nucleo


« Reply #15 on: May 03, 2012, 8:06:48 pm » 

And on a quiz with a simple question: "Who is the murderer in this story?" there is always at least one student who responds with two answers, hoping one is correct. So I tell my students that I only accept the first answer; they can't just make a list.
Wow, you accept the first answer? I mark a zero on the whole thing if the question clearly has N answers and someone puts greater than N things. I explain to my students that answering, "What are the three colors in the American flag?", with "Red, Orange, Yellow, Green, Blue, Purple, Black, and White" tells me that the student has no clue and is hoping to get lucky whereas "Red, white, and dang, I forgot" is a pretty good answer. The way I'd mark that one is that for each missing or extra answer, I take off half a point until there are no points left. So "Red, white, and dang, I forgot" and "Red, white, blue, and green" would both lose half a point, while "Red, white, and green" would lose the full point (and these types of questions are typically worth one mark for me). Well, I thought about that when I was making up the policy, but the second answer seems more wrong to me than the first answer. I told you three and you put four so not only don't you know the answer, but you can't apply basic logic, either. Ah, good point. You see, I'd never actually tell them how many answers they should be giving...



Logged




lohai0


« Reply #16 on: May 03, 2012, 8:08:19 pm » 

I teach Calculus II and III. I have developed a technique where all problems are worth 5 points. This makes partial credit easier.
1 if you make an arithmetic error 2 if you used the correct technique but didn't make it all the way to the end. 3 if you made an H.A.M. (horrible algebra mistake, like thinking (x + y)^2 = x^2 + y^2 4 if you were in the room while a test was taking place and wrote something down.
So, in the case of the problem you are talking about, I would grade it 2  not confident enough to stop at the right answer.
Welcome, fellow math person! My 5 point scale is a little more cynical, but pretty much the same: 1: not blank 2: at least semirelated fake work 3: approximately half right 4: minor error (dropped minus sign, etc) 5: correct with correct work ptarmigan, in this case I'd give a 3/5, because I hate half points. Otherwise 2.5.



Logged

I hate helicopter zombie grandparents. They are the worst.



ptarmigan


« Reply #17 on: May 03, 2012, 9:26:47 pm » 

I teach Calculus II and III. I have developed a technique where all problems are worth 5 points. This makes partial credit easier.
1 if you make an arithmetic error 2 if you used the correct technique but didn't make it all the way to the end. 3 if you made an H.A.M. (horrible algebra mistake, like thinking (x + y)^2 = x^2 + y^2 4 if you were in the room while a test was taking place and wrote something down.
So, in the case of the problem you are talking about, I would grade it 2  not confident enough to stop at the right answer.
Welcome, fellow math person! My 5 point scale is a little more cynical, but pretty much the same: 1: not blank 2: at least semirelated fake work 3: approximately half right 4: minor error (dropped minus sign, etc) 5: correct with correct work ptarmigan, in this case I'd give a 3/5, because I hate half points. Otherwise 2.5. When I teach my own class, I will definitely be grading things on a 5orsopoint scale. The lecture prof I've worked for this year insists on having each problem (on the quizzes) be worth 20 or 30 points. And, yes, I know that points scale, but it still feels different/wrong. I ended up giving halfcredit for the style of answer I was asking about here. The poor kids had an exam Tuesday followed by a quiz on Thursday and I graded it a bit generously to help them out.



Logged

He's on my roster, but if I've taught him anything, it isn't math.



tuxthepenguin


« Reply #18 on: May 04, 2012, 12:18:55 am » 

