9-24-93
 
TO: Tom Mead, Dartmouth, and MEDLIB-L Readers
 
WHOA!!! What is this "Yet-Another-Article ..."??? Where did it appear?
Where are the others? I want to keep up with the literature like the rest
of you and I try as best I can, but am I missing something? How about a
little help?!
 
I think you raise some interesting points, but how do we know it's as bad
as you paint it? Maybe you're having a bad Friday (you're allowed), or are
flaming for good reason. I would be more sympathetic if I knew more about
what you are talking about!!!
 
And frankly, if your point is that good and worthwhile, and I am willing to
give you the benefit of the doubt that it is, why not spend your Friday
afternoon WRITING THE EDITOR OF THE JOURNAL THAT PUBLISHED A POORLY
RESEARCHED ARTICLE!!! (And I want to know the name of the journal -- if
it's that poor in quality, I don't need to support it with institutional or
personal dollars, right?! That kind of action will send an even stronger
message!) Don't preach to the converted -- use your energy and preach it
the offending party! TEACH THEM THE ERRORS OF THEIR RESEARCH -- maybe
suggest the article should be retracted! And indexed appropriately with a
MeSH heading!
 
And if I missed a citation in your long message, please let me know -- I
maybe having the bad Friday (and I'm allowed too)!
 
Thanks for the raising the issue, and I really do look forward to your
bibliography!
 
Sincerely, Jim Shedlock, Director, Galter Health Sciences Library,
Northwestern University, Chicago [log in to unmask]
 
