Advice on How to Prevent Academic Dishonesty

Introduction

We ask all instructors and their teaching assistants for their help in preventing academic dishonesty in our courses and in reporting incidents to our program's office. The advice on this page is designed to supplement the University's Academic Integrity Policy by providing guidelines and examples specific to our courses.

Past examples of academic dishonesty in our program

Our students, alumni, instructors, teaching assistants, and staff members have each reported instances of cheating or attempted cheating which include all of the following, despite specific and repeated warnings:

Copying a written homework solution from another student;
Copying a written homework solution from a previous year's solution;
Copying in-class programing quizzes;
Copying programing assignments (C++, MATLAB);
Copying final project programming code from another student;
Copying from a neighbor during an in-class exam;
Submission of a previously graded final project from another course as a new final project;
Copying a final project report and code from another student who had taken the course in a previous year.

We ask instructors to help strictly enforce the University's Academic Integrity Policy, with the advice of your graduate program director, and report suspected instances to our program's office.

Why it matters

The harm caused by cheating among students is described in the University's Academic Integrity Policy. In addition to obvious academic integrity concerns, cheating

Damages the reputation of our program;
Hurts the morale of honest students;
Erodes respect for our instructors and our program when cheating goes unaddressed;
Negatively affects placement.

The issue has become a major issue for new graduates and some alumni say they would be reluctant to hire students from our program until this issue is addressed decisively.

How to prevent cheating

The methods described below are not intended to be exhaustive.

General

Remind students of the academic integrity code, in class and in writing, how the code applies to your specific course, explain the difference between permissible collaboration and unapproved copying, and warn students of the consequences of cheating;
Ask any students who have concerns about cheating to bring them you or to our office;
Please share with our office any tips you may have to help eliminate cheating.
Cheating often occurs when students feel they have no other alternative: please consult our guidelines for new instructors for advice on how to avoid such situations;
Consult our academic integrity resources, including links to the University policy and guidelines provided by other departments (for example, Computer Science) to combat cheating;
Use the class roster to take attendance. Our program office receives complaints that some students arrive late or leave early, disrupting the class.

In-class exams

Please include the following methods in your efforts to prevent cheating on exams:

If using Rutgers blue books for exams, be sure to collect all blue books, whether used or not;
Use alternating copies of exams, with different colors and questions in different orders;
Randomize seating order (using assigned seating) to avoid friends or cliques sitting together;
Require all students to bring government-issued IDs or RUIDs and check them against your class roster;
Space students at least one seat apart;
Have at least one proctor per 20 students and remind them of the importance of monitoring students carefully for the duration of the exam;
Request additional proctors (our office can help) if your TAs or graders are unavailable;
No electronic devices (calculators, mobile phones, Blackberries, etc) of any kind allowed for any purpose during the exam;
All papers or books, except exam papers and blue books, must be stowed during the exam, unless open book;

Homework assignments

Since collected homework can never be fully screened for copying, it may be more effective to put greater effort into preventing cheating on in-class exams and quizzes which can be closely monitored or in final projects, and instead de-emphasize the role of graded homework in determining course grades.

Limit grades for collected homework to no more than 10% of the total course grade, except in special courses where regular practice is deemed essential (C++ or other programming, final project work, etc) and which cannot be replaced by in-class quizzes or exams;
Encourage TAs and graders to give essentially full credit for obvious effort, lowering the incentive to copy and increasing the incentive to use homework as a learning experience and not a determining factor in the course grade;
Replace the balance of customary homework assignment grades (traditionally, 25-30% of the course grade had been typical) with a second midterm or in-class quizzes closely based on homework assignments;
Rotate homework assignments so that no two similar assignments are used in back-to-back years (most of our master's students graduate in under 2 years);
Explain clearly what level of collaboration is acceptable;

Final project assignments

Vary final project list so that no two similar projects are assigned in back-to-back years (most master's students graduate in under 2 years);
Randomize project assignments, so friends or cliques are not assigned the same projects;
Never allow students to use their own (or their firm's) non-public data;
Never allow students to use a project which they propose instead of one on your project list (they could be attempting to recyle a work project or a project from another course);
Do not allow more than 4 or 5 students to work on the same project (Math 16:642:623 does not allow more than 3 students per project);
Check final project report code and report text carefully for evidence of cheating, comparing code and text with projects from other students in the class and past students;
Schedule oral exams (15 to 20 min per student) using TAs and additional help as needed, asking students to explain their project code and report text face to face;
Require students to include a signed statement in their report stating that the project report represents their own work and includes complete citations to all references or sources of help, including other students or work colleagues;
Explain clearly what level of collaboration is acceptable, keeping in mind that students must submit individual reports and are graded individually.

Use grading schemes designed to limit opportunities for cheating

Explain your grading scheme carefully to your students on the first day of class and provide it in writing. Below are two examples of grading schemes used for our program's courses:

Example 1. Math 16: 642:621 & 16:642:622, Mathematical Finance I & II. This is a traditional mathematics graduate course with little or no programming (aside from occasional use of MATLAB or Excel-Visual Basic) and no project: 45% final (3 hour), 20% for each one of two midterms (80 minutes), 10% homework (weekly, lowest 2 dropped), and 5% attendance.

Example 2. Math 16: 642:623, Computational Finance. A mathematics graduate course emphasizing algorithms and methods, with both analytical and C++ programming assignments: 40% final project, 20% for each one for two midterms (80 min), 15% homework (weekly for first 8-10 weeks, all counted), and 5% attendance.