123

Example:

Q: Does Venus exhibit retrograde motion? (1 mark)

A: No. This is because Venus orbits the Sun and not the Earth.

The first part is correct: Venus does not exhibit retrograde motion. But the explanation is incorrect: the reason Venus doesn't exhibit retrograde motion is because it's closer to the Sun than we are. Mars for example also orbits the Sun and not the Earth but does exhibit retrograde motion.

Do I award 1 mark or 0? On the one hand, for obvious reasons, the grading scheme only covers whether the student said "yes" or "no". Based on that, I should award 1 mark. Further, if the student hadn't written the incorrect explanation, then the answer is perfect, and it feels wrong to penalize the student for going beyond what the question asks for.

On the other hand, the explanation is clearly incorrect and the student should've known the correct explanation (it's part of the curriculum). It also feels wrong to award full marks for semi-incorrect answers. For example, if the student had written something silly such as "This is because Venus is made of Swiss cheese", do I still award 1 mark?!

Ideally, I'd award 0.5 marks, but for various reasons, fractional scores aren't permissible.

Anton Menshov
  • 6,030
  • 4
  • 32
  • 54
Allure
  • 127,528
  • 50
  • 325
  • 493
  • 1
    Answers in comments and discussions about astronomy have been moved to chat. Please read this FAQ before posting another comment. In particular, all further comments about the astronomical accuracy of the example will be deleted. This really belongs to chat or on [astronomy.se]. – Wrzlprmft Jan 24 '19 at 18:45
  • I'd think usually this would be all part of some list of true/false questions. So did the student give a justification for each question? – BCLC Jul 04 '22 at 15:28

17 Answers17

284

You asked for a yes/no answer (which, as you've discovered, has its disadvantages) and got one plus some other stuff. You should grade the yes/no answer and ignore the other stuff. If you like, you could add a note like "You got lucky! This is actually because..."

The whole point of yes/no or MC questions is that you grade only the answer, and assume that type-1 and type-2 errors cancel out or are normalized out. That paradigm doesn't work if you don't uniformly ignore everything other than the answer.

More concretely: other students likely got this question right using the same incorrect reasoning, but didn't write their reasoning down. There is no way to identify these students; so, you need to make sure they get the same score as this student.

