21

My family originates from Russia but I only have experience with American academia. My parents constantly describe oral exams whereupon a student is required to choose one of many cards that have a certain amount of questions on them and then after some time preparing be able to answer the question verbally to the professor, who can then ask follow up questions.

In my mind, this offers numerous benefits; it's much harder to cheat this system as you have to verbalise your answer so copying is not an option, verbalisation is shown to help with memory's retention, because you don't know what topic the teacher/professor will follow up with you have to more deeply to impress the professor, and many more.

The drawbacks seem limited to me; more people standing around waiting to do the exam isn't great, nor is the fact that this will be a bit more subjective to the teacher/professor.

Is there some specific reason as to why this isn't done?

2
  • 4
    Note orals exams are common in grad school in the US.
    – Kimball
    Apr 15, 2018 at 13:37
  • 1
    Much as it pains me to suggest that there were sound pedagogic reasons for the status quo ante pestillum (in the US or anywhere else), the Google Scholar search "oral examination" "validity" "reliability" does not paint a happy picture. May 30, 2021 at 17:38

7 Answers 7

23

An oral exam takes me about 30 minutes (including preparation and discussion afterwards). I have about 110 students in my class. Doing an oral exam for that class would take me about 55 hours. This is not the only class I teach. I teach about 5 classes per semester. In order to avoid some of the subjectivity my department requires that an additional staff member is present. So implementing this for my courses would take about 55*5*2=550 person hours. That is just not practical (not to mention that my colleagues won't thank me for the additional work I give them as they have to sit in for these exams).

I know how long an oral exam takes because I do use it sometimes, but for practical reasons I can only do that for a small number of courses and I have to justify why this is necessary to my colleagues.

6
  • 3
    For comparison, how long does it take you to grade a short answer exam generally? I feel like it would take a similar amount of time as administering an oral exam which would then be graded. Also, how do you choose when you do an oral exam versus a written one?
    – mettle
    Apr 13, 2018 at 18:19
  • 3
    If I'm reasonably efficient about it, a 2 hour written short answer exam takes me about 5 minutes per student to grade. Even that would be impractical if I had 500 students. Apr 13, 2018 at 23:44
  • 3
    Coming from a Russian academic system, I've seen oral exams organized for a cohort of 100 students. It required around 10 instructors, so that each instructor had to listen to 10 students, which is doable in 5 hours. A great deal of subjectivity was inherent to those exams, alas.
    – svavil
    Apr 14, 2018 at 8:42
  • 5
    @svavil I've seen oral exams organized for a cohort of 100 students. It required around 10 instructors... which is doable in 5 hours — In my department, each of those 10 instructors would be teaching their own cohort of 100-400 students, so doing what you describe fairly would require 100-200 hours, or roughly a continuous week without sleep.
    – JeffE
    Apr 14, 2018 at 17:44
  • 6
    @JeffE In fact, I have classes ranging from 50 to 120 students and I usually dedicate a couple of weeks to oral exams (8 to 20 students per day, depending on the class: for some classes I do a 20 min oral exam, for others up to 45 min - 1 hour). Apr 15, 2018 at 8:12
12

As others have pointed out, oral exams take a lot of time.

I'm not familiar with the US education system, but I'm in a country where, from the elementary school to the university, verbal tests and oral exams are quite common, and I've administered oral exams for about 20 years (with a duration from 20 min to 1 hour, depending on the class). I'll thus try to outline what factors, apart from tradition, allow to administer oral exams in a manageable way.

The most significant factor is probably the structure of courses and exams. In my country, there are usually several exam sessions in a year and students can take an exam in any one of those sessions. And, frequently, for very small courses, professors allow students to take the exam whenever they wish along the year. This means that if you have a course of, say, 100 students, 50 will probably take the exam at the first session, 30 at the second and 10-15 at the third session and the rest along the year (of course, at each session there's also a bit of backlog).

Second, for many courses, there are both a written and oral tests. Those who don't pass the written test are not admitted to the oral test and fail the exam. This means that of 50 students that try the exam at the first session, maybe only 20-40 pass the exam (in the past, just 10 would not have been uncommon), and this further limits the number of oral exams that you have to deliver each session.

Third, there is a certain freedom on how to administer the exams. Some professors will thus make the oral test optional: students who decide to not take the oral test cannot get a top grade (e.g. they can get a B, or equivalent, at most). And, usually, it is remarked that the oral exam can also worsen the grade. This, again, allows to further reduce the number of oral exams (in my experience, only about 10% of the students try the oral exam if it's optional).

In addition to the above points, there is also the acceptance of the stress of the oral exam and the subjectivity, or perceived fairness, of the evaluation (and a few comments here show that this is a controversial point, probably worth of a different question).

8
  • 1
    Very interesting. Based on your comment to another answer, another important factor seems to be that you're asking questions of varying difficulty, in order to challenge the students and understand how far they can go. Thus, it seems like oral exams can give professors a better idea of students' abilities at the top end, especially if written tests are more aimed at distinguishing passing from failing or if many students score near 100%. Apr 15, 2018 at 22:49
  • 2
    The schedule is a big difference, then. In the US system, a course typically lasts one semester (15 weeks), and final exams occur right after the end of the semester, usually all within 1-2 weeks. A permanent course grade must be assigned very shortly after that, so there is no way for a student to take the exam at a later date. And there aren't large blocks of unscheduled time during the final exam period, so it would be impractical to schedule all oral exams in that short period even if it weren't an excessive amount of work. Apr 16, 2018 at 2:23
  • How can students who take the exam 1/3 of the way through the class be adequately graded for the content that they haven't covered? Or is it only over the first third of the material? Do the students who take it at the final session get tested on the whole of the material or only the final third (to make a comparable breadth)? Jul 2, 2018 at 10:20
  • @guifa I don't understand your question: the exam is taken at the end of the class, not 1/3 of the way. There are no intermediate terms. And students get tested on the whole of the material. Jul 2, 2018 at 10:34
  • 2
    @guifa Prerequisites are checked only for attendance, so you can take the exams in whatever order you wish. Of course, if you take them in the wrong order, you'll probably have a higher chance of failing. Jul 2, 2018 at 13:15
7

I actually did that once, as a TA in basic computer science for CS majors. 90ish students, 10-15mins each group of 3 students, it took me the best part of two weeks, including days from 7am to 9pm.

