[Eril-l] LexisNexis OpenURL failures

[Mandi Schwarz]

Hello colleagues,

(Apologies for cross-posting.)

We have noticed a distressing number of OpenURL failures for LexisNexis content. Using our Summon instance, I located 100 articles (each in a different, mostly-randomly-chosen publication [1]) to test; 34 of them failed to get from our link resolver (360 Link) to the full text in LexisNexis. There are two errors that we encounter: "No Documents Found: You can edit your search and try again." and "Error: The document section name you entered is invalid. Please check for spelling errors or duplicate information." Comparing the citation in Summon with the article in LexisNexis, I often see no glaring errors. Serials Solutions maintains that the disconnect occurs with LexisNexis' unique indexing; to date, LexisNexis has been uninterested in rectifying this.

A 34% failure rate also indicates the inverse: a 66% success rate. Our librarians will be reviewing this to see if this meets the minimum success threshold; if not, we will discuss what we can do to minimize the impact on our users, which may include removing LexisNexis article-level results from Summon.

Our question to you is: Have you noticed similar failures at your institution? If so, have you changed how you present LexisNexis content to your users?

Thank you,
Mandi

[1] Using an updated title list from LexisNexis, I would randomly scroll through the list with my eyes closed and click on a cell. I would check that publication for language (I searched only English publications, because I was using an English search term) and format; I discovered that may publication types are not fully indexed in Summon, including: Boards, Panels & Tribunals; Cases; Decisions; Legislation, Statutes & Codes; and, Web-based Publications. As such, I chose publications listed as: Law Reviews & Journals; Magazines & Journals; or any type of 'News' item. I would then combine a publication title search with a truncated keyword, such as 'increas*', and work down the list until I found a title that returned at least one article for testing, which I would also choose by blind scrolling. Then I'd scroll the spreadsheet for the next title and repeat the process. While this method isn't scientifically random, it met the needs for this test.

Mandi Schwarz
Library Assistant - Electronic Resources
University of Northern British Columbia
250-960-6455; Mandi.Schwarz at unbc.ca<mailto:Mandi.Schwarz at unbc.ca>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.eril-l.org/pipermail/eril-l-eril-l.org/attachments/20160425/e8ca90c1/attachment.html>