75

A professor has a habit of starting oral examinations by asking for or looking up grade averages and grades from other subjects and then taking them heavily into account when evaluating the student in question. For example, I was allowed to be even considered to get an A in Real Analysis III, because I got an A in Real Analysis II and had a good overall grade average.

Is it ethical or helpful to ever consider student's performance outside the course? If not, what about subjects tightly interconnected? But then, most mathematics is interconnected in one way or another...

  • 76
    I don't think that's fair. I think grades should be based solely on performance from that class alone. That's what you're receiving a grade for. – Compass May 4 '15 at 15:40
  • 39
    Is it ethical to limit marks based upon past performance? — Simply stated: No, of course not. – Mad Jack May 4 '15 at 17:24
  • 14
    What country is this taking place in? Classroom culture differs a lot throughout the world. In the US, most major course practices should at least in theory be listed in the syllabus. If this manner of assessment is not listed in the syllabus, students have a right to complain. If it is, they have something to hold in their hand as they complain! – Pete L. Clark May 4 '15 at 20:24
  • 8
    I've had a half-dozen professors that made a point of not even looking at the name on an assignment/test while grading (or require it only be printed on the cover/answer sheet) so they could avoid such a bias! I've never encountered one who intentionally did the opposite! – BrianH May 4 '15 at 23:22
  • 14
    I tried playing devil's advocate and writing an answer of why it is ethical, but every argument I came up with had major flaws. This is definitely unethical. – Evorlor May 6 '15 at 17:06

10 Answers 10

143

What you describe is, in my opinion, horribly unethical!

Yes, past performance is often a predictor of present performance, but there are so many other factors involved as well. What if a student did poorly before because of any number of reasons, but have since stepped up their work, caught up, and really mastered the material? Or what if the student has been focusing heavily on this subject and has consequently fared more poorly in another subject?

It is manifestly stupid and counter to the entire notion of education to ignore good work by a student in one location simply because the student did bad work in another course or another time. Ethical violations are a different matter---it is reasonable to be suspicious of a student with a history of cheating---but grading a student poorly for having the audacity to exceed expectations is a rank betrayal of the most basic responsibilities of an educator.

  • 7
    Never mind that sometimes people have to choose a class to focus on, if they're all time-intensive courses. Sometimes grades fall simply because they had a really unbalanced semester. – Seiyria May 4 '15 at 19:03
  • 13
    Not to mention, a 'predictor' of performance does not mean it is part of that performance. – Zibbobz May 6 '15 at 20:02
33

No, I do not think that basing grades on past performance is fair. Grades should be based on the performance being graded, and it should be clear up front what exactly is being graded.

(Of course, if the syllabus explains that a grade will be based on homework and a final exam, then of course the grade would fairly be based on the final and the past performance in homework.)


One possibility for using past performance would be an oral exam. If I know the candidate sitting in front of me is very good or very weak, I might ask more targeted questions - focusing on basic definitions for a weak candidate, or on advanced understanding for a strong one, to avoid wasting everyone's time with questions that are too easy or too hard. If (!) the oral exam can then follow the natural development and the examiner adapts the difficulty of subsequent questions to the candidate's performance on earlier questions, then I'd argue that this kind of "customization" is acceptable.

  • This approach seems unethical for the same reason, as it seems to assume the current strength of a candidate based on prior performance. Wouldn't this be solved more conveniently by asking what is the grade you would like to receive and adjusting the difficulty accordingly? One disadvantage of this approach I see is that it could allow for an overly confident student to unnecessarily undermine his position, but it is still better than giving him no control at all make ill-founded assumptions. – mirgee May 4 '15 at 16:51
  • 6
    Or just start asking questions that are moderately difficult, then scale the difficult up or down based on the responses you get. – Kevin May 4 '15 at 18:50
  • 4
    @mirgee I don't think there's much of an ethical problem with Stephan's method, assuming that he rapidly moves on to more advanced questions if a candidate who did badly at Widget Studies I turns out to have mastered the basics of Widget Studies II. – David Richerby May 5 '15 at 9:14
  • 10
    @mirgee Stephan's suggestion seems to basically apply a sort of binary search to speed up the exam by efficiently estimating performance and readjusting as required. This is fine as long as the examiner doesn't fall into the trap of holding exceptional students up to a higher standard than is required for full marks. – Lilienthal May 5 '15 at 14:10
11

In general, what you're asking is an unethical practice. I could see some exceptions to this where such considerations would be reasonable:

  • You are taking a multi-semester course sequence, and the instructor is willing to base final grades on progression and improvement, rather than a strict numerical average. For instance, if you've improved from a B- to an A over the course of the year, your final grade in the second semester would be an A instead of say a B+ average.

