Instructors of courses with two hundred or more students often avoid assigning writing because they cannot imagine grading it. Multiple-choice exams feel safer, even when the instructor knows that writing would deepen learning. Recent debates about handwritten exams and AI have made the problem worse, since many of the proposed fixes, such as in-person written tests, simply do not scale to a lecture hall of that size. A deliberate system makes writing possible even in the biggest courses.

The first challenge is consistency. When ten teaching assistants each grade twenty papers, small differences in how they interpret the rubric add up to significant unfairness. Students notice quickly, and complaints often go straight to the instructor. A shared rubric, clear anchors, and a regular calibration routine are the foundations of any large-course writing program, so the instructor should expect to spend real time on rubric design before the semester begins.

The second challenge is time. If each paper takes fifteen minutes, a set of two hundred requires fifty hours of reading, divided across the team. That workload is realistic for a graduate TA only if tasks are small and expectations are clear. Designing assignments with the grading load in mind is just as important as designing them for learning, and an honest estimate at the start of the term prevents the rushed, thin comments that appear in week twelve.

Design assignments for scale

Choose shorter assignments that target one skill at a time, such as a one-page argument memo or a paragraph analyzing a source. Short tasks are quicker to grade and easier for students to complete well. They also allow multiple opportunities to practice and improve across the semester. Reserve longer papers for one or two major assignments and give them extra grading support.

• Use short, single-skill writing assignments instead of one long paper
• Run a one-hour norming session with sample papers before each assignment
• Have TAs second-score a few papers each other graded
• Track how often TAs change draft scores by criterion
• Share common errors and strong examples with the whole class

Calibrate the TA team before every assignment

Hold a short norming session before each assignment, where all TAs score the same three sample papers and discuss disagreements. The instructor should facilitate and make final calls about how the rubric applies. Keep notes on these decisions and share them with the full team. Calibration takes an hour or so and can prevent weeks of disputes later, and recording the decisions in a shared document means new TAs can catch up without repeating the whole discussion.

During grading, have TAs exchange a few papers for second scoring and review the results together. If one TA is consistently harsher or more lenient, offer support and adjust expectations. Keep the tone collaborative, since TAs are learning to grade too. A supportive team produces more consistent and fairer results, and a ten-minute check-in midway through the grading window gives everyone a chance to raise questions before scores are final.

Use tools to speed the first pass, with TA review

A rubric-based AI tool can give TAs a starting draft of criterion scores and comments for each paper, which they review and revise. This can shorten the reading time substantially while keeping a person responsible for every score. Instructors should specify the review expected, such as checking every score against the paper and rewriting any comment that sounds generic. Confirm that using such a tool is consistent with university policy and student privacy rules.

Track how often TAs change the draft scores and on which criteria. A criterion with frequent changes may need clearer descriptors, while a TA who changes nothing may need coaching on reviewing. Sharing these patterns in team meetings keeps the process honest. It also helps the instructor see where the rubric is working and where it is not, and the numbers give the instructor concrete evidence for deciding which descriptors to rewrite before the next assignment.

Give students feedback they can use

With hundreds of students, individual comments cannot be extensive, so focus on a few high-value elements. A criterion-level score report with one or two sentences of specific advice can be enough for students to improve. Share common errors and strong examples with the whole class in lecture or on the course site. Students learn from seeing patterns across the class.

Offer simple pathways for students who want more help, such as office hours that focus on one paragraph or a short revision opportunity for partial credit. Make clear how to request a regrade and what evidence will be considered. A transparent process reduces frustration and saves time. Over a term, it also teaches students to use feedback productively, and posting the policy on the course site means TAs can point to it instead of explaining it repeatedly.