I've been Googling "learn* analyt*" and "librar*" from time to time for about a year now. A year ago, the only things that I could find were reports, not written by librarians, that mentioned academic library data as possible fodder for learning analytics and, of course, instances wherein "library" was used in a mere data sort of way, i.e. "data library." That's starting to change. Today I found:
The Library Impact Data Project, a JISC-funded project in the UK in which systems librarians at the Univ. of Huddersfield seek to describe the relationship between library usage and student attainment; identify courses that require a high level of library usage and describe the practices that lead to this; and identify courses that do NOT require library usage and improve services to those courses. I'm definitely going to look into this one more: http://library.hud.ac.uk/blogs/projects/lidp/
Stephen Abram's blog post from Jan. 3, 2012 that points to EDUCAUSE's "7 things about analytics" article, but without any analysis of it: http://stephenslighthouse.com/2012/01/03/learning-analytics/comment-page-1/#comment-19972
In 2010, this was published: Kumar,
S., Edwards, M. & Ochoa, M. (2010). Analysis of Online Students’ Use of
Embedded Library Instruction in a Graduate Educational Technology Course. which used data logs from the LMS in which library instruction was embedded to examine student usage of library resources. Far beyond the usual self-reported "preferences/how did you like it?" type of research that we often see in library-related publications.
The Sept. 2011 issues of College and Research Libraries had an article by Wong and Cmor describing a study that linked GPA to library workshop attendance with n=8,000+ students. The librarians had access to the school's ERP to do this; interestingly, the study was done at a Hong Kong university and there is little mention of any consent from the students involved, although student names were unlinked from GPA, and also no mention of an IRB-type approval process in the way we would think of it in North America. The article is at http://crl.acrl.org/content/72/5/464.full.pdf+html I imagine the only similarity between this study and true LA is the link between the library and the ERP -- a link which was enacted outside the ERP, at any rate.
The Georgia Perimeter College Libraries' GPC blog by Rebecca Rose explains,
A library application for learning analytics could be a report alerting the library to order more materials in specific subject areas, based on enrollments in specific courses. Another idea would be to get lists of students who have never experienced library instruction. http://gpclibraries.wordpress.com/2011/04/21/technology-trends-in-libraries-pt-3/
And, finally, lots of library-related blog posts referencing the 2011 Horizon Report which listed LA as a technology with a 4-5 year time horizon. Right now, these were mostly informational posts, and did not offer any visions or analysis for LA applied to libraries.
So, overall, it appears that tech-focused librarians are becoming aware of LA, but outside of a couple of projects, libraries are not involved with LA very much at all. If I'm wrong, please leave a comment and give me a clue.