Do your standards change as you grade?

[malvolio]

Am I alone in this? As much as I try sometimes, I find that I sometimes grade the last few papers differently than the first few.

Sometimes, after I see 1/3 of the students make the same mistake, I realize that the assignment was poorly designed, or that something wasn't clearly explained. I'll stop counting that mistake as heavily and usually go back and change earlier grades to reflect that change.

But other times I just have an uneasy feeling that I get either worn down or angry and that shows up in a 1/2 grade easier or harder for those at the bottom of the pile.

If you ever feel this way, do you go back to the earlier papers and review? I'd like to do that, but I'm so overwhelmed with an overload and several committees, I can barely get things graded the first time. Any suggestions for heading this off?

Or maybe I just feel that way and my grading is fair. After 40 freshman research papers, I'm not even sure of my own name, let alone my standards.

[historian]

I find that sometimes I have to go look up facts while grading. AFter 30 or so different wrong answers to something, I find myself confused.

As to grading fatigue and standard shifts? I used to worry about it and then I found that what I am doing is grading *faster* toward the end when I see certain "big errors" or common faults in essays, (and say) something has hit the "B" threshold,  I skim the rest seeing if there are any big markers that would lower it further. Same with the other grad ranges---once the categories for essay grades are firmly established in my mind.

[magistra]

No, I suspect this is a common problem.  I do go back and check.  I also started grading papers by the following method.  Please note, however, that this was with two papers a semester that were a relatively small percentages of the grade, not a 50 percenter or a serious research paper!

I grade on a scale of 5s.  A great paper that does everything right and shows real thought is a 100.  An A paper is a 95, an A- a 90, etc.  If I feel a grade is truly in between I will give it a 92 or 93 or whatever; no point in sticking too rigorously to any system.  I'd been giving a paper a 92 that was at least as good as that 93 from the day before, and I was going crazy; this is much better.

Students have never had a problem with this method (I suspect they think in terms of 5s too).  I point out, rightly, that within a couple points there's likely no affect on their final grade; and that they could be getting an extra point or two, not losing them.  It's subjective within a few points anyway, and they know that.  The time and sanity saved is well worth it.

[malinche]

You're not alone in this. I find that my mood affects how I grade. I've learned after so many that if I've been at it for hours, my comments get snippy and the grades go down.  I no longer grade in huge chunks of hours.  I grade 5-10 papers (depending on class size) in a sitting and then take a long break or wait until the next day.  I also write the grade in pencil b/c I end up changing some grades by the time I'm done.  As the semester progresses, I get to know who are the "really good" students and who are the "middle of the road" ones.  I read the "really good" students first and then the "middle" students' papers to check and see where the median of the class is.  Everyone else is graded in comparison to these two.  When I get worried that I've been grading too harshly, I go back to the "representative" papers of the "really good" and "middle."  It's one way I try to keep consistent.  But, let's face it - we are not machines and true consistency is the ideal.

[galactic_hedgehog]

I often will go back and check to see how I graded similar answers if they're a little weird -- if I get something unexpected and incomplete I often have to decide on the fly if it's worth 2 or 3 points out of 5.  And if I see it a few times I might decide, no it is worth 3 points after all, and go back and give the others the extra point.  The important thing is for me to be consistent.  Well, and to make sure I grade things correctly.

[scheherazade]

I sometimes grade them piecemeal.  I can't focus on a paper without noting spelling/grammar stuff, so I'll go through the stack just marking that.  I give myself a break, then go back and grade them all on another section of the rubric.  Etc.  It doesn't take any longer that grading them all the way through (I've had to do this repeatedly with about 100 6-8 pg papers) as long as each paper has an attached (photocopied) rubric on which I can write that section's grade.

I like the attached rubrics, because students tend to get overwhelmed by all the comments on the paper, and they don't read them.  So if there's a brief summary attached, they at least get something out of it.  They often go into the paper looking for specific things if they scored particularly low in a section.  Granted, rubrics are kind of k-12, but I've had professors do it before in classes I took.  They're especially nice when a grade challenge comes up.  You could always keep the rubric for yourself and not pass it back - just use it as a way to keep track of what you're looking at in the paper.

Anyway, when I grade across by rubric section, I grade more consistently.  And my brain hurts less - always a good thing.

[math_prof]

I'm not sure that this method would apply for your discipline, but I have found that grading one page at a time causes me to grade much more consistently.  Once I am past the first page, I have no idea whose paper I have in front of me (except for the rare student whose handwriting is easily identifiable) and I can concentrate solely on the content.

[cc_alan]

I'm not sure that this method would apply for your discipline, but I have found that grading one page at a time causes me to grade much more consistently.  Once I am past the first page, I have no idea whose paper I have in front of me (except for the rare student whose handwriting is easily identifiable) and I can concentrate solely on the content.