cag51
  • 67,924
  • 25
  • 181
  • 247
  • 158
    +1 for "other students likely got this question right using the same incorrect reasoning... [but you can't] identify these students" – J. Chris Compton Jan 23 '19 at 15:22
  • 25
    I'd say "the whole point" of any test is to grade a students understanding of the material. It was decided ahead of time that a yes/no answer was good enough to determine that understanding, but now this student has given us clear evidence that they do not understand something. That makes the answer wrong. As far as the other students, all we know is that they got the right answer. They presumably understood. They also understood the question didn't ask for further explanation. – JPhi1618 Jan 23 '19 at 19:32
  • 9
    The last paragraph is pretty sharp. – Randall Jan 24 '19 at 02:56
  • 2
    Because there may be more students with wrong reasoning, if I were an instructor I would take a copy of that answer into the next class and tell the whole class that this particular reasoning is wrong and what is the correct one. – Alexander Jan 24 '19 at 11:21
  • 12
    @Alexander I wouldn't take that specific example (or say 'one student has put...'), because it effectively embarrasses or shames that student in particular (who went above and beyond to explain why), when more than one could have the understanding wrong, perhaps for differing reasons. It's simpler to simply (re)introduce the topic of why Mars is retrograde, and if asked why, simply say that based on the test results you feel students need a refresher. – SE Does Not Like Dissent Jan 24 '19 at 14:29
  • @JPhi1618 Why would you presume other students understood when the only insight into the student's thinking is indicating they didn't? – Frank Hopkins Jan 25 '19 at 19:55
  • @DarkwingI was trying to counter the argument of other students likely got this question right using the same incorrect reasoning. My point is you can't infer anything about the other students. They got the answer correct based on the instructions - The End. However the student in question certainly doesn't understand the question and therefore got the answer wrong. – JPhi1618 Jan 25 '19 at 20:02
  • 1
    I agree with giving a whole-point for the incorrectly reasoned answer, but suggest adding another important reason: you don't want to give students a reason not to provide such an explanation, because that reduces your awareness of their knowledge/reasoning and hence your ability to help them. Of course, you could go the whole hog and ask for explanations on every question and would then have much more such insights to work with, but the MC question is a practical time-saving compromise. You can still circle the explanation in red, write "nope" or a longer explanation if you have time. – Tony D Jan 26 '19 at 03:51
  • @SSight3 - why should an anonymous student get embarrassed? Being taught how to answer questions is as important as getting facts right. Make a point of it in class! – Tim Jan 27 '19 at 12:56
  • 29
    @JPhi1618 : You say "That makes the answer wrong." This is incorrect. It makes "that a yes/no answer was good enough to determine that understanding" wrong. Since the defect is in the assessment method, the corrective action is to improve subsequent assessments. – Eric Towers Jan 27 '19 at 16:34
  • 4
    @JPhi1618 That approach would be extremely unfair, and bordering on dishonest. A multiple-choice exam (of which this is a variant) is structured with an implied promise that a student who gets the right answer will get a mark and a student who gets a wrong answer will not get a mark, regardless of whether they understand the reasoning or not. If the examiner wants to be able to grade the student's understanding then they choose an exam structure which demands more than just a MC answer. It is dishonest to trick a student into thinking you have given them a MCQ and then grade them differently – JBentley Jan 27 '19 at 20:27
  • 3
    @JBentley, I like your reasoning, but no one "tricked" this student. The directions were clear, they did not follow them, and they ended up writing down the wrong answer because of that oversight. If you were the student, could you honestly argue that you should get credit for this answer, knowing that you either didn't follow the directions or you didn't understand the material? – JPhi1618 Jan 28 '19 at 18:21
  • 1
    @JBentley 1) The OP did not state that this was a multiple choice exam, just that one question happened to be a "yes/no question" 2) The question does not actually say "answer only yes/no". IMHO as a student seeing a question and a line to write on I'd expect to have to justify my claims hence I disagree with OP opinion that that question of the exam is "just a yes/no question". 3) What other students might have done is irrelevant to what you certainly know of a given student. 4) Adding something when not required comes with a chance to get a bonus and a risk to get a malus. – Bakuriu Jan 28 '19 at 20:04
  • My two cents: when OP gave a T/F exam, they accepted that some students would get credit when they didn't deserve it. Even randomly guessing will lead to 50% accuracy -- but after normalizing and averaging, students get approximately what they deserve. While it may be justifiable to penalize students for providing unsolicited, incorrect reasoning, this will create a lot of rancor (which you may or may not find to be justified, based on the factors discussed on this page), and will harm the exam's ability to provide an apples-to-apples ranking of the students' understanding. – cag51 Jan 28 '19 at 20:48
  • @JPhi1618 Yes, I could argue that - "the grading scheme only covers whether the student said "yes" or "no". (from the OP's description). In the majority of cases MCQ grading schemes will direct the marker to ignore anything that the student writes in addition to the answer. – JBentley Jan 29 '19 at 00:41
  • @Bakuriu I disagree. As per my comment above, the grading scheme only covers a yes/no answer, and as a student, I would expect this because the question states that it is worth 1 mark. Questions demanding explanations usually give more marks than that. I strongly disagree that "adding something when not required comes with a chance to get a bonus". That directly contradicts the fact that it is a 1 mark question. Giving bonus marks where the grading scheme does not provide for it is unjustifiable. – JBentley Jan 29 '19 at 00:42
  • 1
    @Tim Just because the student who made the error is anonymous to the class in terms of exam doesn't mean you should isolate them as the sole example of an error, because they will know what they put, and will feel such remarks are directed at them (whether or not you know it's them is irrelevant). Given it's unlikely they are the only ones to have made an error (not necessarily the same kind of error), highlighting that specific error is redundant; you're teaching an entire class, so just highlight the correct answer. The student is smart enough to work out for themselves it was an error. – SE Does Not Like Dissent Jan 29 '19 at 13:35
  • @SSight3 - what I mean, actually, is highlighting to everyone that it wasn't the best way to answer, regardless of who wrote it. That part is inconsequential. No need to tell who it was, except that person, and all his fellow students, will benefit from it being brought to everyone's attention. Part of being a student is to learn how to answer questions properly, and with this one, instead of purely 'no' which may get the point awarded, he shot himself in the foot. If I was marking, he certainly wouldn't get the mark, as it's obvious his knowledge is lacking. But that's why tests are given. – Tim Jan 29 '19 at 15:21
  • 1
    @Tim There's no need to highlight how 'not' to do something (this induces unnecessary cognitive load. It's also why telling people 'not to do something' often leads to failure, as you have to think of the thing you don't want to do, then not do it, which increases the odds of you actually doing it). Tell them simply what they should do (which in turn, allows them to forgot the incorrect way). – SE Does Not Like Dissent Jan 29 '19 at 15:37
  • 1
    @Tim Example: Do not think of a pink elephant. Do not think of Mars being retrograde due to two moons. – SE Does Not Like Dissent Jan 29 '19 at 15:39
  • 2
    @JPhi1618 There's more to this than just whether the answer is right or wrong, I think. Creating a comfortable learning environment makes it easier for students to learn, and their perceived ability to trust the instructors/graders is a big part of that. Whether or not you actually trick them, if they feel tricked it will be harder for them to learn, because now they're focusing more on trick questions than on whether they understand the information. Awarding the mark with a note makes it clear to the student that they can trust you to keep your word, leaving them to focus on learning. – DarthFennec Jan 29 '19 at 21:17
239

I think that if you would allow full marks for just yes/no without an explanation at all, then you should allow it here. Otherwise the question is flawed and can't be properly and fairly graded. But a note to the student would be good, also.

