Which Database(s)?

Choosing and Using Databases

Choosing among the many databases

How on earth is one to make sense of the landscape of choices among databases? And is it ever going to be less confusing? How do I know when I've done everything I can?

Difficult questions. The array of available databases seems to change all the time, and each has its own peculiarities in coverage, search interface, possibilities for e-mailing or printing what's on view, etc. Here are some comments on the databases now available:

PAO (Periodical Abstracts Online)
PAO has 1600+ periodical titles, but they tend to be 'general', and don't cover all that much of the really 'scholarly' research literature of the sciences.
Verdict: good place to start, but quickly outgrown.
FirstSearch (which contains more than 60 specialized databases --access via Library Gateway => Research Resources => Periodical Indexes)
The Life Sciences and Medical subparts of FS have several essential databases, including BasicBIOSIS, AGRICOLA, BioDigest, MEDLINE, CINAHL. The basic search interface is the same for all (and isn't particularly flexible --no truncation except + for -s), and it's easy to e-mail records to yourself, to help build up a bibliography.
Verdict: good enough that one has to use it, but not very sleek in presentation.
Cambridge Scientific Abstracts is VERY useful, providing some new databases and also another form of access to MEDLINE and BioDigest. Its details are a bit labyrinthine, but well worth the time to get them under control.
Lexis/Nexis (consult General News and Medical Abstracts and General Medical and Health particularly --access via Library Gateway => Research Resources => Periodical Indexes)
An excellent source for full text of newspaper articles and some periodicals (notably Lancet!! [use General Medical and Health, select 'Medical and Health Journals' and add publication(lancet) under 'Additional terms:') which we don't have in our library.
Verdict: often the solution to otherwise intractable problems.
UnCover (words in article titles and author names --access via Liberty's Commercial Databases or Libraries & Research => Periodical Indexes for a web version)
The greatest advantage is a scope of 16,000+ academic journals, and very rapid updates of tables of contents (PAO and most FirstSearch databases are much slower to update). Really useful if you know an author's name but not what he/she has written. Somewhat clumsy interface.
Verdict: finds things that nothing else can.
Online Mendelian Inheritance in Man (OMIM) (Libraries & Research => Natural and Physical Sciences => Biology)
Absolutely essential for diseases with a genetic component. Constantly updated, with an astounding depth of detail and bibliography, and connections to gene databases and other resources (including MEDLINE references).
Verdict: essential for some subjects, and educational if you only look up "jumping frenchman"
NSF and NIH Grants databases (Libraries & Research => Natural and Physical Sciences => Internet Resources for the Sciences)
When you need to know who's doing what research and what they said about it to get funding, this is where you'll find a whole lot. You can search by keyword, by name, by institution. The records aren't research reports, but you can learn a lot about what's current and how science is presented to $ agencies.
Verdict: for most topics, here's the cutting edge of research.

A few words about annotations:

The idea is to summarize in your own words, and it doesn't have to be elaborate. Most articles don't have that many ideas --they're bricks in a towering edifice of scholarship, and each should contribute something new, some new insight or finding. Your task is to comprehend that contribution and express it in your words. If you've done the necessary background reading (and understand the terminology), you'll be able to do this pretty easily once you've had a bit of practise. It's a skill worth cultivating.

Try to avoid "In this article...", "This article is about...", "The authors of this article found that...". Cut to the chase and say what it is that the article has to say that makes it worthy of inclusion in your precis of the literature for your topic.