I need help grading math tests...

[apstaats]

Hi, I am a new high school math teacher, and I am really comfortable with all aspects of my job, except grading tests. Grading seems so subjective to me, even in math, unless you use an all-or-nothing method of grading. I am really uncomfortable deciding how much partial credit to give incorrect answers, and I am constantly second-guessing myself. I really wish there was a formula for this, because I could definitely follow one. I've looked at so many math rubrics, and they still seem so subjective to me. Every time I ask a colleague, she can only give me examples of how she would grade a specific problem, but nothing that I could use for all of my exams throughout the year. It seems like something that I need to get a "feel" for, but I've been a TA for the last three years before my current job, and I still don't get it. I'm hoping that someone here might be able to give me some advice...

[mathwhiz]

It does take time to develop a "feel" for grading if you are giving partial credit. You need to decide in your mind (or in writing so that you stay consistent) how much you will deduct for a computational error vs. a concept error. Also, don't double penalize on the same question for the same type of error. I have a collegue who does the all or nothing system and it works for them. Whatever method you use, use positive points, not negative points when making deductions. I hope this helps a little.

[apstaats]

Thanks, mathwhiz, I really like your suggestions. I would like to try coming up with specific point deductions for computational and conceptual errors, and see how that works. I need something methodical, and I think that might do it for me. May I ask, what point values have you decided on for those types of errors?

[annettemcd]

When I was in college (in the Dark Ages when we still used slide rules), I have a physics professor who did not require us to do the computations on exams (since only a few had calculators and slide rules took too much time). We were required to set up the entire problems with a horizontal line and vertical lines separating pairs of equivalent units. Then we had to determine the units of the answer, but not the number itself. If we had had calculators, it would have taken seconds to multiple all the numbers above the horizontal line and multiple all the numbers below the line and then do the final division calculation, but that did not earn us any points.

I use this as an example only. In math, understanding the concepts which are involved in solving a problem and in setting up the work to be able to get the right answer may be considered at least as important if not more important than the final numeric answer.

I would recommend a system of requiring that students show all work on tests and then give, say, 5 points if everything is correct, but 3 points if the problem is set up correctly, but the answer is wrong and 4 points if the set-up and the units are correct (if there are units), but just the numeric answer is wrong. That would also mean that just the right answer with no work and no units would only be worth 1 point. (Hopefully, thus convincing them of the importance of showing their work.) Just a thought. (On the other hand, I have seen students that are so careless and/or have such poor arithmetic skills that they make stupid errors which they should learn to overcome and/or double check. You did not want to reward sloppy work by having the final answer worth too little.) You might also want to consider changing things as the year progresses. In the beginning, when you are trying to convince them that showing their work is important, weigh the set-up higher, but as the year progresses, weigh the final answer higher. These are just ideas.

Whatever you do, let the students know what you expect and be consistent in your grading. If you do that, you can't go wrong.

[Unregistered]

You will get better at this with experience.
As you grade the tests, you will find that there are common errors made by many students. It will become easier to assign values to their work.

You really need to think about this as you write the test.
I outline the concepts being tested.
I carefully write each question to reflect the progression of concepts.( some basic questions moving to more difficult application questions)
I think about what skills are being tested and assign a point value to each question. Usually I use one point for each skill.
For example:
Student is asked to solve and graph the solution -5( x - 3) < 25
1 pt for distribuing correctly
1 pt for solving
1 pt for flipping the sign
1 pt for the correct answer
1 pt for open dot
1 pt for correct direction of the arrow

I write these on my key so I'm consistent

The hard part comes when they mess up early in the problem. Then I look to see if the rest of the work is correct as they carry their error through the problem. So if they'd done the distibutive incorrectly but carried that error through the problem correctly I'd take off 2 pt (distribution and correct answer)

It sound like a lot of work, but if you carefully write test items to reflect what you want to assess and look at the errors they make, you will figure out their weaknesses and what you need to re-teach.