  • You are being judged for something like a qualifying exam, in which performance in coursework over one's career can help to ameliorate poor performance on the exam itself.

  • 3
    Notice that all of the situations you describe as being ethical are ones that work in the student's favor, where the OP is describing the exact reverse. – jakebeal May 4 '15 at 16:10
  • 2
    Thanks for the answer, but just to clarify - the question was about considering performance outside currently graded "unit". – mirgee May 4 '15 at 17:00
  • 1
    There's a problem with your first point. What if the student gets first an A followed by a B- later? If you also give an A, then it doesn't reward improvement, but if you give lower, how can you say that it is fair if it could clearly be a case of variation in performance over time? Even if performance is relatively stable, of a large class there will certainly be a few whose observed performance is unstable. So your justification is not too good. Even if the exact way improvement is taken into account is known to the students, it would be unfair. – user21820 May 7 '15 at 2:57
8

I think this is unethical. All instructors are vulnerable to bias and subjectivity in grading and ought to do everything they can to minimize that; looking up a student's previous performance introduces a strong source of bias that was completely avoidable. An instructor who looks up a student's previous grade on an exam is going to go into the exam expecting the student to do well or poorly based on that previous grade, and people tend to be biased towards seeing what they expect to see.

I do understand why it's tempting for instructors to do this, though. I am a part-time instructor myself and have been a TA for many professors, so I see professors do things like this, and have been tempted to do it myself. Psychologically, I think the reason is that instructors are aware that grading is somewhat subjective, but that is also something that makes them uncomfortable. And they tend to feel uncomfortable if their grades look random: if people who did really well last semester do really poorly with this time, or vice versa, or if a student got an A- on one paper and a B- on the next and then an A on the next. Checking what grade a student got last semester, or on the last paper, feels like a way of double-checking your own grading, of getting a second opinion to double-check your sense of the student's performance on this particular task.

But the problem is that, far from being an effective way to deal with bias, the above procedure just magnifies its effects. The way it works in my experience is that if the student got a high grade last time, it's tempting to think, "Well, they're smart/a good student" and then give them the higher grade; if they got a low grade last time it's tempting to think "Well, even the lower grade is still better than they got the last time, so that's fine," or "This student is just lower-performing," and then give them the lower grade. And that's just not fair. But it feels secure and reassuring to an uncertain grader to know what other graders have given a person before assigning them their own grade, so unfortunately it happens.

4

I disagree with the notion that limiting marks based on past performance is necessarily unethical. If done without a valid reason (for example as a case of laziness on the professor's side), then it would very well be unethical, but there exist cases in which such a system is perfectly reasonable.

One reason that would justify the professor's system of grading would be one similar to the GRE's adaptive testing system.

If the professor knows that the student did poorly on his/her past course, the professor can then choose to administer an easier examination, which better distinguishes students of medium to low ability from each other. Similarly, if the student managed to obtain an A for the past course, the professor can then administer a more difficult exam to better distinguish students of relatively high ability from one another. Students who are given the "easier" exam would then not be able to achieve the maximum scores in the final grade.

This would be analogous to the GRE's system which uses the results of the first section to produce adaptively a second section which is scaled to an appropriate level of difficulty.

