Research School Network: Academic Integrity, Cheating and Online Exams A brief exploration of the evidence around minimizing unethical behavior and maximizing validity of online assessment

With the most recent period of remote learning, many teachers and schools are facing up to the challenge of assessing students remotely. This is of particular pertinence for secondary schools that may have been unable to complete mock exams for GCSE and A‑Level groups prior to the partial closure of schools.

The problem is that remote assessment is highly challenging, and there will naturally be questions over academic integrity, the extent of cheating and as such the reliability of any data collected. It is a conundrum that is pretty hard to resolve. As Margaret Williamson, at the University of North Georgia, puts it ​“there is no fool proof way” to prevent unethical behavior, such as cheating, in online assessment, but it can be ​“minimized”.

As with much of the existing research on online learning, studies into effective online assessment or exams mainly stem from works focusing on university or adult courses which, even prior to the pandemic, often involved some form of online learning and assessment – so we must consider any findings or suggestions with this in mind.

The first thing to consider is whether or not our concerns around the integrity of online assessments are valid. Although early literature was somewhat inconclusive on the extent of online ​“cheating”; with some studies findings no discernible difference in its prevalence between online and conventional assessments, more recent studies have begun to collate mounting evidence that it can be substantial if the correct conditions are not put in place (Dendir and Maxwell, 2020). For example, Alessio et al (2017) found significant score inflation on unsupervised online assessments completed by undergraduates at Miami University versus their peers that completed proctored assessments.

If we are aware of the possibility of unethical behavior in online assessments, then we must plan to try and reduce this. To do so we must consider why students may be tempted to cheat. Hylton et al (2016) state that cheating occurs when the examinee has both the motive as well as the opportunity to do so. Motivation can be influenced by a lack of interest or preparation and performance consequences, while opportunities can be influenced by the exam design/​environment (Dendir and Maxwell, 2020). On top of this students may hold vague ideas of what constitutes academic dishonesty, for example a study by Watson and Sottile (2010) noted that while students reported equitable rates of cheating in face to face classes versus online classes, students did report that they did have help from someone in the online classroom but perceived this an ​“acceptable collaboration” rather than cheating.

Attempts to minimize unethical behavior/​cheating may therefore be split into strategies addressing motivation and strategies addressing opportunity. One way the research suggests to tackle the motivation issue is to reduce the stakes of assessment so that performance consequences do not drive cheating. This can be achieved through more low weighted tests rather than few heavily weighted tests; however, this is not always feasible for summative assessment points such as mock exams. It is therefore vital that in the build up to any online exams educational settings remind students of the importance of academic integrity and honesty, which may also involve sharing assessment polices or a code of conduct with students. In doing so teacher can attempt to reduce the motivation to cheat by explaining the additional benefits of exams/​testing linked to learning and allowing teachers to identify gaps in student knowledge that need filling. Any code of conduct should clearly set out what is and is not acceptable so that there is no confusion amongst students.

Reducing the opportunities to cheat may be more complex, although perhaps more tangible. There has been a significant amount of research into the value of proctored (supervised online assessments) with research showing that when online assessments are supervised remotely (i.e. via webcam) then the likelihood of cheating is significantly decreased. However, proctoring all online assessments for large cohorts is challenging, can cause anxiety amongst students and with larger numbers to monitor the opportunity for cheating increases.

There are other strategies that can be used, although there is a growing consensus that these should supplement supervision rather than replace it. Many authors encourage teachers to create their own assessments or questions rather than using test banks as these can often be accessed by students. In addition to this it is common suggestion that questions, where possible should be conceptual or apply knowledge to a situation so that students cannot browse the answers online or in books etc. Some authors have suggested that online exams be treated in the same way as open book assessments, eliminating the need to cheat and then taking this into consideration when awarding grades. However, this has numerous issues such as is not being reflective of most terminal exams. Other suggestions include randomizing the question order or having a large bank of questions which are selected at random for each student, however again the feasibility and fairness of such systems is questionable, particularly in subjects where questions set in exams often have a set or logical sequence of content and challenge.

The use of time limits on assessments can also limit student’s opportunity to cheat by limiting the time they have to complete assessments, so that there is insufficient time to look up answers and complete the assessment. For those using Google Classroom there are extension available which allow you to set time limits on quizzes etc. The timing of completion is also a consideration, with the opportunity to share answers reduced by ensuring the assessment is done simultaneously and that feedback (especially self-marking answers) is not released immediately, so that any students who complete the assessment early are unable to share feedback with their peers.

To sum up then, I think it best to refer to my opening point – there is no fool proof way of ensuring academic integrity when completing online exams, but we can take steps to encourage and promote it.

By Ben Crockett (@bencrockett1)

Ben is an Associate Senior leader at Durrington High School. He is also a Research School Associate for Durrington Research School and will be delivering our training on ​“Effective Use of the Pupil Premium Fund” with Marc Rowland and our ​“Curriculum, Teaching and Assessment Training” with Shaun Allison. 

References:

Alessio, Helaine & Malay, Nancy & Maurer, Karsten & Bailer, Albert & Rubin, Beth., 2017, Examining the Effect of Proctoring on Online Test Scores. Online Learning

Hylton. K, Levy.Y and Dringus. L.P, 2016, Utilizing webcam-based proctoring to deter misconduct in online exams Computers & Education, 92 – 93

Seife Dendir, R. Stockton Maxwell, 2020, Cheating in online courses: Evidence from online proctoring, Computers in Human Behavior Reports, Volume 2,

Watson. G and Sottile. J 2010, Cheating in the Digital Age: Do Students Cheat More in Online Courses?, Online Journal of Distance Learning Administration

Williamson Margaret, 2018, Online Exams: The Need for Best Practices and Overcoming Challenges, The Journal of Public and Professional Sociology