9

This question is related to, but (I think) distinct from, the recently asked Best day of the week to hold an exam?. In that question, the author asks for the opinions of other instructors and students. I am more interested in what the research has to say on the topic (if anything).

Question: In the context of relatively high-stakes, summative assessments, does the day of the week have a measurable effect on student performance? That is, what is the effect (if any) of holding exams on a particular day of the week? Has this been studied at all?

I am looking for answers which are supported by research, rather than anyone's gut instinct or experience (these things are valuable, but I would prefer to see evidence which supports the conclusions or recommendations).

One of the answers to the above-cited question references Pope and Fillmore, "The impact of time between cognitive tasks on performance: Evidence from advanced placement exams" Economics of Education Review, Volume 48, 2015, Pages 30-40, https://doi.org/10.1016/j.econedurev.2015.04.002 . However, this paper is primarily interested in the effect on AP exams, which seem (to me) to be somewhat different from regularly scheduled midterm exams in a college course.

I have spent some time looking for other articles on this topic but I have not found much that seems relevant. That being said, (a) I am not an expert in this area of education research (hence I am not even sure which journals I should be looking at) and (b) my institution does not provide me with much access to academic resources (we are a small community college and lack the budget for access to many journals). Further input from this community would be greatly appreciated.

7
  • 2
    If everyone picks the same day for exams, based on "evidence", then your goal will utterly fail if it includes making it possible for students to study effectively.
    – Buffy
    Commented Aug 30, 2021 at 14:47
  • The solution isn't the "day" of the exam, but the type of exam. You can also avoid "high stakes" exams, no matter the day. One option is to have a daily very short quiz at the start of a lecture. One or two questions. Students hate it, I think (I did), but you got in the habit of studying every day.
    – Buffy
    Commented Aug 30, 2021 at 14:51
  • 2
    What is the metric by which you define "best"? Do you want your students to get the highest scores? Your students to be the most relaxed? (E.g., after a weekend) Do you want to make it easiest for you to fit grading into the schedule? (Monday) Do you want to make it easiest for your students to fit the exam into their schedules? (Friday 7-9pm) There are so many notions of what "best" might actually mean. Commented Aug 30, 2021 at 17:09
  • 3
    Those of us who have taught large enrollment courses, e.g., 300 or so students, know that other professors also teach big classes and, like Mick Jagger said: “You can’t always get what you want!” Good luck trying to find a solution.
    – Ed V
    Commented Aug 30, 2021 at 18:27
  • 1
    @XanderHenderson Effect with regard to which variable? Exam grades? Student stress level? You will have to be specific about what effect you want to consider. Commented Aug 30, 2021 at 20:10

2 Answers 2

2

While there is little research on this topic, I did find this paper from SUNY-Oswego. The methodology simply entailed giving surveys to 204 students several times a day over 10 weekdays. They claim that "productivity is lowest on Friday's [sic]."

Clearly this methodology leaves a little to be desired (and the apostrophe use is abominable). Further, it provides at best only a partial answer, since it considers only affective states and neglects scheduling issues. But you asked for evidence based, and this seems to be what there is.

-1

With no research needed I'll declare (perhaps at some risk to my reputation here) that there is no solution. There is no best day. There are many reasons for this. First, is that if there were and every professor decided to use it, then it would automatically become the worst day since students would all get jammed up.

Second is that every student is different. Some have to work on week-ends (perhaps), some use it to study. Some to party. Some have a very heavy schedule, perhaps with report deadlines as well as exams. Every student is different.

I believe that there are a few universities that have a policy that every class has an exam on the last day of the week (there once were). This is "fair" at least, since everyone is in the same situation, though it isn't especially effective for actual learning. A slightly better system is to schedule major exams long in advance so that students that choose to have the opportunity to plan.

Another factor is the type of exam. Those that rely primarily on short term memory force some (many?) students to cram for the exam, losing sleep, perhaps, and doing worse than otherwise. Those exams that depend more on some competence learned only with practice and feedback will be less susceptible to such effects.

There are too many other variables than "day of week" that affect the individual performance and too many different student study techniques to make the question meaningful.

But, again, if there is a "best" day and everyone uses it, then it is likely to automatically become the worst day for many students.

My advice is to spread them out if possible, schedule them long in advance.

My more serious advice is to rely as little as possible on high stakes exams. There are better ways that actually enhance learning, not test-taking. And this is from a pretty effective "test taker" in my day.

Finally, optimize everything for learning, not for grading.


While it is true that this is just the opinion of an old codger who has taught something approaching 10,000 students, I think that is about all you will get for "evidence". The experience of teachers. The problem is that the number of variables that you would have to control to do a real research study with scientific validity is too large: field, course level, type of exam, student variability, ....

But if you simply want something that might have some, perhaps limited, validity for your own students, you could just ask them. I suspect you will get a variety of answers, but perhaps not (depending on all those variables). But if you also ask them "why" they prefer one day over another you might get an idea of the difficulty of the problem. And, of course, their own preference might be the deciding factor or not.

4
  • 4
    "With no research needed I'll declare..." -- leeway that we gladly provide to members of this group with a reputation of 214k ;-) Commented Aug 30, 2021 at 20:11
  • 1
    @WolfgangBangerth, thanks for that, but 40 years in the saddle at all sorts of colleges and universities provides the anecdotal background.
    – Buffy
    Commented Aug 30, 2021 at 20:15
  • 5
    I do feel it does not really answer the question asking specifically about something evidence-based. Your insights are overall valuable, though, and "optimize everything for learning, not for grading" is the real highlight here. Otherwise it's a severe XY problem - OP should be concerned with what makes them to teach better, not how they should grade things...
    – Lodinn
    Commented Aug 31, 2021 at 8:48
  • 2
    @Lodinn My posing of the question was somewhat tongue-in-cheek and insincere. I was rather shocked that the question I linked in my question here got such a positive reaction, when I considered the question to be nonsensical. Buffy's answer is (essentially) the one that I would have given. That being said, if there is actual research on the question, I would be interested to see it (since, as nearly every commenter here (though not at the other question) asserts, day-of-the-week is swamped by other variables). Commented Sep 30, 2021 at 2:32

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .