Tests are and have been the standard for measuring ... what? Learning? Skill? Knowledge? In the past, tests and testing maintained this role as the standard for assessing a variety of states of students’ being. However, the disruption brought on by COVID-19 has caused us to re-evaluate the role tests plays in education.
At my department at Drexel University, there has been some debate about the function of two different types of tests and how we might adapt and accommodate them. First, is the role of the Physics GRE and, secondly, the general GRE. In the US, these two tests have been the standard for the graduate admissions process in physics. Similar subject specific tests exist for other disciplines and play similar roles. Recently, a movement among academics across the US, with the accompanying hashtag #GRExit, has pushed departments to drop the subject specific test (and in some cases the general GRE) as a requirement for admission to graduate school. The reasons for dropping these tests are myriad. Inherent bias in the tests is one reason. Departments that use cutoff scores as part of their admissions criteria, limit access to graduate study for under-represented students. Simultaneously, these tests fail to predict the success of those students completing doctoral degrees (Miller et al., 2019).
Since the COVID-19 outbreak, a number of departments have announced they will be suspending the requirement for the GRE and subject specific GRE, as it was onerous for departments to expect students to complete these tests, especially during a pandemic. Not only are the tests expensive to take ($205 for the general test, $150 for subject area tests, plus an additional $27/school if you want to send the scores to more than four schools), they often require students to travel to examination centers to complete these tests. So as a consequence, departments have decided they will use a variety of other measures to evaluate students for graduate programs instead.
Concurrently, one of the challenges that has arisen across college campuses is that while instruction can be moved online, assessment is a bit more difficult. This is because the backbone of the assessment of students, particularly in the sciences, has been through tests. Faculty are wrestling with developing testing procedures for online teaching, with some involving elaborate proctoring methods or relying on student codes of conduct and academic integrity. However, faculty are finding websites like Chegg or Yahoo Answers present a challenge because they post solutions to many standard problems or questions online. One professor who searched for a problem prior to including it on an exam, could not find it on Chegg and happily proceeded with the exam, only to find about halfway through the online exam that the problem and solution had been published to Chegg. Other professors have gone further, developing erroneous solutions, which they then post to Chegg and then penalize students for cheating when the erroneous solution is used.
To me, the disruption caused by COVID-19 has forced me to reconsider tests. What do they measure and what are they intending to measure (and how well do they match up)? I have decided in my classes not to include tests. Instead I’m applying standards-based grading and focusing on assessing student writing and presentations, which are harder to copy.
The overall challenge of exams and testing is they are intended to measure something. They are generally presumed to measure learning or knowledge but in cases where students can look up problems and solutions via Chegg, what are they really measuring? The GRE and Subject GRE are supposed to measure preparedness for graduate study. But if there is no relationship between the student’s scores and their success in graduate school, then what are these tests assessing? And will departments successfully identify candidates worthy of graduate study in future, without the need to resume these expensive tests?
Miller, C. W., Zwickl, B. M., Posselt, J. R., Silvestrini, R. T., & Hodapp, T. (2019). Typical physics Ph. D. admissions criteria limit access to underrepresented groups but fail to predict doctoral completion. Science advances, 5(1), eaat7550.
 For a more robust description, see https://ericbrewe.com/post/standards-based-grading/