Auto-plagiarismo em ciência

This month sees the launch of CrossCheck, an anti-plagiarism system for academic journals created by iParadigms to help publishers verify the originality of submitted work. It will cover 20 million journal articles from major publishers including Elsevier, Nature Publishing Group, Oxford University Press and Sage.

John Barrie, chief executive of iParadigms and the man who developed the technology behind Turnitin, the plagiarism-detection software, described self-plagiarism as a "huge" problem.

"Academics receive tenure based on their publications - it is publish or perish. That system creates this massive conflict of interest," he said.

"Anybody who has done any research knows it is very difficult to do. You just can't crank out five, ten papers a year unless (...) you have a research team of 20 people."


A pilot study by Tracey Bretag and Saadia Carapiet [ver também] from the University of South Australia found that 60 per cent of authors in a random sample of 269 papers from the Web of Science social science and humanities database had self-plagiarised at least once in the period 2003-06. Self-plagiarism was defined "quite generously" as occurring when 10 per cent or more text from any single previous publication was reused without a citation.

"The truth is that if these authors had self-cited in each case, it is unlikely that the editors would have published their work because they would have seen that it had all been published before," Dr Bretag said.

[Fiquei tão admirado, que mais não faço que copiar e colar pedaços do texto do artigo de Rebbeca Attwood, no THE. Como será a situação no campo das Ciências Exactas e Naturais?]