Rethinking exams: Do universities need a smarter balance in their assessment models?
Over the past couple of years, I have pursued two undergraduate degrees concurrently: an entirely in-person programme at a private university in Bangladesh and an entirely online degree at a university abroad.
At my university in Dhaka, everything builds toward the midterm and final, while the online degree breaks things up into a steady rhythm of smaller tasks: fortnightly tests, short projects, and ongoing participation, without any midterms or finals. Living with these two systems side-by-side and observing how each one shaped my study habits, stress levels, and ability to learn convinced me that the question isn't which system is "better", but which pieces of each system we should keep, discard, or combine.
In our part of the world, high-stakes exams are familiar for a reason. Everything from school to university entrance exams, as well as exams in universities, is a high-stakes, singular exam that provides a compact, standardised snapshot of performance.
For instructors and administrators managing large class sizes, this efficiency matters: grading, quality control, and comparisons across batches are simpler when assessment is standardised. A high-pressure exam can test a student's proficiency in certain skills, such as fast problem solving, timed reasoning, and crisis decision making. How a student performs under constraints is indicative of their abilities. Although this method does involve a degree of clarity, it comes at a cost.
When a large fraction of a course's grade rides on one or two tests, it pushes students towards short-term cramming and rote memorisation. I myself have pulled all-nighters memorising and learning for a final, only to forget it weeks later. In addition, high-stakes exams also magnify anxiety. A single bad day, illness, or family emergency can eclipse an entire term's worth of work.
The continuous-assessment world feels very different. In my online course, assignments and quizzes arrive like steady drumbeats. Each test carries a small weight; therefore, the pressure of any single deadline is low. Frequent tasks give students opportunities to correct mistakes. A poor quiz becomes a moment to learn, not a catastrophe.
For students who struggle with test anxiety or need structured repetition, this system is humane. Adding to this, testing small bits of learning rather than having to internalise a huge chunk of the syllabus feels more manageable and cements better learning.
This model, nonetheless, brings its own set of pitfalls. Without careful scheduling, it can become an endless parade of deadlines. Consistency is also way more important. For people who prefer to learn in short, powerful bursts, rather than maintaining a consistent moderate effort, this can lead to tasks piling up, especially if the later course content builds on earlier concepts.
From my experience, the best system is a mix of both. For example, the curriculum could opt to include a smaller final exam or a project at the end of the semester to bring everything together, coupled with regular, low-stakes quizzes and short assignments along the way. These little check-ins should be there to help students practise and learn, not to punish them. And they only work if students are provided quick feedback. Teachers should be able to drop the lowest quiz or two, because life happens, and no one should fail because of one bad week. Deadlines across courses should be coordinated to ensure that students don't drown in overlapping submissions.
Most importantly, university assessments need to be more equitable. Continuous assessment can be a huge help for students, but only if it's done with clear instructions, flexibility, and compassion. At the same time, teachers also need support in the form of proper training, realistic workloads, and the right tools.
Ultimately, the goal is not to abolish finals or to make students live under perpetual assessment. It is to align assessment with what we want graduates to do: solve complex problems, integrate ideas, and keep learning beyond exams.


Comments