What is the point of a timed exam?

Is there any utility to determining who is the fastest test-taker?

If you have paid a lot of attention to my thinking about tests, you know that I find them to have absolute utility for helping learners learn, and almost no other utility whatsoever in the learning process. This is at odds with the testing culture of American-style education, which uses tests as much as sorting devices for students as it does for anything even approaching useful testing pedagogy.

One of the joys of moving to Fancy International School is that I no longer have to pay attention to almost any standardized or otherwise time-limited testing experiences for my students. I still give tests, but I have virtually unlimited leeway to give students as much time as they need when testing to recapitulate their knowledge to the best of their ability. It's incredible how good the ability to not care about test time-pressure feels when coming from 14 years of American publication education, where time-pressured tests are functionally devotional in their placement within a course of instruction.

There is only one time-pressured exam in my FIS life— the AP Biology Exam. The exam is a three-hour endurance fest of multiple choice questions and free response "essays." It's a beast of an exam and has been ever since the College Board redesigned it in 2012. Its beastliness is not universal for all students. The quickest readers and writers in the room are going to be able to finish it and feel like they have completed it on a level that comes pretty close to their best effort, as long as they haven't dawdled for too long during the experience. But many students are going to feel less-good about their effort because they will get caught out; their reading|processing|thinking speed is not as fast as it needs to be to complete the thing comfortably in the time they are given.

Until this past year, I hadn't really focused on the time issue. I'm not sure why that is, but I think that in my pre-FIS situation, many of my amazing students were hitting the time-concern but also had other issues related to the exam that commingled with the timing thing. As a teacher, if a student indicates that exam timing is a problem while also demonstrating knowledge|skill gaps in various other contexts, it's pretty natural to focus on the later. This year, my students generally had fewer knowledge|skill gaps, and so I think that the timing issue became a more common sole-source of exam-related grief. This led to some pretty lame "exam training" that I found myself doing, which mostly revolved around speededness of taking the test. As an educator, I didn't love that I was doing this, but as a teacher of this group of kids, it was what they needed to feel like they had as good a shot as possible to demonstrate their best thinking on this over-stuffed exam.

All of which is why I found these two recent episodes of "Revisionist History" so interesting. Malcolm Gladwell frames a discussion of this paper on the role of speededness in LSAT performance through a gimmick-attempt to do better than his research assistant on the same exam. Among the most resonant points are the discussions that he has with test-prep experts, who proffer a variety of strategies for LSAT success that sound a whole lot like the kinds of approaches I was teaching my own AP Biology students when dealing with their exam. All of which leads me to wonder:

Do time-limited exams serve any actual purpose in the learning process?

I've felt for a long time that the answer to this question is "not a lot," but these days, I think the answer is "not at all." I'm not sure what the role of a time-limited exam is, except to select out those students who are the best at taking time-limited exams. And I'm not sure there's any research that suggests that the quickest students at doing well on a subject's time-limited exams are the students who best learn the subject. Certainly, there is some degree of overlap, quite possibly even a large one, but the suggestion that there's a causal relationship does not seem to be supported by anything, even at the level of the LSAT. What is supported is the notion that placing a large value on time-limited exams in determining a student's grade in a subject is pretty damaging for those students who don't do well on time-limited exams.

Of course, what is time-limited is the amount of time a course meets, and that includes the duration of testing periods. So even if you agree with me that time-limited testing is a stupid thing to do to students, we still need to be able to administer assessments within the constraints of the school environment. How do we accomplish this? I have some thoughts:

1. Design any exam to be shorter than the time you have allotted for it. I think a good rule of thumb here is to design your exams so that they will occupy approximately half of the allotted testing period for approximately half of the students in a "general" education level course. The designed length can probably be increased as the level of the course "advances," as long as you hew to rule #2 that follows below. The point of doing this is to make sure that any unavoidable timing considerations for your exams have as little effect as possible on the ability of your students to complete the exam without feeling rushed.
2. Never take an exam away from a student who hasn't finished it. Let's say that even with your best efforts to be mindful of designing an exam that can be completed by all of your students within the time you have to administer it, you still find yourself with a student or two who have not finished the exam at the end of the allotted time. This happens. And when it happens, don't take the exam from the kid at the "end of time." Or if you have to take it (to avoid burdening a colleague with a late student), set up a time for the kid to come in to finish the exam. If you don't do this, you're going to be grading an exam that isn't actually reflective of the student's understanding; it's reflective of the ability of that student to demonstrate their understanding within an arbitrary time-period. Unless somehow your course is explicitly constructed so that speededness is a skill you are working to develop, this is not something you should be evaluating.
3. Balance the impact of assessments on a student's grade. Grading has its own host of issues, but I'm not going to get into any of that here. Let's assume you need to generate a grade for a student, like the vast majority of teachers do. If that's the case, then don't give any single type of assessment undue influence on a student's grade. This isn't just good practice for dealing with time-limited exams. It's good practice more broadly.
4. Train for unavoidable standardized tests with compassion. Even for all of the above, your teaching situation may be such that at the end of your course, your students are going to need to take a time-limited exam. In that case, it's not wise to totally ignore the reality of the thing. You definitely want to give your students the kinds of skills they will need to do as well as possible on that exam. This is going to involve training them in all the sorts of exam-taking skills that they will need. But even so, you have the power to do this training compassionately. You have the ability to decide that you aren't going to use the exigencies of a standardized test to determine your in-class testing protocol. And you have the authority to let kids know just how terrible an idea it is to test in the way that AUTHORITY X is going to require them to. Test-taking skills definitely have a place in a class that is going to end in a time-limited test, but nothing requires us as teachers to give that place more prominence than it's worth.

These are a few of the ways that I think teachers can navigate the reality of having to give time-limited tests even while knowing just how terrible an idea time-limited tests are for assessing the extent of a student's knowledge. I'm sure there are others, too, but I haven't thought of them here. One thing I do think of a lot these days is just how unfortunate it is that we have decided that time-limited exams are anything worth using when trying to establish a picture of student knowledge. That seems like a shame, but in the era of needing to do more with less in education, it's also completely understandable how we got here. It's hard to think of something easier to use than a time-limited exam when measuring student learning. But as is often the case, the easiest things don't always work the best.

Sometimes, they barely work at all

Have I made a convincing argument against time-limited tests? Or maybe my commentary has roused you to their robust defense? Leave me a comment or drop me a line if you yearn to let me know. I do not check social media.