Title Vrijednost postupka otkrivanja plagiranja u biomedicinskom časopisu : doktorski rad Title (english) The Value of Plagiarism Detection Procedure in the Biomedical Journal Author Ksenija Baždarić Mentor Lidija Bilić-Zulle (mentor)Committee member Amir Muzur (predsjednik povjerenstva)Committee member Bojana Brajenović-Milić (član povjerenstva)Committee member Vedran Katavić (član povjerenstva)Committee member Lidija Bilić-Zulle (član povjerenstva)Granter University of Rijeka Defense date and country 2012-01-01, Croatia Scientific / art field, BIOMEDICINE AND HEALTHCARE Universal decimal classificationUDC ) 61 - Medical sciences Abstract Ciljevi: Razviti postupak otkrivanja plagiranih radova u časopisu Croatian Medical Journal (CMJ) koristeći računalne programe za otkrivanje plagiranja te izmjeriti učestalost i značajke plagiranja.
Materijali i metode: Svi radovi (N=754) zaprimljeni u CMJ, 2009. - 2010. godine obrađeni su s pomoću programa za otkrivanje plagiranja (eTBLAST, CrossCheck, WCopyfind). Radovi koji sadrže udio podudarnog teksta veći od 10% smatrani su sumnjivima i raščlanjeni usporednim čitanjem.
Rezultati: Programima eTBLAST i CrossCheck pronađeno je 105 (14%) radova s više od 10% podudarnog teksta. Nakon provjere usporednim čitanjem 20 (3%) radova isključeno je iz daljnje raščlambe, a za 85 (11%) je zaključeno da su plagirani od čega su 22 (3%) rada bila samoplagirana. Zemlje podrijetla autora plagiranih radova su: Kina 18 (21%), Turska 16 (19%), Hrvatska 12 (14%) i ostale zemlje 39 (46%). Udio podudarnog teksta bio je podjednak u plagiranih i samoplagiranih radova (medijan=25%, 5-95 percentili (11%-59%) vs. medijan=28%, 5-95 percentili (14-47); P=0,634). Udio podudarnog teksta u plagiranih radova značajno je veći u radovima u kojima izvornik nije citiran u odnosu na radove u kojima se izvornik citira (medijan=36%, 5-95 percentili 16%-62% vs. medijan=23%, 5-95 percentili 13%-58% P=0,011), dok u samoplagiranih radova nema razlike (P=0,793). Najčešće se preuzimao tekst odjeljaka "Rasprava" (N=51), "Uvod" (N=51) i "Materijali i metode" (N=50). Podudarnost teksta u samoplagiranih radova najveća je u odjeljku "Materijali i metode" [medijan=61%, 5-95 percentili (41-68)], te je značajno veća od udjela podudarnosti u odjeljcima "Rasprava" (medijan=26%, 5-95 percentili (15-35), P=0,005) i "Rezultati" (medijan=23%, 5-95 percentili (17-36), P=0,009), dok u plagiranih radova razlike nema. Razvijen je postupak otkrivanja plagiranja i oblikovano izvješće o podudarnom radu.
Zaključak: Računalni programi za otkrivanje plagiranja uz kontrolni pregled rada usporednim čitanjem mogu se preporučiti kao standardna metoda za otkrivanje i sprječavanje objavljivanja podudarnih tekstova u biomedicinskom časopisu.
Abstract (english) Aim. To assess the prevalence of plagiarism in manuscripts submitted for publication in the Croatian Medical Journal (CMJ) and develop a standard operating procedure for scanning submitted manuscripts for plagiarism.
Materials and Methods. All manuscripts submitted in 2009-2010 period were analyzed using plagiarism detection software: eTBLAST, CrossCheck, and WCopyfind. Plagiarism was suspected in manuscripts with more than 10% of the text derived from other sources. These manuscripts were manually verified by the investigator.
Results. Of all 754 submitted manuscripts, 105 (14%) were identified by the software as suspected of plagiarizing. Manual verification confirmed that 85 (11%) manuscripts were plagiarized, of which 22 (3%) were self-plagiarized. Authors of plagiarized manuscripts (n=85) were mostly from China 18 (21%), Turkey 16 (19%), Croatia 12 (14%) and other countries 39 (46%). There was no significant difference in the text similarity rate between plagiarized and self-plagiarized manuscripts [median=25%, 5-95 percentile (11%-59%) vs. median=28%, 5-95 percentile (14%-47%), P=0,634]. Higher text similarity rate was found in true plagiarized manuscripts where the original source was not cited than in manuscripts in which the cited source [median=36%, 5-95 percentile (16%-62%) vs. median=23%, 5-95 percentile (15%-38%), P = 0.011], but no such difference was found for self-plagiarized manuscripts (P = 0.793). Differences in text similarity rate were found between various sections of self-plagiarized manuscripts (P=0.001). The text similarity rate of self-plagiarized manuscripts was higher in the "Materials and Methods" [median=61%, 5-95 percentile (41%-68%)] than in the "Results" [median=23%, 5-95 percentile (17%-36%), P=0.009] and or "Discussion" [median=26%, 5-95 percentile (15%-35%), P=0.005] sections. A standard operating procedure for scanning submitted manuscripts for plagiarism was developed and a plagiarism report was created. Conclusion. Plagiarism detection software in combination with manual verification may be used as a standard operation procedure to detect plagiarized manuscripts and prevent their publication in a biomedical journal.
Keywords
Plagiranje
Računalni programi
Recenzija
Višestruka publikacija
Znanstvenoistraživačka čestitost
Znanstveno nepoštenje
Keywords (english)
Plagiarism
Software
Peer Review
Research Integrity
Research
Scientific Misconduct
Duplicate Publication.
Language croatian URN:NBN urn:nbn:hr:188:348771 Study programme Title: Biomedicine Postgraduate (doctoral) study programme Catalog URL https://libraries.uniri.hr/cgi-bin/unilib.cgi?form=D1130217099 Type of resource Text Extent 92 str; 30 cm File origin Born digital Access conditions Open access Terms of use Created on 2017-01-19 17:30:44
