Grading: students give correct answers, but in the wrong places

Question

In the last assignment I've given my students, one of the questions has two subquestions that go as follows.

1. Explain how Theory X, as developed in Smith (2010) accounts for dataset A.
2. Briefly summarize the additional data that, according to Smith (2010), fall under the scope of Theory X.

A number of students have submitted something along these lines.

1. Something that is technically correct, but doesn't actually answer Question 1.
2. Something that is technically correct and answers both Question 2 and Question 1.

What is the proper grade for this students? The three possibilities that I have in mind are:

• Full credit: they have provided correct answers to both questions, even though the answer to Question 1 is embedded within the answer to Question 2.
• Full credit only for Answer 2, no credit for Answer 1: they not only have to provide the correct answers, they also have to provide them in the right place.
• Half credit for Answer 2, no credit for Answer 1: half credit because the Answer 2 contains more information than I asked for.

I'm inclined towards the second option (full credit only for Answer 2). Would this be appropriate?

Update: in response to some of the answers below, the problem is not that my questions are unclear, as a majority of the students have managed to answer them properly.

Answer 1

From the level of detail you provide it is not entirely clear to me which option is the most fair one:

• Possibility 1 seems fair if the questions were not entirely clear, i.e., it is reasonable to assume that a student really just misunderstood. If a large fragment of your course actually answered as you said, this may be a reasonable assumption.
• Possibility 2 is my default policy for exams. It is not ok to answer "something" factually correct at some point, the answer also needs to fit the question. That is, if I hypothetically ask "What is A?" and "Also describe the alternative approach B", and the student brings up the right answers but mixes up A and B, she will get (fairly, as I think) a grand total of 0 zero points, or close to it. Of course this is only fair if it was obvious what was being asked. For example, if I had asked instead "Name and describe a really cool standard idea to do foo" (thinking about A), and the second question is then "Name and describe a more uncommon alternative way to do foo" (thinking about B), then purely mixing them up is not worth 0 points anymore.
• Possibility 3 seems somewhat iffy. Detracting points for "too much information" opens a can of worms, and easily brings students into a "doomed if they do, doomed if they don't" situation where they lose points for too much as well as too little info. However, this is again something that can be fixed by having sufficiently clear questions. If I ask "Define C according to McFoo", and the student offers 3 alternative definitions (one being the one of McFoo), the two other definitions are not just additional information, but objectively wrong. The main point here is to prevent students from gaining points by just blurping out a braindump of "every piece of info on topic C", without actually understanding the question.

As said, from your question I am not entirely sure which case your situation actually falls into, however, it kind of sounds like Possibility 1 is fair in your specific case. If you are adamant that the question was clearly phrased and the students just misunderstood based on their lacking understanding of "Theory X", then Possibility 2 can also easily be justified. I would steer away from Possibility 3 in your case.

Answer 2

What is your purpose of a grade? If it is a numerical certification of their knowledge, full credit is the right choice, as they indeed have shown to know the answer to both questions. If the assignment didn't have time pressure, you may consider applying a small penalty for the not so good organisation, but be prepared to have it challenged.

Don't penalise adding extra, factually correct, information. Other professors welcome it, and even, the expect you to understand that the question has unwritten sub questions that should be answered too for the full credit (I had one of these in high school, and "give three examples of mammals" meant also "explain what they are and why are they mammals, or I will only give you half").

Lastly, if several students have made the same mistake, consider that maybe the questions are not clear. I must say I don't quite understand the difference from your excerpt.

Answer 3

• Possibility 1 would not be fair as the students didn't manage to understand the questions correctly.
• Possibility 2 could be applied but is very rigorous.
• Possibility 3 is unfair as the students answered the question correctly and haven't written anything wrong.

I think you are missing a fourth option, full credit for answer 2 but only partial credit for answer 1. While I understand that one could argue that the students answered both questions correctly I also think that understanding a question is part of the correct/perfect solution and should therefor influence the credits. But it should also be acknowledged that the students have shown that they have the knowledge to answer both questions correctly.

If some students then want to argue with you regarding their answer you can always tell them that if they think a question is unclear they should always ask.

Answer 4

The purpose of an exam is to establish the students' mastery of your subject, and absolutely not to test their mastery of taking exams.

