Learn

about good practices and sound policies that accelerate the effective adoption and use of technologies in teaching and learning.

Vernon Smith, Vice President of Academic Affairs, Rio Salado College

Rio Salado College Succeeds in Tackling Student Plagiarism

Interviewed by Russell Poulin, Deputy Director, WCET
January 2011

 

Rio Salado College's Peer-to-Peer Plagiarism Project detects among its 41,000 online students, using exclusively internal staff and resources. The system is intended to deter cheating, which will greatly benefit students by helping to ensure the long-term value and integrity of Rio’s academic programs.

 

Russ: Congratulations to Rio Salado College on receiving the 2010 WCET Outstanding Work (WOW) Award. WCET has been active in promoting good practices for academic integrity, student authentication, and student assessment. It is great that the WOW Award can honor a community college that is being proactive in addrssing this issue. Could you give us a little background about the history of plagiarism at Rio Salado College? What events lead you to decide to creat this plagiarism tool?

Dr. Smith:  Rio Salado College faculty have had a long history of promoting academic integrity.  Starting from the early days when the college first offered distance learning courses, the understanding of what plagiarism was and how to avoid it was highlighted in the course materials and shared by faculty members.  As online learning emerged with new technologies, the faculty quickly adopted tools and technologies (such as Turnitin.com) and using search engines to check for external plagiarism sources. 

Many assignments were specific to the context of the individual courses, so we needed a tool that checked individual student submissions against previous sections or semesters.  This type of peer-to-peer plagiarism is extremely difficult for a faculty member to detect, especially if one of the students involved is not in his or her section or took the course during a previous semester.  The Peer-to-Peer Plagiarism Detector was an innovative response to this need that reinforced our culture of academic integrity.

Russ:  Why did you decide to create your own plagiarism-checking software? Did you research any of the services or software products already available?  What features did you create that you did not find elsewhere?

Dr. Smith: Technology systems capable of identifying this form of plagiarism are available, but they are very expensive and are difficult to implement.  In light of these challenges, Rio Salado elected to construct a detection system internally using existing resources. The college determined that an in-house solution would not only save money, but also make it possible to customize the process around specific institutional needs. The detection system was built using existing staff and resources, which saved the college tens of thousands of dollars on third-party hardware and software.  It also created long-term savings by avoiding annual licensing and technical support fees that would likely have been required with an external software application.  Rio's plagiarism detector was conceptualized, researched, developed, piloted, and implemented over a five month cycle in 2009.

The plagiarism detector runs continuously, scanning incoming assessments and comparing those to historical submissions. It is driven by a mathematical text mining model that compares essays word for word. The system is extremely accurate in detecting cases of plagiarism and is successful in delivering a low false-positive rate. This system is also very robust. It can detect plagiarism even if students attempt to conceal the similarity between their submissions and the previous submissions of others. In fact, the detection system routinely identifies students who have taken a source document from another student and have altered it, either by modifying the paragraph order or replacing words with synonyms.

Cases of cheating are typically detected within 12 hours of initial submission. Most all submissions are scanned automatically, meaning that faculty members do not need to request for an assessment to be included in scanning.

When a new case of peer-to-peer plagiarism is discovered, faculty chairs are alerted automatically via email. These “Plagiarism Alert!” emails indicate how similar the two submissions are (i.e. 85% similar). They also contain all relevant student and course information, along with the full text submissions from both students. The submissions are presented together with exact matches highlighted to help facilitate accurate analysis. Although the system detects both semantic and exact matches, only the exact matches are highlighted within the emails for simplicity. Upon receiving an email alert, faculty chairs follow through with their own department-specific protocols.

In summary, it is fast, accurate, and very robust.  It provides very specific information to faculty chairs that are then able to make clear decisions to reinforce existing policy.  The Peer-to-Peer Plagiarism Detector surpasses our expectations of its ease of use and reliability.

Russ:  That sounds very impressive. Could you provide more information about what faculty have to do?  If a work is flagged for plagiarism problems, what is the faculty person instructed to do?  How do they handle the incident? What do students have to do?

Dr. Smith:  As described above, if a case is detected, a report is sent to the faculty chair with all the important information.  The faculty chair then investigates the case—and checks, for example, if this is the first time a student has been in violation.  With this information, the faculty chair applies their departmental policies to the situation which may involve requiring the student to take a remediation module on academic integrity, to assigning a “0” on the assignment, to withdrawing the student or giving an “F” for the class.  Other, more serious consequences can be recommended by the faculty chair to administration as well—including expulsion.  

Students typically learn from their mistakes.  Cheating is not automatically the motivation for plagiarism—sometimes it is an understanding of appropriate documentation or a change in an educational cultural expectation. 

Russ:  What has been the feedback from faculty and students? 

Dr. Smith:  For students, word of mouth has been the best feedback.  Students share with each other. 

John Jensen, Mathematics Faculty Chair and Rio Salado College Faculty Senate President commented that:  “The peer-to-peer plagiarism detection system is a key element in fortifying Rio Salado’s commitment to academic integrity. What is more, it reinforces in students the formation of ethical behavior and self-reliance as critical values in their character development and their learning as a personal responsibility. In the end, we feel that this system supports authentication in the online learning environment that meets or exceeds the capabilities of the physical classroom.”

Russ:  What advice would you have for others who are seeking to address problems with student plagiarism?

• Be sure that your policies are in place.  Academic integrity is most effective when it is an expression of the academic culture.

• Be sure that technology systems are in place with business processes.  If an incident of plagiarism occurs, what are the next steps?  Are those steps documented and explicit in the syllabus and college policies?

• Start with the faculty.  This is in the domain of faculty and the success and consistency depends upon faculty support and enforcement.

Russ: Thank you, Vernon and congratulations once again on your success with this innovation. 

See our webpage on student authentication issues for good practices and other publications on this issue.