I asked for opinions on how search behavior has changed over the
years and whether or not people have become better searchers.
Overall, it seems that there's still a long way to go.
Thanks for your responses.
[An aside: whenever I make up one of these summary messages, I notice
that most people start their messages with something like "Hi Hal". I
never do that. I wonder if it has to do with when one started using
email, in what environment, etc. Or Maybe I'm not as friendly as the others...]
-- hs
=====
From: "McLaughlin, Gail G" <[log in to unmask]>
Hi Hal,
As a Web user since 1994 when all pages were gray with black text,
and maybe a small square picture for visual relief, I have noticed my
that my own search arguments are better formed to achieve optimal results
I tutor people at my local senior center on searching, among other
things. They often have a hard time understanding what a search
engine is, how it works, how to form a search argument, what results
to expect, and how to choose the result that best provides the
information they are looking for.
People have become better searchers through practice, trial and error
and learning from others. In my experience, a new searchers tend to
enter one word, form a question or string together a bunch of words
(sometimes without spaces). They don't give much thought to the
hierarchy of the words initially. I teach them to present their
search argument as a hierarchy of words, to use Google, and click on
"cache" in a promising result. So, if someone is looking for a good
recipe for chicken tortilla soup, I have them enter "recipe soup
chicken tortilla".
=====
From: "Simon Forrest" <[log in to unmask]>
Hal,
A couple of observations, with no claim for general applicability, from a
recent test session:
*The site being tested (a business information site) had a standard
search box on each page with an option for advanced search. The
advanced search had some checkboxes and select lists to filter the
results. Fairly standard stuff. What surprised us is that every
person who went to the advanced search page and picked one or more
options from the filters expected it to increase the scope of the
search, not restrict the scope of the search. Hence in each case the
person was surprised that after they asked it to bring back more
content (in their view of the way the advanced search worked), they
actually got back less content than before. From their use of search
throughout the test, these people appeared to be less sophisticated
searchers - the more sophisticated searchers ignored the advanced
search and simply rephrased their query if they didn't get the
answers they expected.
* One person, when presented with the test script setting a task,
typed the entire paragraph from the test script into the search box.
When this returned no search results, this person also tried the
advanced search filters to broaden rather than narrow the scope of the search.
The second example is no doubt rather extreme and unusual. However,
we were genuinely surprised that a) everyone who used the advanced
search misunderstood the way it worked, and b) those with the less
successful search strategies were the ones who turned to advanced search.
Among other changes we made as a result of this session was to improve the
general search capability (eg adding a best bets feature) and to remove the
advanced search as it appeared to be consistently confusing and problematic.
Simon Forrest
[log in to unmask]
=====
From: Hagan Rivers <[log in to unmask]>
Hal --
Jared specifically studied this in a series of usability tests he did
a few years ago. I don't know if the data still holds, but I can't
imagine why it would have changed. Might be worth a phone call or
email to him or Christine Perfetti (on his staff) to find out.
Anyway, here's the article:
http://www.uie.com/articles/learn_to_search/
- Hagan
=====
From: Jeff Johnson <[log in to unmask]>
Hal,
The base of Internet users is still expanding rapidly
(http://www.pewinternet.org). The number of Americans who have
Internet access at home is still rising, as is the number who are
switching from dial-up to broadband.
Anecdotally, I can think of several relatives and friends who are
just starting to use the Internet now.
Thus, at any point in time, a large number of people are just
beginning to use search engines.
So assuming that Internet users are becoming more sophisticated
searchers is like saying teenagers are becoming more sophisticated
readers: there aremillions of new ones every week.
Jeff Johnson
UI Wizards, Inc.
Product Usability Consulting
www.uiwizards.com
=====
From: "Claudia Alden Case" <[log in to unmask]>
Hal,
What the person might be talking about is Ask Jeeves, which I know
was the first search engine my kids used because that was what the
teachers (about 6 years ago) were teaching kids to use at
school. Ask Jeeves, which is still around as Ask.com
(http://www.ask.com/) allows users to type in an question and the
results page will be sites that should have answers to the question.
It was very useful for people who weren't familiar with the idea of
search engines.