  • 4
    You're using "past" to mean "earlier in the same day". That's not at all the sense in which the question is using "past performance". – Ben Voigt May 4 '15 at 17:50
  • 3
    "This would be analogous to the GRE's system which uses the results of the first section to produce adaptively a second section which is scaled to an appropriate level of difficulty." No, it wouldn't. It would be analogous to having your GRE score depend on your previous coursework. Unless the link between a course and a past course is made explicit in the course syllabus, this policy violates a basic tenet of the university culture: that the assessment in a course is based on the student's performance in that course. – Pete L. Clark May 4 '15 at 20:19
  • 8
    So if the student did not manage an A in the past course it should be impossible to manage an A in this one?, and the next? oh and the one after that? Sorry you can never get an A again. – Mr.Mindor May 4 '15 at 21:11
  • 3
    A single test and an academic career are such vastly different scales, that what is reasonable for one may not even be remotely relevant to the other (as in this case). Refining a single test may be reasonable: There is no time or room to dramatically improve your ability between sections of a test and refinement allows for more detailed examination in less time/effort. Months, or even years may bridge the gap between courses, there is ample time for remediation of whatever problems lead to the poor grade. – Mr.Mindor May 4 '15 at 21:57
  • 4
    But even if the professor starts out with easier questions for people with poor past performance, he should make them harder and harder if he keeps getting good answers - so a formerly bad student might get an easier question at the start, but if he has learned enough to get an A, he will get to A-Level questions quickly and get an A in the end. That's how one of my profs did it and I thought it was entirely fair. – Sumyrda May 5 '15 at 7:30
3

This reeks of typical halfassed professory. Instead of working to create an assessment (ie grading) method wherein he can comprehensively determine how well a student understands the topic, this professor has willingly created and accepted a sub-par assessment which requires reviewing previous performance.

"Well, the student failed my test... however, on related courses they passed. Therefore I must have a flaw in my assessment or my course, so I should let them pass because they should have understood it and my methodology can't tell me whether they did understand it."

That's what I'm reading as the inner dialogue.

However, at the same time, I love it. This is a very realistic method for complex topics which are not readily conveyed and/or the understanding of which are not readily assessed.

And then I come back to hating it again, because that's the professor's job. Professors generally don't like the course work and teaching and testing... too bad. I never liked folding shirts when I worked at a retail store, and I never liked NBC warfare training during my military service, but they were part of my job. I didn't get a pass just because I "don't like" them.

So yes, it's extraordinarily unethical!!

Not only because it's not giving you the grade you earned, good or bad, but because the professor is doing this in lieu of a portion of the job they don't like/find hard. The university (and thus, you/your creditors) is paying them for a job they're not actually doing! That's unethical; that's fraud.

But it's quite clever and realistic, so maybe... :(

2

You should check your professor's grading criteria against his grading rubric - if 'prior courses' is not listed on that rubric, it's not only unethical, but duplicitous and could be brought up to the Dean if your grade or others' grades reflect poorly because of it.

If it's not on the rubric, you could still bring it up with the dean as being duplicitous, but there would be less actual evidence to bring against the professor, and since it's an oral exam, you'd be hard-pressed to provide proof without some type of recording of the exam itself.

If it's on the rubric, you could bring it up with the Dean and ask if it's really allowed for professors to do so, but if they say 'yes' then you're completely out of luck.

However, it is completely unethical regardless of what the dean says - you may have some paths of recourse for it, but regardless of whether or not those pan out, it is still unethical, and you have every right to feel it is unfair.

2

It's very unethical in my opinion. I quickly explain my way of handling things when I examine students:

Questions and answers for oral and written exams are set before the exam. For each question I expect certain key-content to be mentioned. If all is mentioned, it's full score for that question, if not, the score depends with mathematical precision on the amount of content mentioned compared to the amount expected.

After an exam I publish questions and key-content. Every student can check what he did wrong and how his grade was made. I find this the only fair way: If you know what's expected, then you have full score and you have the right to know what you did wrong.

0

Although I agree it's unethical, similar patterns are sanctionned by the academic world.

To play devil's advocate, maybe your teacher is adequatly preparing you to the way things work in higher education ?

One of the most important criteria to get funding, a bursary , a grant, a publication or even a TA job is if you had that same sort of thing granted before.

It's a vicious circle where if you've never had one of these, chances are you'll never get one and therefore probably won't have the extra "humfes" (jobs, contacts, money, opportunities) necessary to be in the top of your field.