To be more precise, if it is possible to answer a question with inconsistent parts it isn't a valid question for examination. It should be clear and clean.

But your job is to educate, not to grade. Give the marks and write the note. And think harder about the questions you ask and how they are presented.

If the explanation is required, it is a different situation. In that case, and if you weight the explanation heavily for other students, then probably 0 marks is better than any other alternative.

Buffy
  • 363,966
  • 84
  • 956
  • 1,406
  • 165
    "your job is to educate, not to grade" Rarely said. Thank you. – MickeyfAgain_BeforeExitOfSO Jan 23 '19 at 12:37
  • 7
    This is a very good question - I hadn't even considered that the question itself doesn't specify the student needs to provide an explanation - the fact that they do anyway should not count against them, but leaving a mark that this is wrong (and why!) is absolutely the correct way to do this. – Zibbobz Jan 23 '19 at 13:41
  • 8
    @Zibbobz - the fact that they do anyway should not count against them - I disagree in general. The issue, as SolarMike points out, is not having a clear grading scheme in advance. I often give problems where no justification is required, but state that they will be graded on what they write. This includes taking off points for writing down bad justification (serious conceptual misunderstanding, not minor calculation error) for the right answer. (Much more often, this rule means they get partial credit for the wrong answer.) – Kimball Jan 23 '19 at 14:37
  • 7
    @Kimball Actually I completely agree with you - when the question specifically asks for an explanation as part of the grade, that explanation should be part of what they get graded on - but the key words there are when the question specifically asks - and in this case, the question did not specify that an explanation would be graded. In this specific case, it would be unfair to grade the student on this answer - and in the future, the professor should make explaining the answer part of the question explicitly, complete with its own allocation of points for an adequate explanation. – Zibbobz Jan 23 '19 at 15:50
  • 9
    I disagree with the conclusion. "Your job is to educate, not to grade"; giving a bad grade for writing nonsense is also a form of education. – Martin Argerami Jan 23 '19 at 16:50
  • 2
    @MartinArgerami In a good world, the examination is also a teaching moment. – Captain Giraffe Jan 24 '19 at 22:18
  • 2
    @Kimball that encourages students to write as little as possible. To make answer only, and not provide any explanation. Which in turn leads to a mindset where no "feedback" is given to what they do. That leads to people who don't comment code, don't explain the mathematics and try to keep as most information to themselves as possible: also while in a work environment. – paul23 Jan 25 '19 at 16:22
  • 2
    @mickeyf: My employer, a public university, pays me to grade as much as it pays me to educate. Certification is an objective as much as is learning. I don't completely like that it is so, but it is so (though in engineering, medicine, law, etc. certification is necessary). – Dan Fox Jan 26 '19 at 08:17
  • @Kimball I think here the grading scheme is perfectly clear. It's a yes no question. That fact that OP set the wrong question, and tried to test the students knowledge rather than understanding is a very different issue. That is the lesson I think OP needs to learn from this. – UKMonkey Jan 28 '19 at 12:42
  • 1
    'Your job is to educate'. The whole point. The point of testing is to ascertain what knowledge has been understood and retained. That answer certainly made it clear that the student wasn't educated in that particular part of the curriculum. Giving a point for something that the student obviously didn't understand will make him think everything is o.k. But to the teacher it's a clear sign that more needs to be done. Or should be! It's a poor question anyway! – Tim Jan 29 '19 at 15:29
45

This is a discretionary matter, and different lecturers will treat it differently, depending on their own educational preferences. However, I disagree strongly with some other commentators on this thread. In my view, there is nothing unfair in marking a student down for unsolicited and incorrect information. Indeed, I would say that this is generally a good practice, since it ensures that the student is held responsible for the correctness of their assertions, even in cases where they offer unsolicited information. This implicitly gives the student some broader training in the importance of ensuring that they give correct information even when they choose to advance information that is unsolicited --- something that is a broader life-skill of importance.

In my personal practice, if a student gives me more information than was requested, and that additional information is wrong, this incurs a marking penalty just as if that information was part of the question. I warn my students in advance that this is my practice, but it is a justifiable practice even without giving a warning. In this particular case, if I were marking the question, I would not give the student full marks.

What kind of graduates do we want? We are training students to become professionals in difficult fields. So, in considering this issue, I think it is important to consider the implicit lessons we give students by what we penalise and what we don't. Imagine that this student graduates and practices in your field. Would it be okay if this practitioner gives unsolicited information to people on the subject area, and that information is wrong? Would you be comfortable working with a colleague who gives information to you or others that is sloppy and incorrect, but then he faces no penalty just because that information was not requested by others? Is that the lesson you would like to impart to your students? Is that what you want to teach them about the world?

