Here is what our main screen looked like for the October 2002 user testing sessions.  It still replicated the “Browse/Search Resource” screen of the ENCompass 2.0 interface quite closely, though we did choose to remove the direct searching connections and integrate article level searching into the interface later.  From the page you see, users could only search and browse metadata about resources, as was the case with the e-Reference interface.
	The 2.0 design included a shaded table format with expandable/collapsible categories and checkboxes and presumed that the user knew to type a query term in the search box and then select one or more categories and/or specific resources below it to run a search.  We found that users were confused by this design, for several reasons.
	Browsing and searching capabilities were presented in a way that was not intuitive.  It took users some time to figure out how to navigate the series of checkboxes beneath the search box.  This was compounded by the expanding/collapsing checkbox categories design; expanding a category meant reloading the screen with more subcategories showing and more shaded rows in the table to scroll through.  Users were disoriented as they clicked deeper and deeper into the same screen.
	Further, some users did not understand that they could select and search for more than one resource at a time.  While cross-collection searching was a familiar concept to our users in other contexts, it was new to the e-Reference collection and required clearer presentation.
	As had been the case in the past, users still attempted to perform article level searching, despite the explanatory text at the top of the page.  The e-Reference interface had included similar text, but clearly, many users were not reading it.
	Finally, the name “e-Reference” appeared to hold less meaning than we anticipated for most of our users.  In this early prototype, you can see that we retained the “e-Reference collection” name at the top.  We wanted to preserve the identity of this heavily used system and had assumed that keeping the same name was important to our users.  While it is certainly true that experienced users identified with the name and that many librarians favored keeping it, our sample indicated that the name was not particularly meaningful to new users.  And, as we looked at adding the article level searching functionality, the name became problematic.