Examples of unintentional(ly amusing) errors in citation databases
Maxine Clarke
Thursday, 05 June 2008 09:39 UTC
Debbie Chaves of Wilfrid Laurier University, Waterloo, Ontario writes in Correspondence this week (Nature 453, 719; 5 June 2008):
In his Correspondence Hall and Keynes join Arbor in the citation indexes -(see this Nautilus post), Daniel Postellon describes the distinguished careers of Milton Keynes, Walton Hall and Ann Arbor. In the last case, I note that Professor Arbor has an h-index of 1 from the Web of Science database provided by Thomson Scientific’s ISI Web of Knowledge. This is based on her five citations for the year 2007: two articles, two letters and one abstract.
An author search in the Web of Science reveals that Chevy Chase MD (not to be confused with Chevy Chase, Maryland) has co-authored a letter with Howard Kaplan (H. Kaplan Am. Sci. 96, 3; 2008). My own institution, Wilfrid Laurier University, is also an author (S. Cadell et al. J. Palliat. Care 23, 273–279; 2007).
Irrespective of how these errors are created, the rising use of systems in which citation information moves directly from the search of a database or citation index to a bibliographic management system, and then into a reference list, means that inexperienced students and researchers who are not savvy enough to detect these errors will propagate them further.
Vigilance is required by all users of citation indexes and databases.
Any other examples?
Updated 17 September 2008 14:02 UTC
-
Replies
-
The query Author=(leuven ku) (for my institution K.U.Leuven) yields 37 results in Web of Science: 24 meeting abstracts, 2 book reviews, 1 editorial and 10 articles. See (hopefully) my reaction in Correspondence.
-
(It didn’t make it to Corres, so why not posting it here?)
In the Correspondence ‘Learning to navigate the geography of citation indexes’ (Nature, 453; 2008), Debbie Chaves calls for vigilance of all using citation indexes and databases because of the many errors in the Web of Science database provided by Thomson Scientific’s ISI Web of Knowledge.
Errors are indeed abound: institutions are sometimes listed as authors (Author=(Leuven KU) (for my institution K.U.Leuven) yields 37 results), some authors are listed with the wrong initials (the two records Author=(Muys A) should be Muys B), and abstracts quite often carry the wrong information (spelling errors in species names, numbers become letters and letters become numbers). The latter is particularly treacherous because it is to be expected that some authors cite directly from the web or the database, without accessing the original publication.
But it is easy to point a finger at Web of Science. In my opinion, it is also the duty of the authors themselves to perform a quality control, like it is their duty to proofread their own papers. As key expert of the record, the corresponding author should check the information in the record, and if any errors are detected, report them using the online correction form. According to the reply to one of my correction requests, it is possible to change “source author names and addresses, terms in a record’s title, author keywords and abstracts, some types of cited references, and information about document types and language codes”. Although, according to the same source, “data change process usually takes three or four weeks to be identified, updated and uploaded to the database” (where more recent publications receive priority above older records), I have seen none of my suggested corrections implemented thus far.
-
Oh yes, it was this thread that reminded me to post about citations of mis-citations. It’s sort-of relevant.
-