Ben
  • 68,453
  • 9
  • 142
  • 263
  • 43
    Do you want graduates who never say anything for fear of being wrong? Marking off for incorrect, unsolicited information seems reasonable on an essay, where clear argumentation is important, but for a true/false quiz, it seems highly unfair to say "this is true/false" and then "surprise! I marked your correct answers wrong because I also saw incorrect stuff!" – cag51 Jan 23 '19 at 06:30
  • 33
    Actually, unsolicited information that exposes a misunderstanding is hugely valuable. Even more so when communicating cross culture/language. I think you're solving the wrong problem here (even though the intention is good, and the problem of off-topic information is real). – Sean Houlihane Jan 23 '19 at 09:32
  • 3
    I agree it's very valuable, and so are (solicited) wrong answers on an exam. Both show the student that there is something they are getting wrong, and the penalty spurs the student to get it right next time. – Ben Jan 23 '19 at 10:27
  • 5
    ...or to keep their mouth shut. – JeffE Jan 23 '19 at 11:08
  • 3
    And keeping one's mouth shut instead of giving wrong information is also a useful skill! – Ben Jan 23 '19 at 11:27
  • 22
    -1 If this student had been trained to engage with the question minimally, in order not to lose marks, they would have just answered "no" and missed an opportunity to learn about retrograde motion. Whether or not learning to "keep quiet in one's ignorance" is valuable (very disputable), it must come second to learning Physics in a Physics class. – Nathan Jan 23 '19 at 14:49
  • @SeanHoulihane: Information like that is incredibly valuable in a formative context, where you're trying to help the students learn. The question is whether tests are supposed to be formative, or whether they're purely evaluative. – Michael Seifert Jan 23 '19 at 15:16
  • 17
    “_What kind of graduates do we want? [...] I think it is important to consider the implicit lessons we give students by what we penalize and what we don’t._” Agreed 100%. We should do our best to teach them to be accountable for what they say by being accountable for what we say. So if we said we’ll award a point on an exam question for writing “yes” or “no” (the correct one of the two), that’s what we should do. – Dan Romik Jan 23 '19 at 16:28
  • 15
    ... Conversely, taking off the point teaches students the lesson that while they are required to always be logical and consistent, the people in authority are allowed to act capriciously and say and do anything they please because they hold all the power. Is that what you want to teach them about the world? – Dan Romik Jan 23 '19 at 16:34
  • 1
    @DanRomik "Is that what you want to teach them about the world?" Was that a rhetorical question? You described how the world is, but that last sentence makes it sound like you think teaching that reality is a problem. Yes, we should teach them that the world is unfair. We don't want to teach them to be that way, but we need to teach them that the world in general is that way. – Aaron Jan 23 '19 at 22:29
  • 3
    @DanRomik: There is nothing capricious about marking down incorrect information given in answers, even if it was not solicited. And no, we didn't say " we’ll award a point on an exam question for writing “yes” or “no”" - you just made that part up. Please read my answer again and you will see that I said that I specifically tell my students the contrary. Even in the absence of this, the mere setting of a true/false problem does not imply that there is no penalty for unsolicted false answers. – Ben Jan 23 '19 at 22:45
  • 4
    Maybe you should read your own answer again. You said that you warn your students, and then you added that ”it is a justifiable practice even without giving a warning”. So yes, with a warning it’s perfectly consistent and uncapricious to deduct the point. Without a warning, if you just write “answer the question with true/false” then you are representing to the students that that’s what will get them the point, period. It’s not really subject to interpretation IMHO, and if if someone is writing that and doing something else, I think I can reasonably call that capricious. – Dan Romik Jan 23 '19 at 22:59
  • 5
    Usually that is not what is written. Usually, all that is written in these types of questions is "The moon is made of cheese (TRUE/FALSE) (1 mark)". In either case, there is no explicit guarantee that a TRUE/FALSE answer, accompanied by wrong information, will receive full marks. So if the student writes "FALSE, the moon is made of milk", then in my view, that is not a full-mark answer. You might regard full marks as implicit in setting the question as TRUE/FALSE, but I don't. – Ben Jan 23 '19 at 23:05
  • 1
    ... Unless there is an explicit instruction on the test saying that a correct TRUE/FALSE answer will receive the full marks, I think it's perfectly legitimate to mark an answer like that (comment above) down. Personally, I would probably give it partial marks, but not full marks. Others might give different marks - I think there is room for discretion here. – Ben Jan 23 '19 at 23:06
  • 4
    And incidentally, if you want to be literal about things, if you actually wrote "answer the question with true/false" in the instructions, and they instead answered with a longer elaboration, then that is already a breach of the instructions, and hence could be penalised. – Ben Jan 23 '19 at 23:08
  • 1
    @Ben I actually don’t feel very strongly about this. I disagree slightly with your position, but also think it’s not an unreasonable position to have. And I agree that the professor may have some discretion, depending on the precise wording of the question. Thanks for an interesting discussion. – Dan Romik Jan 24 '19 at 00:11
  • I agree that a wrong explanation, unsolicited or not, should not get points, but the explanation in this answer misses something. Exams, where guessing is better than not answering, are not the same as real life, where “I don't know“ isnt penalized equally harsh as a plainly wrong answer. – DonQuiKong Jan 24 '19 at 13:50
  • +1 for "I warn my students in advance that this is my practice." If it's on the syllabus, it's OK. There might be some loss of learning, but that is mostly due to the question format. If folks want to prevent loss of learning, write better questions that solicit the stuff that maximises learning. Teaching people to try and avoid making wrong statements on tests and other places they will be judged is a valuable lesson. That said I probably would give them the points. – WetlabStudent Jan 25 '19 at 02:45
  • "there is nothing unfair in marking a student down for unsolicited and incorrect information" <- Not true. – einpoklum Jan 30 '19 at 08:20
  • 1
    "<- Not true." <- Not true. – Ben Jan 31 '19 at 00:01
  • "Unless there is an explicit instruction on the test saying that a correct TRUE/FALSE answer will receive the full marks, I think it's perfectly legitimate to mark an answer like that (comment above) down. " I disagree. I have never seen a multiple choice exam where anything but the answer you choose was important. I think it is reasonable to infer that this holds for all multiplice choice exams in my country and a professor who grades otherwise has to explicitly state so. – user111388 Oct 18 '20 at 21:31
  • Say we had a True/False or multiple choice question that could involve a calculation. Suppose I did the calculation as scratch work next to question. Should I be marked off if my scratch work is wrong even though I arrive at the right answer (assuming it wasn't requested to show my work)? – Tyberius Mar 16 '22 at 18:56
40