I teach Calculus II and III. I have developed a technique where all problems are worth 5 points. This makes partial credit easier.
1 if you make an arithmetic error 2 if you used the correct technique but didn't make it all the way to the end. 3 if you made an H.A.M. (horrible algebra mistake, like thinking (x + y)^2 = x^2 + y^2 4 if you were in the room while a test was taking place and wrote something down.
So, in the case of the problem you are talking about, I would grade it 2  not confident enough to stop at the right answer.
Welcome, fellow math person! My 5 point scale is a little more cynical, but pretty much the same: 1: not blank 2: at least semirelated fake work 3: approximately half right 4: minor error (dropped minus sign, etc) 5: correct with correct work ptarmigan, in this case I'd give a 3/5, because I hate half points. Otherwise 2.5. Not saying you're wrong, but I'm curious why. They're supposed to demonstrate that they know the answer. They're just throwing a bunch of things they saw in the lecture down on paper. That shows that they don't understand anything. They're not even claiming that the correct answer is correct. That's about the strongest evidence you can have that they don't understand the material. In a class of 80 students, with three exams per semester, I give 240 exams. I get no complaints about my grading. I explain my reasoning the lecture before the exam, then I write the grading policy on the exam, then I remind them of it before I hand out the exam. I've had a lot of students tell me that they like my policy. It relieves them of having to play games if they don't know the answer.



Logged




lohai0


« Reply #19 on: May 04, 2012, 1:30:38 am » 

I teach Calculus II and III. I have developed a technique where all problems are worth 5 points. This makes partial credit easier.
1 if you make an arithmetic error 2 if you used the correct technique but didn't make it all the way to the end. 3 if you made an H.A.M. (horrible algebra mistake, like thinking (x + y)^2 = x^2 + y^2 4 if you were in the room while a test was taking place and wrote something down.
So, in the case of the problem you are talking about, I would grade it 2  not confident enough to stop at the right answer.
Welcome, fellow math person! My 5 point scale is a little more cynical, but pretty much the same: 1: not blank 2: at least semirelated fake work 3: approximately half right 4: minor error (dropped minus sign, etc) 5: correct with correct work ptarmigan, in this case I'd give a 3/5, because I hate half points. Otherwise 2.5. Not saying you're wrong, but I'm curious why. They're supposed to demonstrate that they know the answer. They're just throwing a bunch of things they saw in the lecture down on paper. That shows that they don't understand anything. They're not even claiming that the correct answer is correct. That's about the strongest evidence you can have that they don't understand the material. In a class of 80 students, with three exams per semester, I give 240 exams. I get no complaints about my grading. I explain my reasoning the lecture before the exam, then I write the grading policy on the exam, then I remind them of it before I hand out the exam. I've had a lot of students tell me that they like my policy. It relieves them of having to play games if they don't know the answer. I don't think it really matters and long as the guidelines are clear and consistent. With the student population I am working with now, there are two general groups that will go with this sort of shotgun approach and hit the right answer by accident: (1) really anxious type A students and (2) students with no understanding. For the first group, losing points for shotgun problem solving is enough of a deterrent that this strategy goes away quickly. For the other group, they are still failing the problem, but are close enough to passing that they might keep trying. FWIW, I grade on total points (usually 5000+ points in a course) so I feel like a point here or there will wash out at the end anyway.



Logged

I hate helicopter zombie grandparents. They are the worst.



wittgenstein
New member
Posts: 16


« Reply #20 on: May 07, 2012, 4:35:36 pm » 

I developed the 5 point scale after TAing for a prof in grad school who counted some problems 8 or 9 points. Those of us who were her graders spent hours trying to figure out a consistent way to grade certain college algebra problems.
Also, the longer I teach, the more I think spending hours obsessing over partial credit is not time well spent. I don't see how this time helps students learn.



Logged

"What we cannot speak about we must pass over in silence."



tuxthepenguin


« Reply #21 on: May 07, 2012, 8:35:16 pm » 

Also, the longer I teach, the more I think spending hours obsessing over partial credit is not time well spent. I don't see how this time helps students learn.
You're 100% correct. Here is my reasoning. I asked a question, they didn't give me the right answer, so it is wrong. "Sort of correct" is not a welldefined concept. It doesn't help the students or anyone else to pretend a wrong answer is right just so the students don't complain. I'm more likely to give full credit for a wrong answer, rather than partial credit, on the basis that they've demonstrated their knowledge but missed a few less important details. If the details really are important, they don't know what they're doing, so they don't deserve partial credit. If I give partial credit it's actually because there are several small questions in one. In that case it's trivial to decide how much "partial credit" to award. In my opinion partial credit is a signal that you need to change the way you're teaching or writing your exams. Partial credit is a sign that you're not being clear about something  normally you are asking vague questions. Even essay questions shouldn't be given 3/5 points. That's why I can't see why the OP is fussing about partial credit. They clearly gave the wrong answer. They should get the score that is appropriate for a wrong answer.



