Legend: NOT STARTED IN PROGRESS STALLED DECIDED
Recommendation
(Harmonization undesirable. Each campus should make the decision based upon their collection and user needs).
Impact
Stakeholder group | Impact |
---|---|
Who does this decision affect? [name of the group] | Explain the significance of the decision to EACH stakeholder group. How will this decision impact this particular group? What will change? Do they need to take any action? If so, when? |
ILL |
|
eResources |
|
Reasoning
Background
Prior to the migration to a consortium, campuses were able to curate their CDI collections. While this required them to turn on access to individual collections, it also allowed them to weed out collections with poor metadata.
The UC Library system is now on EasyActive. “Alma collections that you have active because you subscribe to their full text will become automatically searchable in CDI. For those collections no separate search activations are necessary.”
PPC (Go Live) CDI Models: FullyFlexible or EasyActive
“It is also important to note that while CDI collections in the EasyActive model cannot be “de-searchified” (ie. hidden from users even in expanded results sets), administrators can “override” the full-text activation, effectively hiding these collections behind the “Expand” option (for campuses limiting by availability).”
Campuses are allowed to decide for themselves if they would have their default search limit CDI results by availability or not.
We are revisiting this issue in order to provide guidance and examples of the pros and cons of limiting by availability or not.
What is Filter by Availability?
CDI has tens of millions of records. By default, a search will surface all CDI results regardless of availability to the library.
When configuring a search profile that includes the Central Index search scope, libraries can opt to “Filter by Availability” so that the search profile will not surface results that do not have full text availability. So, the only CDI records that would display would be those with full text availability.
This setting causes the “Expand My Results” toggle to display. Then, if a user wants to include results for CDI records that do not contain full text, they can select the Expand My Results check box.
Options Considered
Default | Filter by Availability | |
---|---|---|
Description | A search will return all relevant CDI results regardless of availability. Users will see electronic article and ebook results that are available online as well as those they must submit a request for. | A search will only return CDI results with full text availability. Articles and ebooks that must be requested are filtered out. |
Pros |
|
|
Cons |
|
|
(Settings at each location and the reasonings behind making that choice at the time?)
UCR: The Articles, Books, and More profile is unfiltered. But we added the CDI scope to the UCR Library (IZ) search profile, filtered by availability, and excluded ebooks.
The idea was that we wanted the UCR Library profile to show users what they could theoretically get “right now” by walking to a shelf or clicking an available online link. The only items they would have to put any kind of request on would be print items that were checked out/lost/in transit. We excluded ebooks to reduce duplication because CDI and IZ holdings do not dedup.
UCSF: Our default search profile (‘Search articles, books, and more’) includes CDI collections, and we have Filter by Availability turned on
We don't have any undergrads and it’s usually more important to our audiences to see everything (rather than only what’s available now).
Dependencies
Questions to consider
How exactly does this impact ILL?
How would they test this?
Why in the world did I approach this decision page and testing backwards?
There was no good option with easy active vs fully flexible. There were pros and cons on both sides. Easy active was less work on the catalog side of things.
Consider that the whole point of this shared catalog thing was to get everything accessible by the UCs into one bucket for people to discover, even if it’s not currently on the shelf or subscribed to. An increased ILL Workload is definitely going to be a part of that.
What options do we have on the Discovery side to reduce the number of “bad” requests coming from inaccurate metadata records?
Action Log
Action/Point Person | Expected Completion Date | Notes | Status |
---|---|---|---|