Planning assessment to promote academic integrity online

Submitted by miranda.prynne on Wed, 24/02/2021 - 09:00
Article type
Main text

Academic misconduct has increased around the globe since the onset of Covid-19, as recently reported in Times Higher Education.  

But it is important to remember that moving learning online does not suddenly turn good students into bad ones and the internet itself is not responsible for increases in cheating. Prior to the pandemic, there were multiple studies that consistently showed there was less academic misconduct in online courses compared with face-to-face courses. 

There is no way to design a cheat-proof assignment but there are approaches to assessment that promote academic integrity. Here I outline five of them: 

1) Just say no to recycling assessments 

Educators must set the example for academic integrity. One way to do that is to stop reusing old tests and assignments. Changing assessments a minimum of once per term is essential. There is a fundamentally unjust double standard if institutions prohibit students submitting previous work for academic credit, labelling it self-plagiarism, and then turn a blind eye to professors who do effectively the same thing when they recycle previous assignments and tests. 

2) Communicate clear assessment criteria 

Unclear assessment criteria can lead to students engaging in misconduct behaviours because they do not understand what is expected of them. For example, if it is not OK for students to collaborate or study together, then the instructor must say so. Do not assume that students will automatically know they cannot form study groups online, just as they would have done in the library when they were on campus. It is essential to talk with students about the boundaries of online study groups and when and how they are convened and used. 

Ensuring that everyone understands the assessment criteria is a shared responsibility. It is up to teaching staff to provide fair and clear assessment criteria and communicate expectations. In turn, students are responsible for asking for clarification if they do not understand. 

3) Assume students are sharing assessments 

We live in a sharing culture. As Susan Blum points out in her book My Word!, many students today grew up sharing all kinds of content online. They share music, movies, memes, and more. It may not be immediately obvious to students that sharing a meme is okay but sharing homework or test answers is not.  

It is naïve to think that assessments you have previously designed are not available to students online in some format. Bemoaning that fact will not change the global online culture in which sharing is the norm. Instead think about how you can use online content to your students’ advantage. Have them engage with existing online content, such as news stories and other media, rather than constantly creating your own new material every term.

4) Avoid pre-made test banks 

In addition to sharing assignments, it is just as easy to share tests and quizzes. Educators would be foolish to believe claims that “instructor-only” course packs, which often include pre-made tests, are not easily accessible to students. Some contract cheating sites specialise in the sale of academic tests, marketing them as “study guides” to students, when the exact same product is marketed to educators as “secure test banks.” 

There have been cases where sizeable groups of students have accessed test questions online prior to an exam, resulting in numerous students being found responsible for academic misconduct. This case with a group of nursing students from Canada is just one example, but the practice is now common for students everywhere so remove temptation by avoiding these tests. 

5) Lower the stakes 

High-stakes assessments are those where a lot rides on the outcome of a single assessment. Students are more likely to cheat on high-stakes assessment so shifting to lower-stakes assessment not only reduces students’ stress, it also promotes integrity. However, this must be balanced with a reasonable number of assessments. Having 100 small assessments each worth one per cent of the final course grade is not going to lower anyone’s stress. Lowering the stakes works when the total number of assessments is still manageable for everyone involved.

These are just a few ideas to promote ethical teaching, learning, and assessment. The common denominator to all of these tips is to focus on students’ learning. When we think of assessment as a process to help students learn, we promote integrity for both teaching and learning. 

Sarah Eaton is associate professor in the Werklund School of Education at the University of Calgary

The ideas presented in this post are from her upcoming book, Plagiarism in Higher Education: Tackling Tough Topics in Academic Integrity

Instructors can minimise the opportunity for academic misconduct through careful online assessment design. Sarah Eaton shares five approaches to online assessment that promote academic integrity
Instructors can minimise the opportunity for academic misconduct through careful online assessment design. Sarah Eaton shares five approaches to online assessment that promote academic integrity



Thu, 25/03/2021 - 16:50

user Id
User name
John McLachlan
Comment body
Thank you, Professor Eaton: some very interesting ideas! But there are also many different contexts in which assessment takes place. Re-using items in large MCQ type tests is part of good practice. It enables the psychometric performance of each item to be explored, and items to be piloted in advance to test for difficulty and whether the item discriminates against particular groups of candidates. It is of course necessary to have a sufficiently large item bank, and when this is the case, if a candidate can answer every item in the bank correctly, they have probably achieved the course objectives. Analytic methods are available to detect item preknowledge (and, indeed, collusion between candidates). Developing a new assessment each time runs the risk of having assessments of varying quality and difficulty, which are then hard to compare with previous versions wit regard to consistency of standards. Multiple small assessments create problems of aggregation: the effective weighting of each assessment is not just its intended weight but is also affected by its variance. Tests with high variance have a bigger effective impact than low variance tests and this phenomenon is not well appreciated by test setters. You say “Some contract cheating sites specialise in the sale of academic tests, marketing them as study guides to students, when the exact same product is marketed to educators as secure test banks”. Is there a link for this phenomenon? Re-using items in large MCQ type tests is part of good practice. It enables the psychometric performance of each item to be explored, Developing a new assessment
jwt token


Fri, 26/03/2021 - 09:02

user Id
User name
Jeffrey Vernon
Comment body
Students know that they can send their question to CHEGG, even during a timed remote exam, and get an answer (visible to all users online) in 15 minutes. The way to counter this is to devise test questions that the call center cannot be bothered to answer at the price they charge, because it would take too long. Two quick wins here: 1)Ask students to associate two or three different parts of the taught course materials in an open-book style exam. 2) Ask the students to prepare a set of slides to address a problem that requires an online search, or analysis of different written sources (I call this Timeslides). In this way, the answer is not 'out there' and cannot be bought so easily.
jwt token