Page 9 - IDEA Study 2 2017 Predatory journals in Scopus

Compiling the database

Box 1 explains on how we constructed the database. Beall's lists were downloaded
on 1st April 2016. The names of individual journals and publishers were searched for
in the Ulrichsweb database on the same day, using an automatic script programmed
in Python. The script searched for the exact name of the journal or publisher and then
saved all relevant entries. Together with each entry from Ulrichsweb it also downloaded
a unique journal indicator in the form of the International Standard Serial Number
(ISSN). Journals whose ISSN were not found in Ulrichsweb, indicating that they had not
been registered with an ISSN or had a false ISSN, were excluded from further scrutiny.

In order to resolve any inconsistencies, for example due to inaccurate names, the entries
found by the automatic script were manually checked. Beall's lists consist of hypertext
links. To check our entries for independent journals we compared the ISSN
on the journal’s website with the ISSN found on Ulrichsweb. If the two ISSNs matched,
the entry was retained; if they differed, the entry was removed from our database.
Publisher’s identity was confirmed if at least one ISSN listed on its web site was found
in an entry linked to the publisher's name on Ulrichsweb. If a matching ISSN was found,
all other journals of the same publisher were also retained in our database. If no
matching ISSN was found, and the entries therefore referred to a different publisher,
they were removed.

In total we confirmed 4,580 ISSNs associated with Beall's lists. Many journals have dual
ISSNs, one for a printed and one for an electronic version. The number of individual
journals is 3,218, of which 281 feature on the list of standalone journals and 2,937 refer
to his list of publishers. Hence, the standalone journals represent only the tip of
the iceberg. Moreover, there are 1,003 hypertext links on the list of standalone journals,
from which follows that more than three quarters of them are not covered in Ulrichsweb,
let alone in more selective databases, and so apart from the unverified information
on their own web pages, we know nothing about them. For this study that is not
a problem, because journals that are not even included in Ulrichweb can hardly be listed
in Scopus. Nevertheless it is important to realize that predatory publishing is far more
widespread than we are able to demonstrate using existing databases.

The next step was to search for these “predatory” ISSNs in the Scopus citation database.
Once again, this search was performed using an automatic script programmed
in Python. For each ISSN detected in Scopus, the script downloaded basic data

7