There are still some Help functions around that tell users to enter a
question. (Clippie, the hateful Office desktop assistant from
Microsoft, allows that format.) I would say that for the bulk of
basic computer users, it might still be a viable interface
solution. My definition of basic computer user is someone who users
the internet only to check mail and maybe visit a website if they
know the URL (but they would likely not understand the term
URL...instead they would call it a website address). This class
includes a large number of users who did not grow up with computers,
especially PCs. For that class of users, I would say "google" is not
yet in their standard vocabulary and smart searching is not one of
their standard skills.
That's my two cents.
c
=====
From: "Caroline Jarrett" <[log in to unmask]>
Hi Hal
I've done quite a bit of search analysis for a major client over the
last couple of years.
Consistently, one-word searches are far more common than 'proper
sentences'. In around 20,000 search entries in a month, there were
only about a dozen properly formed questions.
On the other hand, I've also done a lot of usability testing over the
same time for the same client - so observed people searching as well
as looking at the data. And what I've seen is that consistently,
users try the shortest search they can think of, usually one or two
words. But if that doesn't work, they often then try a longer search
using the same words in the mixture.
What I have definitely observed is that people arrive at the
usabiilty testing generally a lot more adept and confident on the
web. And because search is such a key skill, I'd say that they're
better at searching too.
Best,
Caroline Jarrett
[log in to unmask]
01525 370379
=====
From: Avi Rappoport <[log in to unmask]>
My goodness, that's an interesting definition of "better". In the
site and intranet search logs I've examined people type 1 to 3 words
in a query, and the length has been getting only fractionally longer
over the last ten years.
In my experience, they never did type long natural-language queries,
though librarians and IR people wished they would, to give more
context to the query. The examples they'd use would be "size of
George Washington's false teeth" "What George Washington's were made
of", or "what happened to George Washington's teeth?" All provide
much more for an engine to work on. Sadly, people simply don't do
that, and never have. AskJeeves made us think that people had,
because they offered a list of possibly-related questions to choose,
and then implied that those were the questions typed.
Leaving out articles and conjunctions, only makes search *better* if
you are using a system that matches every query term (Boolean OR,
recall-heavy) and doesn't do stopwords by default. Google and most
other large systems limit document matches to those which contain all
the query terms (Boolean AND, more precise). So the issue of
articles is less important.
I have found users reformulating queries, fixing typing and spelling
mistakes, and otherwise attempting to improve their searches, but I'm
not sure I see specific improvements over time.
Hope that helps,
Avi Rappoport
Enterprise Search Analyst
=====
From: "Claus Zimmermann" <[log in to unmask]>
Its hard to say,
Perhaps its a true statement, and a learned skill. Do a search, see
the results and then start getting a feel how to ask questions.
Socially we learn to with seeing how others search intelligently and
productively and then we try to mimic that.
Good products also help, like google, that prompt and try to give
good quality results. Its difficult when you ask a question and get
the whole of the internet returned to you like some of the older
search engines. Its probably a technological growth, as well as a
testament to usability itself.
Claus
=====
From: Skot Nelson <[log in to unmask]>
I'd argue that this doesn't mean people have become better searchers,
but rather that search has failed to evolve.
Isn't the goal (ultimately) to ensure that people don't have to lower
themselves to the intellectual capacity of computers?
=====
From: "Mary Deaton" <[log in to unmask]>
To: "Hal Shubin" <[log in to unmask]>
HAL,
The Pew Charitable Trust project on the Internet did a search study
about two years ago that looked in great detail at how people search.
You can find the report at
<http://www.pewinternet.org/PPF/r/146/report_display.asp>http://www.pewinternet.org/PPF/r/146/report_display.asp
They have a variety of other reports that might also help you explore
this issue.
--------------------------------------------------------------
Tip of the Day: Quote only what you need from earlier postings
CHI-WEB: www.sigchi.org/web POSTINGS: mailto:[log in to unmask]
MODERATORS: mailto:[log in to unmask]
SUBSCRIPTION CHANGES & FAQ: www.sigchi.org/web/faq.html
--------------------------------------------------------------
|