I do something similar. I try to grade one or two problems at a time so I can be as consistent as possible. If I change my mind about something, I then go through all the previous ones to verify/change the way I graded them.

Alan

[iomhaigh]

I'm not sure that this method would apply for your discipline, but I have found that grading one page at a time causes me to grade much more consistently.  Once I am past the first page, I have no idea whose paper I have in front of me (except for the rare student whose handwriting is easily identifiable) and I can concentrate solely on the content.

I do this on exams.  I grade all answers to essay 1, then all to essay 2, then all to essay 3, etc. 

For papers... I used to do what one of my grad profs suggested and grade alphabetically, starting at A for paper #1, M for #2, S for #3, C for #4, etc. so that they all got the benefit of whatever my tendencies as I progressed were. 

Now I just use check sheets and make sure I grade all of one assignment at once.  I will go back and adjust for various screw-ups on my end as I go.  This is how people end up with B+/A- grades on paper and a 91 in the gradebook. 

I'm a little too in love with the slash grade today.  Lots of slashes.  Some for using Wikipedia and not their brains, but otherwise lots of slashes. 

[carebearstare]

When I grade papers, it definitely takes me 2-3 before I feel "ready" to grade them. I often find that my standards are off the mark, so I have to adjust them; I also almost always read the paper I think will be the best in the class but don't actually grade it until reading it a second time, toward the end.

I try to do all my grading in one sitting to avoid the day-to-day mood swings that can affect how I tabulate paper scores. But of course, this isn't always possible. A friend of mine also gave me a nice rubric that allows me to think about papers that are wildly different in their weaknesses/strengths in a consistent fashion. This helps, for instance, in giving a paper that is well-written but full of crap the same grade as a more awkwardly written but on point paper.

[grad_geek]

A friend of mine also gave me a nice rubric that allows me to think about papers that are wildly different in their weaknesses/strengths in a consistent fashion. This helps, for instance, in giving a paper that is well-written but full of crap the same grade as a more awkwardly written but on point paper.

Is this rubric field-specific, or is it something you can share with us?

I am new to this grading gig, and I do worry that my standards change somewhere between paper #1 and paper #40.  I try to go back and make things consistent, but my brain is usually fried by that point...

[historian]

I've always made lists prior to grading of what should be in an A essay (and so on through the letter grades) in terms of information about the subject.  And, I make a rough list of what sort of analytical conclusions I expect an A paper to have, a B paper and so on.


 I always have one about writing issues as well that doesn't change on any assignment for that level of class.

One thing I do is to skim all the essays on a subject once and do a rough sort (A,B,C, below C). Then, I read through w/o grading and sort again, impressionistically, from best to worst.  It really doesn't take that long and gives me an internal measurement of the class range and I go from best to worst in grading when I sit down to it. 

I found that went I went randomly from great to dismal then to three mediocre then dismal then great and so on, I lost all perception of standards.

[magistra]

If you PM me I'll send you mine, though I inherited it -- it's not original to me.  They're pretty basic -- 5 points for title & heading, 10 for spelling and grammar, 10 for each part of the main topic to be addressed, whatever.  Once you stop and think about what you want you'll realize it's pretty easy to make one (if you give comments, you're basically do the same thing, it's just not as formalized.)  Below is a web address; I wouldn't set mine up like that, but it gives you a place to start.  The great thing is you can give these to your students in advance.

http://writing.umn.edu/tww/responding_grading/creating_rubrics.htm

One prof I had used Patrick Rael's criteria; his site in general is quite useful: http://academic.bowdoin.edu/WritingGuides/

[carebearstare]

I don't think this rubric is field-specific. It divides the paper into thirds--clarity, creativity, and content--and then assigns a third of the points to each category. Clarity consists of the argument, how well it's written, etc. Creativity is self-explanatory; content is how well it answers the questions you determined in the assignment. You could break the points down differently if one category matters more to you than others. For myself, though, I find it's pretty easy for me to think of those as distinct ways to excel (or fail) at writing. So I'm able to say, "OK, student A really came up with a great idea here, but it doesn't fit the assignment." Or the like.

[dept_geek]

I'm not sure that this method would apply for your discipline, but I have found that grading one page at a time causes me to grade much more consistently.  Once I am past the first page, I have no idea whose paper I have in front of me (except for the rare student whose handwriting is easily identifiable) and I can concentrate solely on the content.

I do this on exams.  I grade all answers to essay 1, then all to essay 2, then all to essay 3, etc. 

...

I do the same, only exactly the opposite. I grade the last question first. This has the added benefit of "hiding" the name until I get to the first page. (NB I tell my students to only put their name on the first page. I am thinking of changing to a cover page for name, etc next time so I don't see their name until I am done completely). I picked up this technique from an u/g professor, so it is not at all original.

And like others, I use rubrics heaviliy for the longer problems (like writing code or short essays). The yes/no understand/no clue questions are easy.  :-)