JR has a headache. He’s been fighting off a small cold the last few days - nothing serious enough to go to McKinley, but bad enough to make life miserable. Unfortunately, today is the final exam. There is no choice but to take some aspirin, get some caffeine in him, power through the exam, and hope for the best.
I feel bad for JR in the story above. Final exams often account for a significant percentage of a student’s grade. Still, because they happen just once, situational factors like a headache can significantly impact the final grade. That doesn’t seem fair.
There are several reasons to prefer frequent, low-stakes assessment over one or two big exams.
- A single big exam risks taking an unrepresentative sample of each student’s performance – they may be performing above or below their usual standard on the day of the exam.
- If one or two big exams comprise the bulk of a student’s grade, students may be tempted to put off studying until shortly before the exam. Even if they cram and perform well on the exam, they will likely forget the material they studied at the last minute. Better to set up a system where students are working with material consistently throughout the semester.
- A high-stakes exam that counts for a large chunk of their final grade may tempt students into cheating. Obviously, students should not cheat regardless of how you weigh the exams, but consider lowering the stakes for each exam, so the risk of cheating outweighs the potential reward.
- Some students are better than others at the format of exams. Basing the bulk of a student’s grade on one or two exams risks measuring how good students are at exams, which may not be the same as measuring if they learned the material.
At a minimum, you could take the questions you would ask in the big exam and divide them across several smaller exams or quizzes. This should encourage students to study and practice throughout the semester and lessen the chance that a student’s final grade is based on one unrepresentative performance. As a bonus, you aren’t stuck with a huge pile of grading right at the end of the course.
Even better would be to incorporate other forms of assessment throughout the semester. Consider whether short writing assignments, projects, presentations, or other assessments might just as well measure student learning without disadvantaging those students who aren’t good at the testing format. Plus, out in the real world, your students are more likely to give presentations and write than take exams. Many of the other cards on the student assessment page [link to student-centered assessment homepage] can help you choose interesting and effective assessments for your students.
Bonus idea: if your exams involve problem sets, consider including a few questions where they set up but don’t solve the problem. Maybe you aren’t primarily concerned with whether your students can perform calculations but are more interested in whether they can select the right formula, apply the appropriate principle, or identify the correct variable(s). If so, don’t make students spend time crunching numbers – use the time to assess what you really care about.