ProQuest newspaper problems (when is full-text NOT full-text)

[Stacy Pober]

It's time for me to update our consolidated journal availability
database, and I find that the ProQuest National Newspaper collection has
conflicting coverage dates.  The list contradicts itself, so it's hard
to know which coverage information is correct.

For example, The Christian Science Monitor is listed as full text for
1/3/95-12/3/96 but a separate entry lists it as full-text from
9/30/88-current.  

There are similar conflicts for the Atlanta Constitution, the Atlanta
Journal, the Atlanta Journal/Atlanta Constitution (combined), Boston
Globe, Denver Post, Detroit News, Guardian, Houston Chronicle, Los
Angeles Times, San Francisco Chronicle, St. Louis Post, Times-Picayune,
and USA TODAY.

There's some problem in doing searches and retrieving full-text in the
date range that has the 'conflicts'. For example, the list says that the 
Houston Chronicle has FT from 1/14/95- current and also that it has FT
from 1/1/95-12/3/96.  There's a comment in the second entry that states 

If you do the following search in ProQuest: 
PDN(<11/01/1996) and clinton and source(houston chronicle)

You'll see that only some of the articles are in full-text.  But take a
look at the third and the sixth citations on the list:  "Candidates
still hunting for votes..."  and "Dole has a lead..."  Both are listed
as being citation only.  But if you go into the Search by publication,
go to that date for the Houston Chronicle, you will find the full text
for the "Dole has a lead" article.  You won't find any entry for the
"Candidates still hunting" article.  This isn't a unique glitch, it
seems to affect every newspaper that has these conflicting entries. 

Since most ProQuest users look for newspaper articles by searching
rather than browsing by newspaper title, this means that some of the
full-text is effectively hidden from them.  Few students are going to
try browsing for the full-text in the "for publication" utility if their
search result tells them the article is citation-only.

I first pointed out this journal list discrepancies to UMI in March
2000, but the list still has duplicate entries that provide conflicting
information for most of the major papers, and the search facility still
lacks the links to some full-text that IS in there.  

I've contacted UMI (again), and I'm waiting for a response, but I was
hoping that perhaps someone had a current, accurate list from them. 
(Another reason for posting this is that I figure knowing about the
'hidden full-text' might come in handy to other folks.)

-- 
Stacy Pober
Information Alchemist
Manhattan College Libraries
spober at manhattan.edu
http://www.manhattan.edu/library/