Logged




macadamia


« Reply #22 on: May 08, 2012, 3:22:04 pm » 

What should I do with someone who works the full 10 steps for a problem, but had a finger slip on the calculator to end up with something that is the right order of magnitude, but not the right number? That's a 0 out of 100, really? No, that's a 90 because doing all those calculations three times for consistency is a waste of effort for a freakin' test.
The issue is that some students have so many "finger slips" that it is absolutely impossible that they can solve any 10steps problem correctly. I usually give partial credit, but last week, we had a multiple choice quiz and one (otherwise charming) student lost her cool and accused me of making them all fail by giving a multiple choice quiz that obviously cannot give partial credit. I looked up the grades afterwards and she is indeed one of the very few students who did worse on the multiple choice quiz than on the essay ones.



Logged

A drunk man will find his way home, but a drunk bird may get lost forever. Shizuo Kakutani



tuxthepenguin


« Reply #23 on: May 08, 2012, 11:10:52 pm » 

What should I do with someone who works the full 10 steps for a problem, but had a finger slip on the calculator to end up with something that is the right order of magnitude, but not the right number? That's a 0 out of 100, really? No, that's a 90 because doing all those calculations three times for consistency is a waste of effort for a freakin' test.
For me, that's full credit. The student has given the correct answer by showing how to do it. Hopefully in real life an engineer is doing the calculations three times. If a question really does require 10 steps (I'd normally break it up into smaller pieces) I will specify that each correct step is worth x points. They either do each step correctly for x points or zero points. My post was in response to the OP, who is clearly doing things differently.



Logged




ptarmigan


« Reply #24 on: May 08, 2012, 11:37:12 pm » 

What should I do with someone who works the full 10 steps for a problem, but had a finger slip on the calculator to end up with something that is the right order of magnitude, but not the right number? That's a 0 out of 100, really? No, that's a 90 because doing all those calculations three times for consistency is a waste of effort for a freakin' test.
For me, that's full credit. The student has given the correct answer by showing how to do it. Hopefully in real life an engineer is doing the calculations three times. If a question really does require 10 steps (I'd normally break it up into smaller pieces) I will specify that each correct step is worth x points. They either do each step correctly for x points or zero points. My post was in response to the OP, who is clearly doing things differently. I am usually grading quizzes that have about 4 problems, each worth 2030 points, in a class where I am the TA. I would give partial credit if it were up to me, but I am also required to do so by the instructor of record.


« Last Edit: May 08, 2012, 11:37:37 pm by ptarmigan »

Logged

He's on my roster, but if I've taught him anything, it isn't math.



lohai0


« Reply #25 on: May 09, 2012, 12:52:09 am » 

What should I do with someone who works the full 10 steps for a problem, but had a finger slip on the calculator to end up with something that is the right order of magnitude, but not the right number? That's a 0 out of 100, really? No, that's a 90 because doing all those calculations three times for consistency is a waste of effort for a freakin' test.
For me, that's full credit. The student has given the correct answer by showing how to do it. Hopefully in real life an engineer is doing the calculations three times. If a question really does require 10 steps (I'd normally break it up into smaller pieces) I will specify that each correct step is worth x points. They either do each step correctly for x points or zero points. My post was in response to the OP, who is clearly doing things differently. I am usually grading quizzes that have about 4 problems, each worth 2030 points, in a class where I am the TA. I would give partial credit if it were up to me, but I am also required to do so by the instructor of record. That sounds horrifyingly like someone I used to TA with, but I know he is at a different school. He graded every problem out of 10 points with quarter point intervals.