I don't do this with small quizzes during the unit. I assign 4 points to each problem.
4= correct
3=mostly right
2= half right
1= I can find something right
0= I can't find anything right
I still analyze their errors for re-teaching or review.

Hope this helps. Welcome to teaching and thanks for teaching math!

[Unregistered]

Math grading is a little tricky, especially as the topics get more and more complicated. In my class, quizzes are all short answer and students are required to show their work. Usually each problem is 3 points a piece (my quizzes have to be between 20-50 points, so most quizzes are 10-15 questions).
0 pts = no work shown
1 pt = students showed their work, but it wasn't correct
2 pts = students showed their work and were close to being correct
3 pts = perfect work, perfect answer

Note, if students don't show their work, even if the answer was right, they get zero credit. Students often complain that they did their work in their calculator, but I tell them that they have to at least tell me what they did in their calculator. This cuts down drastically on cheating.

At our school, all teachers that teach the same subject are required to give the same tests and compare their grades, so we usually use multiple-choice tests to get students used to standardized tests that they will face on the ACT. It also makes scores much easier to compare because there isn't an easy grader or a tough grader. It's also much more difficult for students (and parents) to argue about their scores for a grade as big as a test grade.

[Unregistered]

Grading math is very subjective and consistency is important.

I take off more points if they made an error specifically dealing with what was taught in that unit. I take off fewer points for careless, computational errors. Also, if the answer is incorrect and they did not show any work, I deduct all the points.

I grade tests one page at a time (all page 1s, all page 2s, etc.). That way I make sure I deduct the same amount of points for the same error.

[Unregistered]

I think all you teachers should just give us kids 100's. then you'll be the best and you'll get more money when people see the sucess in our reports cards. ^.^

[Unregistered]

Hi, I am a new high school math teacher, and I am really comfortable with all aspects of my job, except grading tests. Grading seems so subjective to me, even in math, unless you use an all-or-nothing method of grading. I am really uncomfortable deciding how much partial credit to give incorrect answers, and I am constantly second-guessing myself. I really wish there was a formula for this, because I could definitely follow one. I've looked at so many math rubrics, and they still seem so subjective to me. Every time I ask a colleague, she can only give me examples of how she would grade a specific problem, but nothing that I could use for all of my exams throughout the year. It seems like something that I need to get a "feel" for, but I've been a TA for the last three years before my current job, and I still don't get it. I'm hoping that someone here might be able to give me some advice...

You set the rubric before the test begins and let the kids know exactly how it works. I'm speaking as a language teacher here, but I believe that there are many similarities in our written testing methods, because we're often expecting students to use simpler, discrete skills recursively to come to solve higher-difficulty problems. Often I did not punish compounded errors if I could see the procedure behind it was sound. In fact, if you want to teach good test taking responsibility, include the point value for each question/question type next to the question or instructions. That way the kids know what you want them to focus on.

Here's something else that I've noticed: with rare exception, the scores on any given test will break down across students in the same _relative ordering_ every time: the brightest kids do the best, the slowest kids do the worst. The up or down shift of the actual values in the distribution might have alerted me as to whether the test was too easy or too difficult, whereas any strange reordering generally means that my grading method was off. This is not to say that the kids at the bottom of the barrel "deserve" to be there if I had done everything right, but they too would benefit from predictability in my grading, because they could focus more on the material and less on the format of the exam. My goal would be to tighten the score differences between the relative ordering, challenging the the brighter kids and supporting the stragglers.

It's not an easy task. Past exams can also be a lot of help to students in order to get them used to format. Good luck.

[shubh]

While checking the assignments or the test answers of the students, you will come across new problems every day. There can't be a thumb rule for marks deduction for wrong part of the answers. This entirely depends upon your judgment and guidelines of board authority.

[Unregistered]

objective:You wll learn to make measuremnt using the metric system