Give him the point. If you want, mark on the paper "wrong reason" in red.

If you wanted to evaluate reasons, you would have made it more points and required an explanation. But you didn't. So treat it like a normal true false or multiple choice problem. Reason not graded, just getting the right answer. Luck allowed. Etc. Similarly right reason but wrong result gets hammered.

If providing an answer was required then I guess you could mark wrong any case where both answer and reason were right.

P.s. This is if you are the teacher. If you are the student, don't debate 1 point. Get it all perfect next time.

guest
  • 826
  • 5
  • 5
  • 4
    I debate 1 point questions on tests and the like simply as a matter of principle (but of course, I am respectful and not trying to bother the professor), though I do make sure to get it next time! – user45266 Jan 23 '19 at 06:17
12

There are a few schools of thought here, and it really depends on your teaching style.

From a fairness perspective, you shouldn't mark this student down. I'm quite sure that there are a few other students in your class who could not explain why they got the right answer (a good number of them probably just guessed at random). Unless you have a system to find out who those students were and penalize them, I think you'll have a hard time justifying why this student gets marked down and all those other students are not.

From an instructional perspective, there is some merit to marking this student down. Students have a tendency to write down random stuff with the hope that something sticks. If you incentivize your students to write less bulls**t and more to the point, you are teaching them a valuable lesson. You will be signalling to this student that you care about how they reason about answers, and not just the final product. I know of some lecturers who give their students 1 point in essay questions if they write nothing, just to provide an incentive to not write nonsense.

I would lean towards a fair verdict, but this is really because your question was very limited in answer scope.

Spark
  • 27,465
  • 10
  • 62
  • 101
  • 2
    +1 I agree with your second big paragraph. I think that changing future questions to evaluate the reasoning separately might fix both problems. – user45266 Jan 23 '19 at 06:18
  • 25
    You will be signalling to this student that you care about how they reason about answers, and not just the final product. – If you want to signal that, you should arguably issue an exam that actually asks for the reason and not just the final product. – Wrzlprmft Jan 24 '19 at 18:55
  • @Wrzlprmft s/arguably// (in other words I don't think it's arguable) but otherwise you nailed it. Rewarding the other guessers who keep mum and penalizing the one who explains her reasoning is exactly the wrong approach. – Philip Jan 30 '19 at 23:05
11

The cause of the problem would seem to be the lack of instructions on how to answer the questions.

If the exam paper clearly said "answer the questions with either yes or no," the answer given is wrong because the candidate did not follow the instructions. That might sound harsh, but that's the way the real world works, unfortunately.

If candidates were expected to explain their reasoning, the answer is also clearly wrong.