>Yesterday I read Yet-Another-Article (published in an important medical
>journal that we all know and get) where someone did a big search on MEDLINE.
>The author bemoaned (with statistics and charts and graphs) about how
>"Medline missed x- percent" of thus-and-so -- and this is why a whole new
>database of blah blah blah ought to be funded, invented, maintained,
>promoted, and made accessible for all the doctors in the world.
>
>The search strategy was displayed in the article. The search was not done on
>MEDLARS, DIALOG, BRS or any other large file-vendor where all the necessary
>MeSH-explosions and other crucial search-tricks are at the disposal of
>experienced, savvy searchers. It was done mostly with text-words. There was
>no consultation or strategy-making with an experienced searcher.
>
>But: "Medline missed" all sorts of important & relevant references, and so
>it is thus hopelessly inadequate to serve these newly defined needs blah blah
>blah...
>
>(...OK. That was the background and here comes the flame....)
>
>Folks: we must, with renewed diligence, get the under-informed to realize
>what MEDLINE is and what it is not. First & foremost, it is not a living,
>sentient being. MEDLINE doesn't "miss things" nor does my new Subaru station
>wagon if I steer it into a tree. It's not a glowing, infinite, exhaustive,
>'last word', medical god-oracle-thingy... It has come to be known to the
>public and the medical establishment as an infinitely large, generic, perfect
>and absolutely complete, mysterious, all-knowing entity. Some of us know
>better and have reasonable expectations of this very nice tool.
>
>These faulty perceptions out there really need to change before something
>happens more awful than just the wasted expenditure of horrendous amounts of
>time & energy & dollars. We need to DEFINE this beast for people and get
>them to understand it in its proper context. (And context &
>searchability-factors are DIFFERENT in different 'mountings' on the various
>vendors, university OPACs, CD-ROM products, and cutsie-friendly 'interfaces'
>to all these...).
>
>We need to DO their searches (on powerful & adequate 'mountings' of the
>file!) if we want to prevent authors from publishing wrong facts,
>embarrassing themselves and their institutions, and embarking on half-baked
>new projects. We need to discuss, promote, and use alternatives to MEDLINE
>(and we need to develop and maintain our competence at using them). Journal
>editors should send these sorts of articles (where MEDLINE retrieval IS the
>topic) to medical reference librarians for review. It took me all of 2 or 3
>minutes to see the fatal flaw in yesterday's article. It didn't matter that
>there was all sorts of lengthy & scholarly-looking analysis of what was found
>and what was not. The search (and the 'mounting' of the file and thus the
>searchability of the file) was completely inadequate for the particular task.
>To me, that rendered useless any subsequent "analysis."
>
>We need to get INVOLVED with these "...LET'S REINVENT THE WHEEL AND MAKE
>ANOTHER NIFTY DATABASE..." projects. For many of these proposals, I feel
>someone ought to call NLM to see if MEDLINE could be tweaked or enhanced into
>doing what they think they need, or to determine whether NLM or NIH has or
>knows-about projects already underway or contemplated to do thus 'n such.
>Rarely are these ideas completely "fresh." They frequently come from a
>particular specialty - and the authors figure that nothing at all like what
>they need to do has ever been done before.
>
>Lastly, when we offer bibliographic instruction (workshops, minicourses,
>demos) for the latest internal or cute or "free" MEDLINE or
>interface-to-MEDLINE, we MUST give a balanced & careful "pitch" about this
>business. (It isn't easy and it takes time). Don't fire up the computer &
>projector and begin raving happily. Be instead the picture of careful,
>concerned scepticism. Start with a low-tech, slow, careful soberness. Tell
>them: "We're gunna put this business into perspective for a moment on the
>board here and THEN we'll go online..." (You'll have plenty of opportunity
>in a few moments for fun & amazement & impressing the audience...) Start
>with a black-board presentation where you outline the computer & paper
>versions (years-of-coverage, number-of-journals-indexed, format,
>availability, etc. etc. as well as reasonings about "WHEN TO USE..." and
>"WHEN TO RUN FOR HELP..." factors) of Index Medicus, Biological Abstracts,
>Science Citation Index, Excerpta Medica, assorted other files & resources in
>Cancer, Nursing, Pharmacology, Hosp. Administration, Toxicology, Ethics,
>Education, Psychology, GovDocs, etc. etc. as well as the "Big Federal
>Medline*" and (finally) the little local version or friendly interface of the
>miniature & partially disabled "MEDLINE" you're about to demo. In this way,
>workshop attendees will see how & where "MEDLINE" fits in to the total
>picture; they'll also see what fraction of total-MEDLINE is really available
>to them.They'll start to get an understanding of the SIZES of things.
>
>*call it "MOTHER Medline, BIG-BERTHA Medline, Real Medline" -- do
>whatever is necessary (and consistent with your style & lingo) to
>DISTINGUISH it from the 'miniMEDLINE' or the disabled-because-of-friendliness
>MEDLINE you're about to launch.
>
>Frankly, I come right out and tell them:
>
>"Don't be publishin' anything where you say 'WE SEARCHED MEDLINE AND......'
>without showing it to ME first...! I don't trust your MEDLINE and I *really*
>don't trust your searchin' "
>
>[This does sound a tad egocentric, so I humor-it-up just a little, and
>provide examples of search-idiocy and how to "miss" things, but I'm dead
>serious. I think they need to know they're just playing around and there's an
>expert (who can afford to be cocky) at the end of the phone or e-mail
>connection...]
>
>Yes, searching is now "easy" and loads of fun and oh so helpful...
>
>But the flip side is that it's very easy to get it ALL WRONG and to discover
>that "MEDLINE missed things..." which to me is like saying "my SUBARU
>crashed..." Duh...
>
>
>=====
>
>Thomas Mead, MLS
>Biomedical Libraries
>Dartmouth College
>Hanover, NH 03755
>
>[log in to unmask]
>(603) 650-1741
>
>(I do indeed LIKE my new Subaru and I drive it after consulting the TREES,
>etc. Bye!)
 
************************************
James Shedlock, Director
Galter Health Sciences Library
Northwestern University
303 E. Chicago Avenue
Chicago, IL 60611
VOICE: 312-503-8133
FAX: 312-503-8028
EMAIL: [log in to unmask]
************************************