Technology and Academic Misconduct
Using technology to detect and prevent cheating.
Technology tools alone cannot foster a culture of academic integrity, nor can they prevent misconduct.
However, when used appropriately and in conjunction with other approaches, they may help to prevent or detect academic misconduct and support academic integrity.
Technology is always changing and these tools are readily available, along with other resources that give step-by-step instructions for specific solutions.
Things to reflect on
Have I taken care to design assignments and assessments so that students need to do their own work?
Have I informed students about what data is available and how it can be used?
Cheating Deterrents
As explained on UBC’s Keep Teaching website, breaches in academic integrity are often results from a combination of desperation plus opportunity. Using cheating deterrent techniques can help minimize the latter.
Question randomization
How can students copy from each other on an exam if everyone has their own version of the exam? While most online testing software solutions (e.g., Canvas Quiz tool) do not provide this level of randomization, instructors can discourage students from working together by providing different versions of questions. A simple type of randomization involves having multiple choice questions display the options in a random order. For quantitative questions, another type of randomization involves having the software choose random parameters and computing the correct answer with a formula you must define.
Why use question randomization?
Having multiple versions of a question will discourage collaboration, and doing this on a number of key questions in a timed exam will make it basically impossible for students to help each other. Three or four versions of a question may be sufficient. Question randomization can be used with in-person and remote exams.
What to watch out for
Test design and creation takes more time, and requires more careful editing and planning.
Question banks
Similar to question randomization, question banks involve creating more questions than any individual student will see on their exam, and having the software randomly choose a subset of them. This can be set up in question groups, so you have considerable control over how many questions are presented from each topic.
Why use question banks?
As with question randomization, having different questions will make collaboration more difficult.
What to watch out for
Test creation takes more time. Additionally, questions in a bank must be of a similar difficulty level, otherwise students may feel that the test is unfair if they happen to get more challenging questions.
Designing questions
In some situations, questions can be designed to deter cheating. For online exams that are not open Internet, consider including words that are commonly found on the Internet to make searches less useful.
Remote/online proctoring
For students taking exams remotely (i.e., not on campus in a proctored environment), there are ways to reduce the opportunities for academic misconduct. Examples of solutions include Lockdown Browser, Proctorio (no longer in widespread use at UBC), and Zoom invigilation.
Why use remote online proctoring?
Whenever a traditional exam format is chosen and students have an option to take the exam remotely, they may feel that they are less likely to be caught if they cheat. They may also feel that their classmates are more likely to cheat. Having tools in place that reduce opportunities for misconduct may make all students feel better about the assessment.
What to watch out for
Any software solution will require extra computing resources and additional bandwidth, and might therefore disadvantage students with older equipment and/or slower internet connections. These tools consume IT or instructor resources because they often require more people to support an exam in order to help with technology challenges students may face.
Cheating Detection
When using deterrent techniques is not enough, you can also rely on technology to detect and analyze submitted work to make an informed decision.
Plagiarism detection
Computers are very good at comparing large strings of text. Tools like Turnitin.com and Grammarly can take a document and compare it against a large database of sources to identify strings of text (phrases, sentences, paragraphs) that match existing sources.
Why use plagiarism detection?
It is an easy way to detect some cases of plagiarism, both copying from existing sources and from other students. Furthermore, if students know you will be using such tools, they may be more likely to submit original work.
What to watch out for
Students may also use these tools to make sure their work “passes” as original, even if it really is not. Also make sure you submit material to Turnitin.com without student names or identifying data, because Turnitin’s servers reside in the US and therefore are not FIPPA compliant.
Analytics
As with plagiarism detection, computers are good at finding patterns. Analytics options are becoming sophisticated enough to help flag suspicious activity.
Why use analytics?
To find possible instances of academic misconduct in exams or other similar assessments.
What to watch out for
A flag does not mean a student has committed misconduct, so anything that is flagged must be further investigated. Also, the software (i.e. in Canvas) is still in development and thus may not behave exactly how you want to.
Resources
