Scalable Test Security
Deliver all your assessments on one platform – from practice tests to proctored certification exams
We mean that you can use a single assessment platform to deliver all types of examinations with a range of stakes: unproctored practice tests to educational benchmark tests to high stakes tests like professional certification or university admissions.
How is this done? Our platform build all the options you need into the test delivery configuration. You pick the various aspects that make sense for you, with a straightforward user interface – no paying expensive publishing fees to us!
The scalable test security can be divided into two aspects: control of the delivery/psychometrics, and control of the test-taker’s environment.
Randomized item sequence
Tests delivered in FastTest and Assess.ai have the option to randomize the order of items for every examinee. So even though all examinees might see the same 100 items, the sequence will be very different.
Randomized option sequence
Tests delivered in FastTest and Assess.ai have the option to randomize the order of answers for multiple choice questions, which makes them look different to examinees.
Linear on the fly testing (LOFT)
LOFT refers to exams that are built uniquely for each examinee, but equivalently. For example, you might have a pool of 200 items, and each examinee receives 100. Rather than selecting at random, you have the option to ensure there is content balancing, so that perhaps each examinee might receive 20 items from each of 5 content areas.
Computerized adaptive testing (CAT)
CAT is a modern assessment paradigm where each test is uniquely designed for each examinee, by adapting the difficulty to their ability. Because each test is unique, the security is greatly enhanced. Moreover, CAT tests are 50% shorter on average, which greatly reduces item exposure in an aggregate sense. There are a number of other advantages; however, CAT is only relevant for large testing programs, typically 1,000 or more per year, because it requires substantial investment and the use of a psychometric approach called item response theory (IRT).
Our lockdown browser, WebLock, prevents the examinee from doing anything other than taking the exam with their computer. It starts by scanning the computer for hundreds of known software programs for image or video capture, and blocking the test if they are found. Once the test begins, the test takes over the screen and the examinee is unable to open another tab to surf for answers. However, we do not know what is happening in the environment; the examinee might have a friend helping or have their phone camera capturing the screen on video. Therefore, it is more useful in concert with in-person proctoring, such as a group of 30 students in a lab, being watched by 1 proctor but with all computers locked down. WebLock cannot be used with remote proctoring services, as it would catch the video stream then kill it.
ASC has partnered with ProctorExam to provide Record & Review (R&R) proctoring services. Record & Review means that the test is not proctored by a live human, but video is recorded and reviewed at a later time.
Because there is no WebLock, the examinee is certainly capable of surfing for answers on another tab or recording video of the exam with their phone; this of course would be flagged during review and the examinee’s results invalidated, but this might be too late if the test content is already compromised.
ASC has partnered with MonitorEDU to provide remote proctoring with live humans in real time. This has the substantial security advantage over R&R that the examinee can be redirected, for example if the proctor sees them checking their mobile phone. There are no downloads, no automated emails, no API setup – just quick, easy, and secure exam delivery.