This is no different in principle from the converse situation where candidates are expected to show their working, and someone simply writes down the correct answer. If they really were smart enough to see the answer immediately rather than do the expected half-page of calculations, they need to learn to be smart enough to also explain why their answer is correct, in real life!

alephzero
  • 4,242
  • 15
  • 19
8

It depends on the actual title above the questions (if there is one). For example:

"Are these statements true?" (or synonyms):

In this case, the actual answer is only "Yes" or "No". There are two options: marking it as incorrect since it's not fully complying with the title (writing further information was not asked), or marking it as correct since its actually complying part ("No") is indeed correct (and potentially crossing out the non-compliant part).

"Are these statements true? Justify your answer." (or synonyms):

The answer would be it as a whole. In this case, I'd mark it as incorrect since it does not answer all the title's parts correctly and thus it's not fully correct.

In case there is no title, it becomes a difficult situation since the mark turns out to be subjective. Besides, the student would be in their right to complain about what they get.

Conclusion: always make your exams with clarity in mind, clearing all technicality loopholes for justifying wrong answers, and no one will be harmed.

Iaka Noe
  • 97
  • 2
  • It is absolutely absurd to mark down for providing additional information just because this was not asked in the title. This is not the case in any exam I have seen and should be explicitly stated. (Is a student also not allowed to scribble on the paper? Or remove a wrong tick when they change their mind because this is an "additional information"?) – user111388 Oct 18 '20 at 21:35
6

I would usually give no credit for a correct answer supported by incorrect argumentation, particulary when the argumentation contradicts the putative answer, is irrelevant, or is absurd. However, given the grade-seeking, formalist/legalist behavior of many students, it is imperative to indicate in the exercise statement that answers will be regarded as correct only if adequately justified. In the current case I would not give full credit, but would not give the 0 I would like to give, unless this had been my practice throughout the course and was known as such to students, because the problem or exam formulation should indicate what sort of answer is required.

It seems to me educationally irresponsible and entirely unfair to award points for an incorrectly reasoned "correct" answer. Such an answer reflects a lack of understanding, and should be graded accordingly.

Here is a concrete example, understandable to some. One asks a student if a given matrix is diagonalizable. One student makes calculations, indicative of conceptual confusion and poorly performed, that would clearly indicate that the matrix has nontrivial Jordan form, but answers that the matrix is diagonalizable anyway. A second student makes a minor arithmetical error that leads via otherwise correct argumentation to the erroneous conclusion that the matrix has nontrivial Jordan form and answers accordingly that it is not diagonalizable. Some seem to think that the first student should be given more credit than the second, when the second student has clearly demonstrated a high level of mastery of material and understanding. (Neither answer should receive full credit).

This isn't a hypothetical example. In mathematics and engineering exams this sort of situation occurs fairly frequently.

Grading serves to indicate whether certain standards have been met. Guessing, or obtaining accidentally, a factually correct answer does not meet reasonable standards for demonstrating understanding of content.

Dan Fox
  • 3,376
  • 1
  • 16
  • 14
6

It seems you were giving a point (a) to students answering "yes" for the correct reason, (b) to students answering "yes" and wasting their time by writing down the correct reason, (c) to students throwing a coin and answering "yes" because they were lucky, (d) to students answering "yes" for the wrong reason, and you ask whether you should give a point (e) to students answering "yes" and writing down the wrong reason.

Since you are willing to give a point to groups (c) and (d), it seems unfair to me not to give a point to group (e). (But if papers are returned to the students, the wrong answer should be marked as wrong).

Of course, asking for an answer and giving reasons would make more sense.

gnasher729
  • 3,554
  • 14
  • 16
4

If you want to be absolutely right, treat the answer as wrong.

The purpose of marks is to accurately reflect the level of student's learning. If there is clear evidence that the student hasn’t learned the correct concept, marks should not be awarded.. unless you are grading their luck. (The student is clearly lucky here.)

But, really... marks are secondary. Won’t hurt if you give them the marks and make sure that they get the concept too. That’s your job as an instructor.

So, give them the points and explain the correct concept to them. Win-win for everyone.

displayName
  • 339
  • 1
  • 8
  • 2
    In most institutions one's job as an instructor involves certifying whether students have learned such content. Grades do matter whether we like it or not, because our employers and our clients give them importance. They should be taken seriously, or one should go teach in a context where there are no grades (with the correspongind remuneration). – Dan Fox Jan 26 '19 at 08:14
  • @DanFox: I don’t object to what you are saying. Therefore I’m not sure about the point you are trying to make. – displayName Jan 26 '19 at 14:14
  • 1
    "But, really... marks are secondary. Won’t hurt if you give them the marks and make sure that they get the concept too. That’s your job as an instructor." In many institutional contexts, from the point of view of many students, and from the certification point of view it is not true that "marks are secondary". Like I think also you, I wish it it were otherwise. That it "won't hurt" just to "given them the marks" seems to be questionable from the educational point of view also. – Dan Fox Jan 26 '19 at 14:21
4

Is this a standardized exam? Will the score acquired by the student affect his chances of getting into college? Will it be carried onto the future? If so, award full marks. You were not testing the mechanics behind it. There is a good chance that other students didn't know the right explanation either, and thus, on account of fairness, you should give the student full marks.