Logged

I hate helicopter zombie grandparents. They are the worst.



ptarmigan


« Reply #26 on: May 09, 2012, 1:54:42 am » 

I am usually grading quizzes that have about 4 problems, each worth 2030 points, in a class where I am the TA. I would give partial credit if it were up to me, but I am also required to do so by the instructor of record.
That sounds horrifyingly like someone I used to TA with, but I know he is at a different school. He graded every problem out of 10 points with quarter point intervals. I usually go through the stack one problem at a time, marking the problems that are completely right (they go in a separate stack) and, for the others, marking mistakes but not giving a score. Then, if there are only a few types of errors being made (in other words, if I only have 3 or 4 basic patterns of wrong answers), I will make up a points value for each style of answer. In the worst case, if the answers are all over the place, I first assign points to each part of the problem (found zeroes  5 pts; sign chart  10 pts; correct answer in interval notation  10 pts) then figure out how many points to take off for smaller mistakes. This pretty much takes forever. I think when I'm teaching my own course next year I will grade everything out of 5 points along the lines some suggested earlier. I do believe in giving partial credit in math. Sometimes we give multiplechoice exams because there is just no other reasonable option (I'm not willing to handgrade 100 exams under current circumstances), but that's not really how math works.



Logged

He's on my roster, but if I've taught him anything, it isn't math.



kiana


« Reply #27 on: May 09, 2012, 8:25:39 am » 

I usually go through the stack one problem at a time, marking the problems that are completely right (they go in a separate stack) and, for the others, marking mistakes but not giving a score. This is pretty much how I end up grading exams. I already know how many points I'll take off each problem for the most common mistakes, but on the first pass anything bizarre goes into the 'wtf?' pile. Then I compare the wtf pile to each other to look for commonalities. Continuing *past* the correct answer loses points depending on the degree of continuance. i.e. A student who simplifies his answer correctly when the directions say 'do not simplify' loses nothing but time. A student who simplifies his answer incorrectly when the directions say 'do not simplify' loses points for algebra errors. A student who, on the other hand, writes the correct answer and then proceeds to do work that indicates he has no idea this is the right answer loses most of them. e.g. If f(x) = the integral from 1 to x of g(t) dt, what is f'(x)? A student who writes g(x) and then proceeds to differentiate it clearly doesn't understand the question, but rather has a rote understanding of the rules to take derivatives.



Logged

If robbers ever broke into my house to search for money, I'd just laugh and search with them.



conjugate
Compulsive punster and insatiable reader, and
MemberModerator
Distinguished Senior Member
Posts: 18,996
Tends to have warped sense of humor


« Reply #28 on: May 09, 2012, 11:35:40 pm » 

Continuing *past* the correct answer loses points depending on the degree of continuance. i.e. A student who simplifies his answer correctly when the directions say 'do not simplify' loses nothing but time. A student who simplifies his answer incorrectly when the directions say 'do not simplify' loses points for algebra errors. A student who, on the other hand, writes the correct answer and then proceeds to do work that indicates he has no idea this is the right answer loses most of them.
e.g. If f(x) = the integral from 1 to x of g(t) dt, what is f'(x)? A student who writes g(x) and then proceeds to differentiate it clearly doesn't understand the question, but rather has a rote understanding of the rules to take derivatives.
For a long time I had a problem with students who would calculate the derivative correctly (as the problem requested) and then set the derivative equal to zero, find the critical numbers, and attempt to classify them with the Second Derivative Test. I thought about taking off a point for being damned annoying and not reading directions, but eventually realized that the penalty for taking the time to do all that without reading the directions carefully was enough. Of course, errors in the search for critical numbers or second derivative would be points counted off, but in that case I think failure to read and follow the directions was selfpunishing.



Logged

Unfortunately, I think conjugate gives good advice.
∀ε>0∃δ>0∋x–a<δ⇒ƒ(x)ƒ(a)<ε



