New series to help test providers protect against cheating and content theft
“An ounce of prevention is worth a pound of cure.” Benjamin Franklin, 1735
Posted by John Kleeman, Founder
In June 2020, an investigation found that many commercial pilots in Pakistan had cheated in exams. They qualified despite not having the technical knowledge they needed. At least one crash seems to have resulted from their deception. More reliable assessments could have helped prevent this.
Perhaps this seems like an extreme example. Surely cheating in most tests or exams is unlikely to carry the same consequences? But test fraud, the term given to any activity which is against the rules of the test or exam, can have serious repercussions. People make decisions based on the results of tests and exams. Test fraud distorts those results potentially allowing for wrong decisions.
When cheating takes place at colleges and universities, or in certification and credentialing, it distorts the quality of the testing and undermines the value of the qualification. When test fraud occurs during workplace assessments, it leads to employers receiving inaccurate information on the skills, knowledge and work-readiness of their staff. An employer might promote or hire the wrong person. A team leader could place a staff member in a situation where they are a risk to themselves or others.
Test fraud can undermine the integrity of an institution, the decision-making process of a business and even the safety of the public. We need to tackle it. And it is better to prevent it from happening than to try and deal with its consequences.
That’s why we’re launching “Investigating Test Fraud”. A new content series to explore how people are attempting to undermine tests and exams and what test-makers can do to tackle it. This blog is an introduction to the series.
The different kinds of test fraud
Test fraud usually appears in one of two forms:
- “Cheating” – where one or more test-takers engage in activities that break the test/exam rules to gain a higher score. Almost every test program will want to resist and reduce cheating.
- “Test theft” – which involves someone stealing, copying or otherwise obtaining questions or test content illicitly to help others. This might not matter for all test programs. For example, some driving test programs publish all possible questions publicly. But for many it can be detrimental to the integrity and the perceived value of test results.
What are the different kinds of cheating?
- Obtaining the questions in advance, giving the test-taker an unfair advantage.
- Receiving expert help while taking the test from a friend or an expert or even the instructor or test proctor.
- Using unauthorized test aids such as searching online for the answers or using a calculator when forbidden.
- Using a proxy test-taker – when someone pays a friend, colleague or someone else to take the test for them.
- Tampering with results after the test is complete.
- Copying answers from another test-taker.
What are the different kinds of test theft?
The International Test Commission (ITC) guidelines suggest the following main ways in which people steal test content:
- Stealing actual test files, e.g., from the computer system used to deliver the test.
- Stealing questions by using digital cameras or other copying mechanisms.
- Stealing content by recording electronically the whole testing session on a computer.
- Memorizing test content and transcribing it later.
- Speaking into a phone or recording device during a test, reading out and so recording the test questions.
- Obtaining test material from a program insider, i.e. an author shares content illicitly.
Some unscrupulous companies seek to harvest questions from widely used tests and sell them to candidates. For other test programs, there can be a culture of friends or insiders passing on questions to test-takers to help them.
Risk, recourse and the route to reliable assessments
Throughout 2022, we will release a series of content exploring the different ways that test programs can prevent fraud and protect the validity and reliability of their assessments. Crucial to tackling bad behavior is to understand that not all test fraud is equally damaging. It’s important to focus efforts on reducing risk and likely impact of fraud.
Take for example, the case of a test-taker facing a science exam who writes formulas on their fingernails to help with the test and so may obtain a marginally higher score. This is certainly test fraud. But the impact on the integrity of the exam is minor. Some effort should go in to preventing it. But not a huge amount. Requiring test-takers to have short nails or inspecting everyone’s nails carefully before taking a test would seem disproportionate.
But what about a situation where a test-taker might hack into a computer system and change their grade? If a system can be hacked, can any of the results be trusted? The impact of such test fraud would be very serious, and it is certainly worth increasing security measures to ensure that this is close to impossible.
The below diagram helps quantify the risk. The least important ones to address are the ones with low probability and low impact (shown in green). The middling ones (shown in yellow) are worth considering, and those highly probable and impactful (in red) need prioritizing.
Conclusions and takeaways
- Test fraud consists of “cheating” where one or more test-takers illicitly seek to get a higher score than they should and “test theft” where question content is stolen to help test-takers
- Test fraud has the potential to significantly reduce the value of tests and exams
- To plan effectively to reduce test fraud, it’s useful to consider risk and drive efforts based on risk impact and probability