The question only asked the student to answer Yes/No to the statement presented. Consider this: on any given test, there are going to be a number of people who answered "Yes", but didn't give an explanation. What if they didn't know the explanation either? What if they just guessed it? What if they tossed a coin? There are lots of facts that we know without knowing why they are true. Plenty of students in your class would not have known the explanation.

If this test is not for some standardized exam and if it'll not have any bearing on the student's feature, we can throw the fairness criterion out the window entirely, and concentrate on the more important aspect of testing: to see if the student has gained the required knowledge. In this case, whether or not you award 1 mark will not matter. You have two options:

  • Award full marks, but leave a note.
  • Award zero marks and leave a note.

What you choose is entirely your preference. Ideally, you'd want to award 0.5 marks; but since fractional scores are not permitted, I'll make a case for (2). Students are most likely to review questions for which they are marked wrong. If something is correct, it's correct and most students wouldn't want to bother with it. If they're marked wrong on a question, they'd want to know why. A student might review the whole paper, in which case your comments would be noted in both cases; but this is less likely to happen. It'll also encourage a student to not go beyond what's asked in a question. You don't have to present everything you know to the examiner in a test. While "sticking to instructions" may not always bode you well in life, in a test, it will certainly help you very much.

WorldGov
  • 520
  • 4
  • 11
3

Whatever you do, you need to do it consistently for all students. For this reason we have a marking scheme. Presumably, you are not the only person responsible for the course? Are you working with a professor or colleagues? Ask them what they would do and do the same.

If there is no marking scheme and each marker makes an independent decision — discuss and work out a marking scheme that everyone can accept. Then write it down and ideally incorporate into the questions to make students aware, e.g. You need to explain your answer - correct answer with incorrect or absent explanation is worth 0 marks. Then simply stick to it.

Dmitry Savostyanov
  • 55,416
  • 14
  • 140
  • 202
  • Except that not giving an explanation is a valid answer and was assumed to be valid before giving the test. The issue is that the student gave more information than needed. – user64742 Jan 25 '19 at 02:39
  • @TheGreatDuck Still, we assess student's understanding and provide constructive suggestions for improvement. We do not just tick the "right" answers. By providing incorrect explanation student gave evidence of their lack of understanding. – Dmitry Savostyanov Jan 25 '19 at 12:17
  • I’m referring to your paragraph about making schemes and the last sentence. – user64742 Jan 25 '19 at 18:05
1

Where does this "grading scheme" come from?

If you made it yourself, cross it out and make a better one that gives zero marks for both "yes" and "no" and only gives marks for an answer with a correct explanation. Students should by this point know that an unexplained answer is worthless, unless the question specifies otherwise, and your example does not.

If you received it from somewhere else, don't ask strangers on the internet to adjudicate, go back to the person or organisation that gave you the grading scheme and ask them for a ruling. This is the only way that the grading can be fair between different graders.

Jack Aidley
  • 11,898
  • 2
  • 34
  • 51
  • The question is only worth one point though. If it were worth five then yeah, one can expect an answer and an explanation, but with only one point it hardly seems fair. Besides it's not hard to word the question to ask for both an answer an explanation - just add "why?" to the end. – Allure Jan 23 '19 at 11:12
  • 12
    Students should by this point know that an unexplained answer is worthless — Only if you actually tell them that an unexplained answer is worthless! If your exam only says "Yes/No", then you're only asking for "Yes/No", which means you must give full credit for a correct Yes/No. Students should not be expected to read your mind. More strongly: Students should be expected not to read your mind. – JeffE Jan 23 '19 at 11:16
  • 1
    @JeffE But the example does not say yes/no. If it did then you'd have to accept yes/no otherwise it's a failure to answer the question. – Jack Aidley Jan 23 '19 at 11:41
  • 5
    @JackAidley It only asks a yes/no question. If you want more than a yes/no answer, ask for more than a yes/no answer. Use your words. – JeffE Jan 24 '19 at 02:13
  • 1
    @JeffE: Perhaps it's a cultural difference, but I would expect any student older than about 14, and certainly any university level student to know not to give yes/no answers without an explanation unless it's explicitly made clear that a yes/no answer is sufficient. Students should be expected to understand the conventions of exams, and answer accordingly. And, yes, a better worded question would be nice, but it's already been asked. – Jack Aidley Jan 24 '19 at 10:21
  • 5
    Students should be expected to understand the conventions of exams — You're assuming that there are universal conventions for exams. There aren't. – JeffE Jan 24 '19 at 13:52
  • 1
    @JeffE If you are so naive as to imagine you've ever set an exam which doesn't rely on the conventions of exams you are very, very much mistaken. – Jack Aidley Jan 24 '19 at 22:15
  • 1
    @JeffE there are universal conventions. If there is a question like in the asker's post and 5 blank lines spanning the page it should be fairly obvious that "Yes/No" is not acceptable. I would say the default is to assume an explanation is wanted unless specifically stated. – user64742 Jan 25 '19 at 02:35
  • 2
    @TheGreatDuck And where did OP say there were 5 blank lines? You and I might say the default is to assume an explanation is required, but not everyone has that default. Even different instructors in the same department might have different defaults. Even if your university has mandated conventions that everyone actually follows (which would be a miracle), you still have to be fair to new/visiting/1st-year students who haven't learned the default yet. Use your words. – JeffE Jan 26 '19 at 09:12
  • 1
    @JeffE There aren't 5 blank lines because explanations were not required for that question. Where did I say that this was the case of the askers question? You said there are no exam conventions. There are and anyone who graduated high school should already know at least by common sense when an explanation is expected. When it isn't expected might be different. However, given the example questions numerous unwarranted edits it is also likely that the actual question says (yes/no) with a one word blank after it. Unless this student writes absurdly small they are ignoring clear signals. – user64742 Jan 26 '19 at 17:10
  • 1
    @JeffE and why do you assume this is some magically university policy? It's just that if someone gives you a question with a large portion of writing space you should logically assume they want a long answer. If it's short, it would make sense to give a short answer. And if the student is unsure they should err on the side of more details than less details because that is the proper thing to do in life period. If someone asks a question via email and you don't know how much information they want you give them more details not less. Not 200 page book more, but more than "yes" or "no". – user64742 Jan 26 '19 at 17:14
  • 1
    because that is the proper thing to do in life period — Except when writing an exam, apparently. – JeffE Jan 27 '19 at 02:43
0

