Please understand that I love FirstSearch. So many databases, so many full-text articles, and all available on a per-search basis rather than at a fixed rate which may never pay for itself. Blessings on OCLC for meeting our need almost before we realized we had it!

I just really wish they hadn't called it FirstSearch. You see, in my book, for most purposes it's Second Search, a system designed for professional researchers who know how to search, who can figure out how any search system operates, and who will read the help screens.

Because of its name, students assume it is THE place to start looking for anything. The trouble is, for the casual first-time user, so much about the search system is far from obvious. For one thing, the design of the opening screen strongly suggests to people that the only available options are WorldCat, Article First, and ECO--none of which are explained. The students have no idea what they're clicking on.

These are all fine databases, but none of them will deliver full-text articles to students on the spot, which is what most of them really want to end up with. The ALL DATABASES option is off to the side on the navigation bar where most users don't see it. (What I have observed is that users do not really look at the full screen; they focus only on the middle.) Nor does OCLC explain up front what kind of materials users will be searching through when they click on any given database.

Having chosen a database, more or less at random, most users type their subject in the box on FirstSearch's initial search screen, which does not permit a complicated search statement. Unless they are trained to, very few of them, will click on the "advanced search" option. That's a pity, because the advanced search template, with its multiple search boxes and logical operators, is so much better at showing naive users how to construct a boolean search that THIS should really be the initial search interface.

Often the Advanced Search template will produce perfectly adequate results. But compared to a search system like Silver Platter or InfoTrac, it is still inferior, because it will not allow complete truncation, and it only allows two "or" statements in one search.

Consider a search topic our Occupational Therapy students get that can be absolutely hellish to search: for a given ethnic group, discuss how that group's cultural background affects its health and health-seeking behaviors.

On our Silver Platter version of CINAHL, I do that statement this way:

Set 1: Native-american* or Indian*
Set 2: health or disease* or diet or food* or nutrition or alcohol* or drug*
Set 3: #1 and #2

In FirstSearch's version of CINAHL, the best I could do in one search with a maximum of two "or" statements would be something like this:

set 1: native American+
AND
set 2: health or disease+ or diet

I would feel better about this if OCLC seemed aware that their search system could stand improvement. But when I have mentioned it in OCLC workshops, their trainers have insisted that "nobody wants to combine sets" or "nobody needs complete truncation." I expect vendors to imitate the best features of their competitors.

Even when my search topic would logically lead me to BIOSIS, which is available on FirstSearch, I will often start my search in Academic Index ASAP instead. This admittedly is somewhat like the drunk who's searching for his lost car keys near the streetlight, even though that's not where he dropped them, because the light is so much better.

I do it because in my book, the first trick is to find SOMETHING, preferably full-text. InfoTrac has a better search system for finding SOMETHING. Once I've read it, I have a more precise idea of how to phrase the question when I get into FirstSearch's inferior searching system.

Which leads me to my 3rd Rule of Information:

---

RESEARCH IS A MULTI-STAGE PROCESS

While many straight informational questions can be answered on the spot, serious research should be a multi-stage process.

In most cases, we begin with a general idea of our topic--we might start with a question like "Why are frog populations declining?" In our initial search, we might find hundreds of articles. But the odds are, by the time we have actually read a few of them, our entire idea of what we planned to do has changed, and the rest of our search results are now only marginally relevant to our new topic.

Perhaps we've decided to investigate the thesis that pesticide contamination is responsible. In this case, we can follow up on the bibliography in the articles we've read. We can go back to our original database and redo our search, using new terms we've picked up, and/or the names of researchers who are pursuing that angle of investigation.

If we began our search in a general periodical database, like Academic Index ASAP, or Periodicals Abstracts, we may now want to pursue the topic further in Biological Abstracts, or Science Citation Index. Or we might want to look at Dissertation Abstracts or databases of conference proceedings. We might even want to go onto the internet to track down home pages or e-mail addresses for the scientists whose work we're using, to ask them questions.

At each stage in this research process we pick up new clues for further research and parlay them, spreading out in wider and wider circles until we finally decide, "Enough already."

---

HOW DO WE STAY CURRENT?

This is a question I am going to be asking each of the internet gurus I interview, because it may be the one single most useful thing I can offer you to apply to your own work life--"What do you read and what sites do you visit to find out about new technology, new sites, new ways of doing things?"

What I read on a regular basis is

• Yahoo! Internet Life, for a good roundup of web sites and tips on searching and technology,
• Searcher, for information on databases and searching skills,
• Wired, for info on technology trends, and discussions of clashes between government, society and technology,
• Choice, which decided early on to review web sites in exactly the same way they review books, and of course
• Library Journal, to keep up with new books.

  The newsletters I read online, or have sent to me include

• Freepint http://www.freepint.co.uk/, for essays on searching and selected sites in specific subject areas,
• Web Site Journal http://www.WebSiteJournal.com, for suggestions on site design and getting discovered by your target audience,
• Danny Sullivan's Search Engine Watch http://searchenginewatch.com/, for updates on new search engines and new features on old ones,
• AlertBox http://www.useit.com/alertbox/, for tips on page design, and
• Jenny's Librarians' Site du Jour http://www.jennyscybrary.com/sitejour.html, where Jenny Levine helpfully suggests important reading on the net along with great reference sites.

  For the sites I consult routinely when I'm looking for sites for NeatNew, see What's New on the Net http://www.sau.edu/CWIS/Internet/Wild/Netnew/netindex.htm

  I look forward to learning what my internet heroes and heroines are reading.