Michael seems to have a very good start at a list of useful data. 

Our general interests for data collection are to inform: 
a) our collection building decisions for image access; 
b) our contract negotiations in providing access; 
c) our technical provisions for providing access; 
d) our provision of services that will effectively assist diverse users in finding what they want.
 
A background question we have about user logging is: how is RLG defining a "session" in the web environment?

Some useful session-level data would be:
 - Number of search statements per session (to give us some information about the complexity of searches and amount of time spent at the computers)
 - Time of day of searches (to give us information about peak usage times)
 - ip address of originating machine (to give information about who is doing the most searching, and where)
 
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~   
Dana Beth                          
Art & Architecture Librarian       Washington University
			           Campus box 1061
dana-beth@library.wustl.edu	   St. Louis, MO 63130-4899
1 (314) 935-5218		   USA
FAX: 1 (314) 935-4362
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~