Development of New Complementary Metrics for the Authenticity of Student Papers

At Zagreb University of Applied Sciences Turnitin software has been in use since 2014 for bachelor's and master's theses plagiarism detection, as well as for authenticity checking of seminar papers. The decision about acceptable authenticity is made upon software generated originality report consisting of percentages showing similarity index of newly created text and already published sources. In this paper, based on the normalized Herfindahl – Hirschman Index, the metrics for authenticity level decision has been developed through new descriptive indicators: Effective number of different sources (ENDS) used in a text and Effective similarity index per source (ESIS). The metrics is descriptive. The usage of ENDS and ESIS indicators is examined at the Department of Electrical Engineering, proposing their introduction as new complementary indicators to the existing originality report for making decisions on student text authenticity.