JISC Answers

Hi James,

If I were you, I'd define an IR for yourselves, for the purposes of your project. Each institution expects a different thing from their IR and so each IR is very different in flavour. Just look up a few institutions on ROAR.

Is it truly an IR if it can't be cross-searched? What about a publications database that isn't on the Web at all?

Each institution probably has a "repository landscape". That might consist of just one repository to do everything (in which case what is "everything" to that institution?), or it might be that they have a repository for full text journal articles & theses, and a publications database for metadata-only records and they have a repository for a specialist collection and one for a department - which is what we've got at Warwick.

We don't even consider ourselves to have a CRIS - yet, but that could muddy waters even further. You could also say that there is a research metadata landscape at each institution, into which the repository/ies fit...

Best of luck!

Jenny

Sadly I can't help you about the specifics of institutional repository projects, but like you I have found the idea of shared understanding to be one of the vital parts of project management. After some difficulties (partly of my own making) in a recent project we used an icebreaker session to develop our own shared definition. Mind you, using an external definition can be useful sometimes when a shared understanding is not possible.  Saying that it's JISC's official version will usually stifle opposition  :-)  I'm trying to find specific advice about this in the InfoNet toolkits as I type this. So not an answer but some sympathy.  .  Paul M   (Teesside University)

Ambiguity is the death of many a project? Alternatively it is the freedom for installations to make their own policy-based decisions. An IR may (or may not) deal with research objects and learning objects and many other kinds of things.

 

Thanks everyone for the responses - much appreciated!

@jenny - I really like the idea of describing the institutional 'repository landscape' as opposed to the IR in general. Probably the mistake we made in drafting the project was in thinking about the IR as a single entity, which of course it isn't. 

Actually the problem for the project has arisen out of the fact that we did try to define the IR for ourselves, only to be told by the 20 other participating HEI that they simply can't agree on the definition as it is to ambiquous a term (hence my posting here).

As our definition was there to indicate the scope of activity, we need to control the extents of scope in a different way. A question like 'what is the preservation policy for your institutional repository' could be better served by using 'how do you manage preservation policy within your repository landscape'

@les - your comments are very interesting and probably worthy of a seperate debate altogether! Freedom to make choices is a great thing, I agree there, but surely its a fine line between freedom for installations and a need for some level of institutional joined up thinking. There is a need I believe to be in some way consistent with institutional approaches to repository use and policy management etc

Thanks again for comments everyone - this is very useful for me

It is a really interesting question.  Clifford Lynch's famous definition of a repository being a set of services is something I don't feel that comfortable with myself. Practically, I think that Jenny is right in saying above that it's pretty much what you make it. The question has been phrased here in terms of content, but it is also a question of its structure, functionality -- and services.

Having said that, if you do need to define an institutional repository, then the way that repositories have grown have typically meant that more and more functionality is being asked of them and because of that, the definition has tended to expand and become woolly. In the end I am quite keen to keep the association with open access, because all of the other things -- research management systems, catalogues, bibliographic databases, publication databases, virtual learning environments, databases, archives, whose functionality is now being added into a repository all have their own distinct names and functions. For me, I think that to be a repository it has to be based on maximising its full-text holdings of research material and making them openly accessible.

For me, that is the core and a working definition: it can have other features, services, etc but those are the things which are built on, rather than being the repository at heart.

Bill