Page 70 - Becoming Metric Wise

60 Becoming Metric-Wise

3.3.1 Plagiarism and Duplication
Plagiarism is defined as “the appropriation of another person’s ideas, pro-
cesses, results, or words without giving appropriate credit” (http://www.
ori.dhhs.gov/definition-misconduct/), while duplicate publication is
defined as an article that substantially duplicates another article without
acknowledgment, where the articles have one or more authors in com-
mon (http://www.nlm.nih.gov/pubs/factsheets/errara.html).
COPE (Committee on Publication Ethics http://www.publicatio-
nethics.org.uk) was established in 1997 as a forum for publishers and edi-
tors. It provides help on ethical issues such as conflicts of interest, data
falsification and manipulation, plagiarism, unethical experiments, priority
disputes, and so on. Publishers united themselves in CrossRef (http://
www.crossref.org) for the development of new technologies leading to
faster publication and easier submissions. CrossRef is also responsible for
CrossCheck, a software program to detect plagiarism, used by member
publishers. Some publishers have their own services for imposing ethic
behavioral, such as Elsevier’s PERK (Publishing Ethics Resource Kit).
Moreover, most universities use special software to detect plagiarism in
theses.
Yet, duplication detection software such as CrossCheck should be an
aid to human judgment, not a substitute for it. Human judgment should
decide if copying methods sections almost verbatim is acceptable and
within the domain of “fair use” (Samuelson, 1994). Zhang (2010) in a
discussion of and introduction to CrossCheck seems to judge that copy-
ing from methods is rarely acceptable. The same diligence applies to
republishing conference articles with little or no new facts or aspects, and
to so-called self-plagiarism, discussed further on.
Chaddah (2014) makes a distinction among three types of plagiarism:
text plagiarism (copying parts of an article without citing it), idea plagia-
rism (using someone else’s idea or hypothesis, again without crediting the
original source), and finally results plagiarism. In the last case, research has
been done but presented as original instead of a replication and verifica-
tion. He notices that plagiarism is not fraud, in which case data are fabri-
cated or unfavorable outcomes removed from the results. He observes
that in each of the events of plagiarism, papers should not be retracted, as
they are scientifically correct, but a correction must be published giving
credit where credit is due, and directly linked to the original paper and
this in such a way that every download of the original paper automatically