I personally would award the point (but disclaimer: I'm not a professor), but probably implement other answers' solutions (like write on the test an explanation). I'd like to point out that if this one question or a similar one were worth a huge amount of points on a test, of course it would be necessary to award points because the question never asked for reasoning. That question could simply ask for an explanation to avoid this whole issue.

user45266
  • 361
  • 3
  • 9
0

There are two methods to use in this situation, and it really comes down to the environment you have made in your classroom.

Let's start with method one. Disregarding the response, giving the question full marks, and leaving a note for the student describing what was wrong with their explanation. There is nothing wrong with this method, and it educates, not punishes the student, because at least they gave effort. However, do cross it out and make sure the note explicitly states that the answer is wrong.

Method 2: Mark it wrong, and leave a note. While this doesn't "punish" the students, per se, though it does help them learn something and prepare for the field that they want to go into. Many of these students will want to go into fields that involve specific (correct!) answers. In this case, you would mark it wrong in the hopes that they may review it. If this seems a bit harsh, maybe invite him to do a correction of sorts.

Inviting him to do a correction seems like a compromise without resorting to fractional grades. It also allows him to make up the question while preparing him for the workplace, and making sure he learns and reviews the material, which will be beneficial later in life. If this isn't allowed, then it is more of a personal preference and a moral compass type situation, in which case you should look at both of the methods above and decide which one most supports the atmosphere that you have created in your class.

ava
  • 121
  • 4
-3

A recommendation I usually give when someone has an analysis paralysis:

  1. Grab a coin. Head for 1, tail for 0.
  2. Toss it high, so that it will spins as long as possible on the floor
  3. During the spinning, your expectation will show up most. Which side do you DON'T want most?

After you have know what you want, you can know why you don't want it.

Ooker
  • 8,266
  • 6
  • 54
  • 114
  • 1
    I use this a lot too but I simply ask if when it landed were they disappointed or relieved. – CramerTV Jan 23 '19 at 20:30
  • 4
    What one person arbitrarily wants is not a good standard for marking a test. – Matthew Read Jan 29 '19 at 21:18
  • @MatthewRead no, this is just a mean to get out of the analysis paralysis, so that the OP can be aware of what they actually prefer. After that, they can justify their preference and stick with it. – Ooker Jan 30 '19 at 02:48
-6

Ask them to give a confidence score (1 to 5 say) for their explanation. A high confidence but wrong explanation suggests a real lack of understanding for example.

Paul
  • 285
  • 1
  • 3
  • 11
    The exam is over. – JeffE Jan 23 '19 at 11:08
  • 2
    I'm upvoting this answer but the real problem is with the OP's himself/herself needing confidence and expertise scores in astronomy. You can see it in comments and in the edit history of the original question. Astronomy has been "butchered" there. I have nothing against the technical terms "prograde" and "retrogade" but there's more to hard sciences than flaunting vocabulary and putting all efforts into making happy-go-lucky tests. I'd say a lot more: confidence, knowledge, proofs, and derivations. Voting was disgusting here. This comment is necessary. – Ken Draco Jan 24 '19 at 00:41