What follows below is a brief description of some of the ideas on grading papers for PHYS/ASTR 109 that should be kept in mind while going through the process. Hopefully when you are finished reading this you will have a basic understanding of what the structure and quality of a “good” paper is and be able to grade effectively and fairly for this course. The ideas are flexible and should anything change in the course you should feel free to edit this page. Please contact Prof David Toback or Joshua Winchell if you are unaware how to edit this page.
Peerceptiv:
(A note about signing in as an instructor: The original instructor can promote TAs from students to instructors in Peerceptiv).
As PHYS/ASTR 109 is a course that doesn’t rely too much on math (practically, not at all), the major way in which we grade a student’s progress is through papers/essays that they submit to Peerceptiv and turnitin.com; these are four per semester (five for the honors students). Because of the sheer size of a typical BBBHNM classroom (anywhere from 60 students up to 200), it is practically impossible for the professor and the TA to grade ALL of the papers. This is why 109 employs Peerceptiv; in this system, each student gets to grade three other students (anonymously) and also send feedback on reviews of their own work.
This is useful in two ways:
1) it saves the professor and TA from having to grade an enormous amount of papers, and
2) it is also a good skill to foster in the students, that of effectively grading and peer-reviewing.
The system is not perfect, and because of this there will always be complaints of unfair grading (expect 5-6 re-gradings per paper assignment); some will be justified, some won’t. Justified or not, the TA has to go through all of them and give the final verdict.
For each paper a general outline to how the grades are distributed is explained in the rubric. This should allow you to quickly identify if the student has covered the major points of the paper in a clear and concise way. The rubric can be found in Peerceptiv. The idea is that a student that understands and can successfully communicate the ideas in a concise, layman language will receive the full score. Asking yourself “Would my mother (or some lay-person) understand what was just said” is a good check if the student has gotten it.
With that in mind, here are some tips on how to effectively go through the re-grading process:
- The subject matter and the nature of the assignments makes grading somewhat subjective, but you must try to be uniform over multiple gradings.
- Remember, you are grading two things: the science and the writing. Sometimes a paper will have (by virtue of how they’re constructed) only a few low scores, yet the quality of the writing and/or science is sub-par, in which case it is ok to bring the score down by an extra point. For example, a student that gets all the science right, as well as the content and formatting, but fails at sentence structuring and grammar may only get a few low scores, but if the paper as a whole is suffering more than just the sum of these parts, an extra score penalty could be in order.
- It is always useful to bring up two screens: one with the student’s paper, and the other with the rubric parts (if available). That way, you can take your time reading the paper and then systematically go through the questions.
- The paper is supposed to be understood by an average lay-person. This usually means not to use any scientific words or words that only a select group of people can understand. If the student is going to use scientific words, the student should also provide a quick and clear explanation of what they are.
What does the grade mean:
You could think of the scores on a paper roughly as follows:
~100% —This is perfect…the student gets it and communicates it clearly and concisely!!
~90% —This is still an A grade…the student seems to get it and communicates this fairly well and concisely
~80% —This is a B grade…the student demonstrates that they get the subject material but their communication is not great or is too long
~70% —Now this is a C grade…the student doesn’t show that they understand completely and their explanation is unclear or confusing
~60% —This is getting into the low grades and is a D…the student obviously doesn’t get it or has poor communication skills
~50% or less— This is an F or worse and the student either needs to have not completed at all some point on the rubric or has done so poorly that the paper is worse for including this section.
General Ideas:
I’ve found that when the students bring their paper to the TA is very useful to take the time and mark each one extensively (this is after speaking with Prof Toback about styles and form). This allows the student to know that you ARE in fact reading each paper and DO want them to do well in the course. I also always offered time to read drafts before they turned in their papers and would sit with each student until they could say in words the structure of the paper, then really emphasize how important getting the language clear and concise when they are talking to me is so that it comes across in their paper.
Be careful with the peer review results:
Since Peerceptiv does not really allow absolute scoring on our part, it’s possible for students to artificially inflate grades by just scoring every paper highly. This should be considered as acdemic dishonesty, and is unfair to other students. Ideally it’s best to check the results systematically, however it’s really not an easy job. Practically if some student complains about some unfair grading, it’s better to check the grader’s other gradings to see if there’s something strange, and start from there.
Even if you see something strange, there isn’t really a great way to prove the student didn’t do his/her work in general (maybe he/she’s just a bad grader and too nice to be critical or something) but if you make it clear from the start that anything less than full credit needs comments to explain why it wasn’t perfect (what you took away points for) and/or helpful feedback (I think we did this by suggesting they balance the good comments with the bad – “You have excellent style in this paper, but the evidence was a little lacking.”) A lot of people won’t mind being factual as well – “You totally forgot to mention that black holes have infinite density! It’s okay, I missed it too, just make sure you include it in your final draft” but requiring some ACTUAL feedback will help prevent that. It’s still a bit of a pain to check but you can spot-check around from time to time, especially when grades are drastically different.