I was only able to do it because I had a lot of "free" time as a TA (MS student, but I didn't have any courses of my own and my work was ahead of schedule).

NEVER AGAIN

Why? Because it takes an insane amount of man-hours (even if you do in parallel) and it doesn't inform you of anything you don't know already. After a while, it is fairly easy to spot the good students, the lazy but good, and the bad students. In that specific course, the marks for that oral exam were remarkably similar to the ones of a regular written exam I applied two months later.

Unless you want to see how the student behaves when "put on the spot", it serves no purpose.

2
  • 1
    I don't think I understand the setup you had. The numbers you give should only have resulted in 7.5 active hours of exams, so it is not clear what took all the time. Apr 15, 2018 at 16:43
  • 3
    The scheduling was not necessarily optimal (unused gaps between groups), sometimes it took more time per student, etc. I might have misremembered some parts, I'm sure of the number of students (default 2 groups of 45) and the weeks (the events are still on my calendar) but it was in 2007... Apr 15, 2018 at 20:41
7

Another aspect not mentioned (I am not sure if this is made concrete somewhere) is that oral exams do put students on the spot (they are in now way anonymous). This opens up the universities of being sued, if a student feels uncomfortable.

Even worse, a student could accuse the professor of some bad behavior in this situation, as a retaliation of a poor grade.

Grading a written exam can be done together with the TAs, and it is easy to get a second opinion, and spread the grading over several people, thus reducing the grounds for a student accusing a teacher for treating them unfairly.

0
7

I am going to speak solely from personal experience as a student in Russia going through 5 oral exams right now. In my university, the lecturer is usually just one of about 10 professors from the department in the lecture hall where we take the exam. These people most of the time have drastically different views on how you should explain one concept or another and what level of understanding deserves an A. This brings me to my main point - grades which are given during such an oral exam simply are not equal to the knowledge of a particular student. I have many friends who didn't deserve their grades and many who deserved much better. This, unfortunately, means that your luck, charisma, and sometimes your looks/gender determine your GPA and consecutively your further education.

5

Cheating can be a problem with oral exams just as much as with a written exam. The issue is foreknowledge of the questions, rather than the ability to verbalize the answers. So you’d be dependent on having a large stack of questions, which leads to a question of fairness. If you randomly select questions to answer, how are you sure that students across the class are getting roughly equal difficulty in the questions they’re supposed to answer. All of this has to go into the preparation time for the exam.

And none of this deals with the enormous time constraints of administering the exam to a large number of students in a class. While I have done such exams when I had 15 or 20 students in a co-taught class, I would never even consider doing it for a lecture course of 50 or more students.

16
  • 6
    It's 20 years that I routinely administer oral exams (actually, first a written exam and then the oral part, with classes of up to 100 students), and I have to disagree with this answer. First, you really don't need a large pool of questions because there are concepts that are difficult to master and explain even if you know already the question. Second, you really don't want to be fair in oral exams, that is, you want to challenge more, with the most difficult questions, those who can aim to top grades. Of course, in the assessment, you will take into account the difficulty of the question. Apr 15, 2018 at 7:41
  • 3
    @JeffE I beg to differ in the interpretation of fair: for me, fair means that to higher grades you get more difficult challenges, not that everybody gets the same type of questions. Otherwise, you have grade inflation. Apr 15, 2018 at 16:05
  • 2
    @JeffE This seems to be a cultural difference between the US and Europe, in my limited experience. In the US it would be outrageous to ask different students different questions, but it sounds relatively commonplace in some other countries. Jul 2, 2018 at 1:48
  • 10
    @StellaBiderman: Just to clarify: What Massimo is probably talking about and what I am familiar with is roughly as follows: Each candidate first gets a basic question from the same selection of questions and depending on the answer, they get either get further questions to weed out mistakes in their answer or proceed with a more advanced question. Every candidate is treated the same way, but it is interactive (which is one of the advantages of an oral exam). I have seen failed exams which ended at high-school level, whereas good exams ended at current research – for the same course.
    – Wrzlprmft
    Jul 2, 2018 at 5:20
  • 2
    And to further clarify: “Fair” doesn’t mean “identical questions” but “identical protocol for assessment” — everyone has the same opportunity for the same grades, and similar responses to similar questions yield similar outcomes. I would consider the protocol @Wrzlprmft describes to be completely fair, assuming there’s a neutral protocol for when to shift the difficulty up or down.
    – JeffE
    Jul 2, 2018 at 11:41
1

I'm teaching in the US system and, because of the pandemic, I started using oral online exams for my midterm and final exams. I find I can get an A/B/C/D/F decision on a grade within about 5 minutes, and a more nuanced grade within 10. This is approximately the same time as I spend grading traditional written exams (I'm a slow grader).

You must log in to answer this question.

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