New Ways to Think about Library Statistics

The Chronicle of Higher Education blog has a really interesting post proposing that libraries begin to make their user data (in anonymized form) available to researchers. In a post titled, "A Good, Dumb Way to Learn from Libraries" (really?! geez!), David Weinberger proposes that libraries

1. Create Stackscores for their materials
These get around the current issues of

A. Privacy/Patron Security by offering an annual computation and perhaps blurring that numbers if it looks as though even that might be hackable down to individual patron.

B. Interoperability - that is making the numbers from various libraries comparable regardless of what automated systems the libraries use, or even how they calculate their stackscores.

He offers an example of what the Harvard Library Innovation Hub created as StackLife. Weinberger, who used to develop the Innovation Hub and thus (I presume) helped develop the StackLife app, explains that the darker blue a title bar appears, the more heavily used it has been. He cautions that depending too much on this can create a feedback loop where the more popular items just get more popular, and lesser known but excellent materials remain overlooked. (Librarians know all about this)

Interesting thoughts. I don't know if he really addresses all librarian issues, but it's a thought-provoking post!

The thinking cap decoration is from another thought-provoking blog page, How would we think without language? http://www.eurolondon.com/blog/en/how-would-we-think-without-language/ Interesting to consider how to build search engines!