Overthinking the basics: timed exams

Like most others in higher education, I’ve responded to students’ pandemic stress by figuring out how to do things more flexibly over space and time. For example, this Spring 2022 class can be attended in person or remotely and also watched afterwards, and similarly my midterm exam could be taken either in the classroom (and handed in on paper) or at home (and scanned to PDF then uploaded). These accommodations all require some technical ingenuity and a lot of time and attention, rethinking and perhaps overthinking the basics of higher education — like timed exams.

For students to take economics exams at home without supervision, I switched to an open-book style but kept the time limit which is our standard class session (90 minutes) for the midterm and then twice that (3 hrs) for the final. But my style has always been to enforce time limits only loosely, and some students who took the exam in the classroom kept writing for about 15 minutes past the time limit.

Then today (the Sunday after the exam), I received this note from a student:

I finished the midterm at home, but there were so many questions that I didn’t finish all the questions in the time frame. And according to some in-person students, in-person students were given extra time to answer the questions, but those who took the exam at home did not receive an email about the extended time. I would like to ask if there is any remedy for me to improve my midterm because I think I can finish all the questions of the exam if I have enough time.

That prompted a bout of overthinking… and this reply:

Thanks for writing about this – the rush of writing at the end of an exam certainly feels very important, and it’s good to ask about. I’ll cc the TA team since they might be interested in this question. 

First, the duration of each exam is indeed a significant part of the experience.  If you were to look again at your exam paper now you would notice a lot of things about what you wrote that are influenced by being rushed.  But the differences in total elapsed time among students don’t actually make any practical difference in scoring, I can assure you.  In my experience, the test could have a wide of durations and would give pretty much the same distribution of scores.  For example if we had a fire alarm and the room were evacuated after 20 or 30 minutes, I am pretty confident that the distribution of students’ performance on those fragments of an exam would be very, very similar to their performance after 90 minutes.  And the same applies to a random change in timing at the end:  if I were to announce a 15 or 30 minute grace period to extend time at the end, the distribution of scores would be unchanged. 

So, why have a timed exam, as opposed to a take-home exam of indefinite length?  When you prepare for an exam of this duration, you devote your prep time to practicing a demonstration of your economics skills at that level of depth — and then spend that amount of time actually doing it during the test itself.   After the exam, students usually forget the specific content pretty quickly.  But the experience of learning and then doing over a sustained period of time builds up your skillset in a more lasting way.   So the basic idea is an exam whose duration is a reasonable amount of time for in-depth thought and quick performance for this particular kind of work.  The content definitely matters.  In this case, it’s the analytical diagrams used in economics.  But there is a more universal aspect of timed exams… In my view the duration of each exam (for us, a 90 minutes midterm and then a 3 hour final exam) is designed to achieve goals like these:

1. To help perfectionists practice working faster and then walking away, so they gain the ability to get more stuff done fast even if it’s not perfect,
2. To help build up the stamina of people with shorter attention spans, so they gain the ability to concentrate for longer periods of time,
3. To normalize the duration of an intense work session around the typical lengths of time that people can focus before needing a break,
4. To experience aha moments of discovery, when you see connections between disparate things revealed by doing them in quick succession,
5. To experience flow that comes from a sustained period of concentration during which you build new kinds of capabilities.

In your case, maybe goal #1 is actually the most important one, but I think everyone can gain along all of these dimensions to some extent.   Also I would flag that beyond test duration there are many, many differences between taking an exam at home vs. in a classroom.  The test-taking environment can make an important difference, for example table size to spread out things in your field of vision, snacks & drinks if you want, presence of other people and mask-wearing if you find that distracting, etc.  People differ a lot in their needs and preferences, and their situations.  The time difference is the least of it, and this whole email is surely more info about test duration than you probably want to know.  But it is interesting and important to think about, and might be helpful to you personally!

All the best,

Will