Speakers: Lisa Shen, Glenda Griffin, Erin Cassidy, and Tyler Manolovitz
They did a 16 week pilot program, and in that time, users selected about 640 titles, selecting them steadily throughout the program. The most expensive titles tended to be reference works and STM titles. The least expensive were humanities driven or public domain. STM and social sciences took up almost half of the purchases. Surprisingly, more literature titles were purchased proportionate to the number available.
They used the YBP content level to compare the patron purchases with the librarian selections. The users were still selecting academic content at a high level, although the librarian selections contained fewer popular titles. They found an overlap between the two in general and advanced academic levels, and interestingly, users selected much more supplementary material than the librarians.
Shortcomings: not all librarians participated in selecting from the thousands of titles, the duplicates were not removed (33% of the titles were already owned in print or from another ebook vendor), and the default catalog display ordered items by publication date (puts the ebook first).
In summary, PDA is a good supplement to but not replacement for traditional methods, and may be an indication of emerging research needs.
Suggestions: Set a title price cap. Consider excluding older materials, journals, duplicates, and titles from publishers with better bundle deals. Use modified triggers like 10 pages viewed, 10min of usage, or anything copied or printed.