the day-to-day work?
Reported grades vary widely, but may not be equivalent between
institutions.  Work is done by one person alone; by one person with one
assistant; by teams of up to 10 editors.  Editors may be Assistant
Librarian equivalent, or a mixture of professional, administrative and
trainee.  One respondent says that aptitude is more important than
grade.

Average number of items input
Ranges from 2 or 3 a week to anticipated figure of c.2000 per year.  

Time commitment
In terms of staff, two institutions quote nearly 1 FTE per week and at
least 1.5 FTE per week.  In terms if individual items, 10-20 minutes for
adding and editing, with checking, clearing copyright, communicating
with academics etc. taking up an additional variable length of time.  

Who inputs/upgrades the records?
The general model is for academic staff or their support staff to do
the basic inputting, and repository staff to do checks and upgrade the
records before making them live.  However, most respondents say that the
repository team will do all the work on request, and one states that in
practice library staff do most of the inputting.  Two institutions also
harvest records from sources such as PubMed.

Does the repository team provide training for academic and support
staff?
All provide some degree of training, including formal hands-on
sessions, practical demonstrations, drop-in sessions, detailed training
documents and FAQs, and email/telephone helplines.  One respondent says
that the process is pretty self-explanatory anyway.

Is there a University mandate to deposit material in the IR?
Four institutions report that deposit is still voluntary, with a
general move towards a mandate.  The one institution with a full mandate
comments that there is no adverse consequence for non-compliance.  One
has a mandate to deposit citation only.  One has a ‘requirement’,
but is waiting for the repository to be running effectively before this
can become a mandate as a formal high-level statement.

How does the repository team approach departments to supply their
material?
All are proactive to some degree.  Emails are used by most respondents,
either to departments or to individuals.  Most spend time on advocacy,
via presentations at departmental meetings, one-to-one and group
sessions, meetings with School Head of Research, distributing flyers
etc.  One institution is working on a faculty basis, and comments that
it becomes easier to persuade others once a number of faculties are on
board.

What categories of material are included/excluded?
Most say that they would
 accept all categories of material, but all are
restricting content to research output, and do not include learning
materials.  The definition of research varies, and one institution
comments on how difficult it is to know what to accept: they place an
emphasis on high quality research, and therefore used to accept
peer-reviewed research only, but HEFCE now recognises ‘grey
material’ in the recent REF consultation.  Where theses are
mentioned, most respondents include only higher degrees, and one
excludes Masters.  Of the two who mention working papers specifically,
one does include them, while one only includes them where they are
equivalent in quality to published outputs.  One institution anticipates
an increase in non-print-based outputs once they begin to work with the
Art Faculty.

Do you include pre-print/post-print/published versions of the
documents?
Most respondents specify post-print and published versions only.  One
comments that pre-prints may never be accepted for publication, and
would therefore not be a good advertisement for the university.  One
institution is particularly keen to include copies of ephemeral items
such as conference papers, working papers (important in fields such as
economics) and publications from faculties.

Do you include both peer-reviewed and non-peer-reviewed documents?
All respondents accept both, but point out that not all categories of
material would go through the peer-review process anyway.  If an item is
not peer-reviewed, two institutions emphasise that it should represent
equivalent research quality.  

Do you accept only full-text documents, or is a link sufficient?
Two institutions prefer full-text, but all accept metadata only, with
link provided where possible.

Do you include material published before the author was at the
University or after s/he has left?
Two institutions include both; one includes neither; four would add a
researcher’s previous output, of whom two would add later output if
the author was a retired alumnus (not happened yet) or if the
affiliation for the output was for the university. 

Under what circumstances would you delete documents from the IR?
This is not generally considered to be a problem.  Potential
circumstances include copyright infringement; offensive content;
plagiarism; falsified research; national security.  One respondent also
mentions withdrawing an item if it is replaced by a later version.

How do you deal with copyright issues?
Most mention checking SHERPA RoMEO database for publishers’ policies
when dealing with journal articles.  Book chapters and conference papers
are more complex to check, and non-print-based outputs are a particular
challenge.  One institution is planning to talk to the PRS about music
items.

Whose responsibility is it to ensure copyright is not infringed, and do
repository staff check anyway?
Three respondents remark that it is officially the author’s
responsibility, but in all cases the repository team take on the
responsibility for checking published material in practice. 

How do you deal with co-authored works?
There is less of a problem if co-authors are at the same institution
(and co-authors can be picked up during the review/verification stage
before going live, so the item will also automatically appear in their
area of the repository), but one respondent pointed out that if there
are many co-authors the item should represent a significant research
output for the institution.  Where co-authors are outside the
institution, one university asks the local author to gain permission
from co-authors before the item is accepted, whereas another leaves it
to the author to decide.

How do you deal with embargoed material?
All use the EPrints embargo feature to restrict until the embargo
expires.

How do you deal with sensitive material to which access is to be
restricted?
One institution advises not to deposit; one embargoes indefinitely,
e.g. for theses; two mention the different levels of access provided by
EPrints.


What file formats do you accept?/Do you convert any to a different
format?
All accept any format.  Two convert everything to PDF, but store the
source files in the background for preservation reasons.  Four mention
specifically converting Word to PDF: one seeks permission from the
author to do this, and uploads as Word if permission is not granted.   
Another mentions converting ZIP files to PDF. 

What means of subject access do you provide?/Who provides the different
types?
Six use uncontrolled keywords; four use LC headings (one specifically
does not use LC headings, and one who does use them says they are not
successful and plans to remove them); one uses broad Dewey; three
mention departments/schools; one uses the RAE headings as ‘Research
areas’; one recommends the use of controlled vocabularies for arts
subjects, but does not police it; one would like to add controlled
vocabulary keywords, but too much work retrospectively.  Uncontrolled
keywords are generally added by the academics or taken from published
documents; LC headings are usually assigned by the repository team (by
the academics in only one case); the institution that uses Dewey makes
it a mandatory element for the academics, but monitors for consistency.
  


Sue Ashby
Deputy Technical Services Manager (Books)
University of Portsmouth Library
Cambridge Road
Portsmouth
PO1 2ST