If we give anything other than full marks for a correct answer, we have fallen into the trap of trying to mark their mastery of exam technique, a non-skill of no use in the real world, as opposed to marking their mastery of our teachings and our subject.

Oh, maybe if we're trying to train FAQ authors? Yeah, then it matters that they give the exact right info at the exact right point in a document I guess. But... I'm pretty sure you aren't.

the problem is not that my questions are unclear, as a majority of the students have managed to answer them properly.

Consider the following sentence: "This road junction is not unsafe because a majority of drivers pass through it safely." What percentage of drivers would have to have accidents at a junction, before you would call a junction unsafe? Did fewer than this percentage of your students make this error?

Of course not. You wouldn't be here asking that question of such a low error rate. Your question was unsafe, and in fact, I'm willing to bet that I can predict exactly how they interpreted it:

A) Explain Theory X, as developed in Smith (2010).

B) Briefly summarize the additional data that, according to Smith (2010), fall under the scope of Theory X, and say whether/how this applies to Dataset A.

That's how they interpreted it, isn't it?

To claim that your question was perfectly crystal clear, based on the fact that fewer than 100% misinterpreted it, is to disingenuously ignore the fact that several intelligent and hardworking students who have fully mastered the question's subject, all made the exact same error when faced with this question, and someone who didn't know the subject or even have the full question (me) was able to predict the specific misinterpretation that they made.

Either:

• Full marks; or
• Be prepared to justify 1) why you are marking exam skills rather than course knowledge, and 2) why you had a consistent error of interpretation from the completely-correct responses, which you disclaim any responsibility for and instead put down to... what, exactly? It's not ignorance, it's not stupidity, so what trait are you ascribing it to, and marking it down for? Failure to re-read a question enough?

Answer 5

I wouldn't create a policy for this. I would go on a case by case basis and judge how much it feels like the student understands it given their answers.

When the student writes too much, there are two options: 1) It shows a lack of understanding, or 2) It shows good understanding. I would penalize #1 and, if I was going to penalize #1, it would only be fair to not penalize or even to give a bonus for #2.

I would also keep in mind that tests are indeed high pressure environments and therefore have pressure induced errors therefore not adequately demonstrating the student's true understanding.

Answer 6

I suppose the real answer is whether you consider these questions to be concatenate or discrete to one another. If you treat them as concatenate, then it makes sense to look at their answers holistically to tease out whether they understand the material. In such a case, they might earn full points, simply because they have demonstrated that they understand the material

If you treat them discretely, then you should simply be as systematic about it as possible.

I would suggest the following logic:

IF student response to Question 1:
    Fulfills the requirements of Question 1, assign full points;
    Partially fulfills the requirement of Question 1, assign 1/2 points;
    Does not fulfill the requirement of Question 1, assign 0 points;

IF student response to Question 2:
    Fulfills the requirements of Question 2, assign full points;
    Partially fulfills the requirement of Question 2, assign 1/2 points;
    Does not fulfill the requirement of Question 2, assign 0 points;

This is an easily-defended manner of grading. If a student challenges you and states that they answered Question 1 in their response to Question 2, you can concede that they did, and inform them that they would have gotten full points had they done so in response to Question 1.

I would suggest, however, that your third option is not only unnecessary, but arbitrary. Have you stated, clearly, that students should add absolutely no extra content in their responses? It seems cruel to reduce points for such an easy mistake, particularly when, ex ante, student's are uncertain of what constitutes 'too much'.

Answer 7

I would probably handle it the same way as if question 1 was not in the exam and got that answer for question 2.

Answer 8

Grading must be completely objective. If I give a lesson: Balls are spherical and the primary color red has a shade called crimson.

and ask 2 questions of you, such as: 1. What is the shape of a ball? 2. What primary color is closest to the color Crimson?

and you were to answer with: 1. Violets are blue. 2. Rubies are red and almost spherical in nature.

I'd have to grade both with no credit as #1 is technically correct but it doesn't answer the corresponding question, and the second has the answer to both in it, but is not used in proper context of the lesson.

Sadly, when students play with their answers, it's difficult to tell if they learned the lesson. But, what I know from first hand experience is when a student plays with answers like this, they're bored and are likely to do well given a chance in an honors or otherwise advanced course.

Happy teachings!