So, do you have any? I'll tell you what I know:

On average, users rarely go beyond the second page of any retrieved set, 
whether in a library catalog or on Google. This of course wreaks hell 
with the FRBR concepts of "identify" and "select" which is supposed to 
be on the results of a "find."

In 1982, when we brought up the first U of Calif union catalog, the 
technology was telnet: 18 lines, 70 characters per line. At most we got 
two items, sometimes only one, on a screen. Statistically, the average 
viewing was 2.5 screens.

Later, maybe 1986? 88? we moved to a web interface. We could now get 10 
items on a screen. Statistically, the average viewing was 2.5 screens.

I've read, but unfortunately right now cannot find, that you never want 
to drop below screen 2 of google results or you're essentially invisible 
(in the SEO literature).

So, given this, and given that in a decent-sized catalog users regularly 
retrieve hundreds or thousands of items, what is the best way to help 
them "grok" that set given that the number of records is too large for 
the user to look at them one-by-one to make a decision? Can the fact 
that the data is in a database help users get a "feel" for what they 
have retrieved without having to look at every record? What is the net 
result of the fact that users don't go generally beyond screen 2? (In 
the U of Calif catalog, it meant that no one looked beyond items  with a 
author whose name sorted in the A's.)

kc


On 9/20/12 10:26 AM, Sean Hannan wrote:
> That's what user research is.  Qualitative evidence, too.
>
> -Sean
>
>
> On 9/20/12 1:18 PM, "Karen Coyle" <[log in to unmask]> wrote:
>
>> And I presume that you have quantitative evidence to show that.
>>
>> kc
>>
>> On 9/20/12 8:49 AM, Sean Hannan wrote:
>>> Every one of these sites is not going to work for everyone.
>>>
>>> Please conduct your own user research for your own audience.
>>>
>>> Our users, for example, have no interest in visualizations of search
>>> results.
>>>
>>> Our researchers actually want just a list of results. They are compiling
>>> bibliographies or reading lists and they honestly just want a really long
>>> page of titles and authors of what we have.
>>>
>>> -Sean
>>>
>>> On 9/20/12 11:03 AM, "Karen Coyle" <[log in to unmask]> wrote:
>>>
>>>> Every one of this suggestions has one major flaw, IMO. The primary
>>>> result of a search is a big set of bibliographic records -- more than
>>>> the user can possible look through. In some of them there are facets
>>>> available, but in no case is there any useful analysis of set in a
>>>> visualization that would allow the user to get a picture of what she has
>>>> retrieved. I'm thinking timelines, a la' WorldCat Identities or the Open
>>>> Library subject pages [1]. Also, none of them tell the user more about
>>>> the person or subject or work that they have retrieved. (At least, in
>>>> the views that I have seen.) I really think that lists of manifestations
>>>> just aren't good enough when searches bring up hundreds of results.
>>>>
>>>> kc
>>>> [1] some examples:
>>>> http://openlibrary.org/subjects/halley%27s_comet
>>>> http://openlibrary.org/subjects/place:istanbul_%28turkey%29
>>>> and see others at: http://openlibrary.org/subjects
>>>> or look for your favorites
>>>>
>>>>
>>>> On 9/20/12 6:03 AM, Hamilton, Gill wrote:
>>>>> My current fav is Digital NZ
>>>>> http://www.digitalnz.org/
>>>>>
>>>>> Gill
>>>>> ----------------------------------
>>>>> Gill Hamilton
>>>>> Digital Access Manager
>>>>> National Library of Scotland
>>>>> Edinburgh, Scotland
>>>>> [log in to unmask]
>>>>>
>>>>>
>>>>> -----Original Message-----
>>>>> From: Code for Libraries [mailto:[log in to unmask]] On Behalf Of
>>>>> Tania Fersenheim
>>>>> Sent: 19 September 2012 20:00
>>>>> To: [log in to unmask]
>>>>> Subject: [CODE4LIB] Seeking examples of outstanding discovery layers
>>>>>
>>>>> Got a favorite discovery interface?  Send me the URL
>>>>>
>>>>> I am doing some quick & dirty investigation into libraries that have
>>>>> successfully and elegantly integrated discovery of various resources,
>>>>> e.g.:
>>>>>
>>>>>     - library catalog
>>>>>     - federated indexing service such as  Serials Solutions or Primo
>>>>> Central, or a federated search system like Metalib
>>>>>     - ejournals
>>>>>     - ebooks
>>>>>     - libguides
>>>>>     - library web site
>>>>>     - worldcat local
>>>>>     - that kind o' stuff
>>>>>
>>>>> I am looking for sites that are both nice to look at and seem easy to
>>>>> use.  I will assume that if you're touting your own site it is
>>>>> technologically sophisticated.  :-D  Got any faves?
>>>>>
>>>>> Tania
>>>>>

-- 
Karen Coyle
[log in to unmask] http://kcoyle.net
ph: 1-510-540-7596
m: 1-510-435-8234
skype: kcoylenet