The increased use of distance and online education brings increased attention to the problem of academic dishonesty. Cheating can take many forms, including cheating during an exam, lying about submitting work, plagiarism, and unauthorized collaboration. It appears that the only category of increased cheating is homework copying, which can be considered plagiarism or unauthorized collaboration. In an MIT study reported by Dave Pritchard (Note 1), cheating was defined as inputting a solution to a previously unread online homework problem faster than the time needed to read the problem. Defining cheating in this way allows instructors to track online cheating and study its effect on exam grades and passing rates. Based on his research about cheating, Gerd Kortemeyer at Michigan State University recommended the following actions to minimize cheating: (1) reduce multiple choice questions, which lead to solution-oriented discussions, (2) calibrate problems to have mid-range difficulty, and (3) give more frequent exams, such as short weekly exams. (Note 2)
Workshop participants discussed a variety of reasons why students cheat: a culture of viewing academic work as a series of requirements, rather than an opportunity to learn; societal pressure to achieve at all costs; the pressure of time, which can cause students to calculate that cheating can be worth the time saved, even if they are caught and receive a lower grade; and the US system of written and multiple-choice exams, which facilitates cheating. In other countries, teachers give oral exams, which make cheating far more difficult.
The discussions about online assessment addressed the question of how to assess so that students are using higher-order thinking, such as analyzing and evaluating information. Participants also discussed how to develop useful formative assessment.
There are various ways to provide the comprehensive assessment that instructors seek. Online homework platforms use randomized numbers to encourage collaboration without cheating, and use symbolic questions and video to assess higher-order thinking. Flipped classrooms were cited as one way to more deeply assess student learning, because these kinds of classrooms can provide time for oral exams. The potential to integrate formative assessment into online systems may be provided by platforms that assess whether students are engaged, gaming the system, or off-task. By determining whether students are engaged in real time, these online homework platforms could give students a tutorial on what they missed or provide them with a reminder that they are not paying attention.
Even a carefully developed and tested educational reform can turn out to be difficult to implement. Melissa Dancy from the University of Colorado, Boulder, argued that one barrier is a faulty model of change (Note 3). Her research shows that many reforms are currently undertaken using a Development and Dissemination model of change, in which specialists develop strategies with proven success, with the expectation that faculty will replicate the changes and have similar success. This model ignores the contextual constraints that faculty face, including the need for additional time to learn new pedagogical methods, resistance from students, and lack of administrative support. In contrast, a model of change involving collaboration is more likely to succeed. Advocates of reform must include other faculty in decision-making, and must expect that changes may be piecemeal and may involve integrating a variety of changes, adapted for local circumstances.
Note 1: D.J. Palazzo, Y. Lee, R. Warnakulasooriya, and D.E. Pritchard, Phys. Rev. ST Phys. Educ. Res. 6, 1 010104 (2010).
Note 2: J.T. Laverty, W. Bauer, G. Kortemeyer, and G. Westfall, Phys. Teach. 50, 9 (2012).
Note 3: M. Dancy and C. Henderson, Am. J. Phys. 78, 10 (2010).