PubMed Changes

[Steve Jones]

David Goodman wrote:

> However, the best way of using PubMed is to not use Boolean logica at all,
> but to just find one clearly relevant article, and then use the Related
> Articles feature. Unlike other systems with that feature, this gives such
> good results that I present it to our undergraduates as the preferred
> way of searching.
> 
>  David Goodman, Princeton University Biology Library
> dgoodman at pucc.princeton.edu            609-258-3235

Yes and no.  I frequently use this method as part of a search, but it is
hardly a sufficient answer.  This method yields citations arranged in
willy-nilly chronological order (when, for example, the original
parameters were only for the most recent 2 years), with both English and
foreign language citations (even though the original search parameters
were for English language), and with citations that frequently only
relate to one or more parts of the original search (e.g., a citation for
paclixatel therapy for breast cancer when expanded in this manner might
end up yielding some articles pertaining to the search, but will also
probably include non-paclixatel drug therapy of breast cancer and
citations pertaining to paclixatel drug therapy in other forms of
cancer).

I keep meaning to send NLM an email asking them to let the search
parameters and chronological arrangement for the original group of
citations flow over to groups of related citations.  On the other hand,
I am extremely grateful that this database exists and think it is one of
the best available.

Steve Jones
U Md Baltimore County Library

"Knowledge is good."  Faber College