When you think about it, it's what a cv is for. While looking at your profile, a prospective employer looks first at what you've done –IF you've done the same job you're applying for, your odds are improving.

  • A CV is for giving someone who's never met you a reasonable idea of your experience so that they having something on which to base their decision to even talk to you to see if you're qualified for their position. At your annual review, an employer is not going to say, "Well, you've done great work here over the last year, however I can't give you a raise because you had trouble at your previous job with another company and/or didn't have enough experience on your CV." Or, if they do, they're an idiot and you should find a better employer. – reirab May 7 '15 at 19:23
  • The CV situation is more analogous to saying, "You shouldn't take this course because you got a D in its prerequisite," not, "You can't get an 'A' in this class even though you've earned it because you got a poor grade in a previous course." – reirab May 7 '15 at 19:26
  • I also think that academia has more scope than you suggest to do extra things and build your way up even if you haven't performed strongly before. – Stephanie May 27 '15 at 12:33
-2

This scenario is ambiguous.there could well be and ethical abuse that the instructor is not doing his/her job, but the use of past performance as one input in the grade decision is not inherently an ethical abuse, depending on how one understands grades.

There is not widespread agreement about what grades measure.

Depending on whom you ask, grades might measure:

  • how much a student learned in a course
  • how much a student knows at the end of the course
  • how much effort a student put into a course
  • how well students complied wth course procedures
  • what a student is capable of doing at the end of the course

And these comparisons could be relative to a fixed standard, or relative to other students in the class ("on a curve").

All of the above have been used in courses I have seen. I'm afraid it seems that grades in a course mean whatever the institution or instructor define them to mean. There are some practices that would be widely condemned as abusive, but grading based on total proficiency and using past achievement as part of that assessment seems to be to be within the realm of defensible practices.

If the standard is an absolute level of proficiency in a body of material, past work in the field is a helpful piece of information about a student's likely level of proficiency. In other words, if a grade in course N is intended to reflect how well a student can do tasks x and y, and tasks x and y require skills from course N-1, then it seems reasonable to use this information.

This is not how I would advocate grading, by the way, but it seems to fall within the universe of acceptable practices. If a professor believed that the only way s/he were comfortable asserting that a student attained the level of proficiency s/he believed warranted an "A" grade were if the student previously demonstrated mastery at the "A" level of prerequisite course material, and also had high overall mastery of other material as demonstrated by a high GPA, this decision strikes me as unfortunate but not unethical.

I would hope an instructor would allow other ways to demonstrate that mastery besides grades in a previous class; this example was unclear whether that was the case. When the instructor told the OP that an A was only in consideration because of past performance, I do not know if this was a general requirement, or the specifics of this case based on other assessments in this class.

If the instructor views grades as a certification to the student's future instructors, employers, clients, etc of the Instructor's assessment of the student's proficiency with course material (far from the only meaning given to grades, but one widely accepted one), then this can be justified as ONE input in the grade decision. Given the set of observations about the student (exams, projects, homework, etc), how likely is it that the student has at least (excellent/good/fair/poor) mastery of course material? If the goal is to give the hugest grade the instructor an justify with confidence level x, past performance is an I out that adds information to the decision function.

If the instructor based grades ONLY on what the OP described, that would be shirking the responsibility of the instructor. If the instructor said that the student's work in this course alone was not enough to definitively convince the instructor that the student was at an "A" level of mastery of course material, but that work combined with evidence of the student's past work raised the confidence level enough to justify certifying to future encountered of the student that the student mastered the course subject, that could be ethical.

  • 2
    You ask a lot of questions in the second paragraph. But it seems to me that one could answer affirmatively to any or all of them and still not be able to justify a practice like "I was allowed to be even considered to get an A in Real Analysis III, because I got an A in Real Analysis II and had a good overall grade average." Clearly the OP is asking an ethical question, so answering "justice is the advantage of the stronger" has been known for a while now to be a not very satisfactory reply. – Pete L. Clark May 4 '15 at 18:27
  • This is an awful lot of words to just say "he might be speaking metaphorically". – Zibbobz May 7 '15 at 19:01
  • How do you get "metaphor" from what I said? – Daniel L May 7 '15 at 19:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.