Usage Statistics for Yales Digital Collections

1
Usage Statistics for Yales Digital Collections

• Jennifer Weintraub
• Digital Collections Specialist, Yale University
  Library
• 5/17/02

2
Digital Collections at Yale

• 400 databases purchased from Vendors
• 12 databases purchased from Vendors but loaded
  locally
• 20 digital collections created here at Yale, many
  from material held by Yale

3
Usage Statistics

• Keep track of the databases of about 30 vendors
• Keep track of all 12 locally loaded databases
• Usage Statistics updated on one central web page,
  in various forms and updated at different times

4
Example of Digital Collections

• Locally loaded full-text databases
• Curiae Supreme Court Hearings
• Visual Resources Collection
• Beinecke Photonegatives Database

5
Full Text Databases

• 12 databases, mostly literature collections
• Encoded text from Chadwyck Healey/UMichigan
• Analog software open source.
• Statistics for Full Text Databases
• Can also get referrer logs

6
(No Transcript)
7
(No Transcript)
8
Curiae Project

• Nedstat page for Curiae
• Where are people coming from?
• Any usage patterns?

9
Visual Resources Collection
10
Problems with Collecting Statistics

• Is it really important to know usage or is
  professor satisfaction enough?
• Granularity
• Luna Imaging to provide a statistics package in
  next release
• Would like to know times of highest use, if
  students can get to it successfully from home, if
  students use the system

11
Beinecke Images Database
12
Why no Statistics?

• Goal is different than most digitization projects
• Statistics not important to the success of this
  project
• Several projects like this at Yale

13
Bottom Line

• Decentralized collections and systems
• Digital Collections are user-based
• Not of particular importance to the librarians in
  charge anecdotal evidence is enough.
• Getting the right statistics to tell librarians
  what they need to know.

14
Differences in Statistics Collection

• Vendor Databases
• Vendors choose the statistics
• Usually the work is done numbers of searches,
  sessions, pages viewed is clear and labeled
• Different statistics for each vendor little
  uniformity

• Home-Grown
• Not as sophisticated as vendor systems
• Can manipulate software to get statistics to suit
  the collection
• Little uniformity, but that might be a good thing

15
Statistics We Do (Should) Collect

• ARL
• Questions about building the collections
• Emetrics
• Asked for usage data that was hard to give
• Digitization project guidelines frequently ignore
  the subject of usage statistics.

16
What Other Libraries do

• IIT Wright Air Development Center Digital
  Collection Statistics
• University of Michigan
• Some collections develop their own statistics
  programs

17
In the Future

• Convey the importance and usefulness of usage
  statistics for people creating digital
  collections.
• Collect number of searches for all databases.
• Look at collecting the number of images viewed
  for image databases
• Look at collecting page views for full-text
  databases
• Look into other methods of determining quality of
  digital collections, besides usage statistics.

18
Links

• Analog http//www.analog.cx
• Nedstat http//www.nedstat.com
• Luna Imaging, maker of Insight
  http//www.luna-imaging.com/insight.html
• Curiae http//curiae.law.yale.edu
• Beinecke Images http//highway49.library.yale.edu
  /photonegatives/

19
Jennifer Weintraub

• Jennifer.weintraub_at_yale.edu