[WEB4LIB] RE: AP study

[Dobbs, Aaron]

Classic quote, Karen.
"start[ing] from the assumption that the user is not broken...would hose
most ILS interfaces right off the bat."

Perhaps you were poking fun, perhaps you meant exactly what you said;
either way, I think you're right: ILS interfaces are hosed.  While I
don't have a magic bullet answer, most ILS interfaces could use major
overhauls (and not merely to the first page splat of the Google results
list format, either).

Maybe an easy to use search field (or two, or three max) with intuitive
Boolean built in, some "smart" algorithms behind the scenes to catch
obvious typos (based on db content) e.g. Google's "did you mean" link at
the top of the results list, and possible grouped results lists (i.e. if
a proper name is entered "items by [proper name]," "items about [proper
name]," with a sub-sort "by format(physical, electronic, webpage,
article, etc)."  When a term is entered, have that term compared to
"official" subject (LC, internal proprietary, whatever) headings and
offer results of those terms or links to those extra results as well.

Like I said, no magic bullets - but if a commercial search engine
(Yahoo, Google, AltaVista. &c) has had some of these features available
for at least two years why don't ILS have similar?  Yes, strides are
being made.  Yes, there is hope.  But will these changes be made before
it is too late and we have lost access to the up-and-coming user market?

Are there efforts underway, beyond the current, first and second
generation federated searhcing initiatives to build such?  
To me, this almost looks like a fresh start would be required with a
different set of basic asumptions: the user knows what s/he wants,
doesn't care where it comes from, and wants it "now."

-Aaron
:-)'

-----Original Message-----
From: web4lib at webjunction.org
[mailto:web4lib at webjunction.org] On Behalf Of K.G. Schneider
Sent: Thursday, October 28, 2004 12:47 PM
To: Multiple recipients of list
Subject: [WEB4LIB] RE: AP study


> The counter to that--as the study in this article shows--is that most 
> users don't make effective use of search engines. That's where they 
> are. So the answer is yes, we either train them or work to create 
> search engines that will lead users beyond the 2 words/first page of 
> results circuit.

Actually, in reading the article, I had two thoughts. 

1. I'd like to see the data.

2. The article didn't say that users weren't finding what they were
looking for. There was a leap from user behavior to conclusions about
user results; I'm unclear whether that is the article's author or the
book itself. The article said users used one or two terms and stopped at
the first page. I do that quite often myself, because I know I will find
what I'm looking for. Did the book conclude that users aren't finding
what they are looking for? 

I've had a lot of novice users direct me through Google searches that
might not have been the way I was taught to search but work just fine
for them, and however Rube Goldberg their devices, they get what they
are looking for in ways that satisfy them.

We spend a lot of time talking about users who can't find what they are
looking for, and yet, users as a tribe are pretty smart. They gravitate
toward the easiest tools, they figure out how to get results that
satisfy their immediate needs, they know to ignore the "hundreds of
thousands of results" and focus on what they are looking for. (I had
novice users advise me of the site based on the Ashlee Simpson SNL
debacle just hours after it went up.) 

I'd be curious to see an interface design approach that started from the
assumption that the user is not broken and that he or she basically
groks the Internet. but then, that would hose most ILS interfaces right
off the bat.

Karen G. Schneider
kgs at bluehighways.com