World literature surveys often include Lu Xun as the anchor for modern Chinese fiction, and enrollments can reach well over a hundred students in a single course. Each student submits a paper on a short text, and the grading burden falls on a professor, a handful of teaching assistants, or both. Without careful planning, the quality of feedback can vary widely from one section to another.

Consistency matters because students compare their grades and notice when similar papers receive different marks. A shared rubric and a calibration session help but do not solve the problem entirely. Differences in how individual graders interpret terms like analysis or insight can still produce uneven results.

Large courses also tend to have students with very different backgrounds, from literature majors to those fulfilling a distribution requirement. A fair assignment must be accessible to newcomers while still allowing strong students to show depth. Offering a choice of prompts at different levels of difficulty is one way to manage this range.

Hold a calibration session before grading

Select five to eight sample papers that span the quality range and have all graders score them independently before meeting. Compare scores, discuss the largest disagreements, and revise the rubric wording where confusion arises. This hour of work prevents many hours of regrading and student complaints later.

• Distribute anonymized sample papers in advance
• Have each grader score and comment independently
• Discuss papers where scores differ by more than one grade level
• Record decisions about borderline cases in a shared document
• Revisit the rubric language after the session to remove ambiguity

Use a tiered feedback approach

Not every paper needs the same depth of commentary, and trying to give every student a full page of feedback is rarely sustainable. A tiered approach assigns one global comment to every paper and reserves detailed marginal notes for students who request them or who are near a grade boundary. This keeps the workload realistic while ensuring that everyone receives something actionable.

Comment banks written by the teaching team can speed up common responses while still being tailored by the grader. For example, a comment about weak thesis statements can include a model revision that fits the Lu Xun text under discussion. Students benefit from seeing a concrete alternative rather than an abstract instruction.

Prevent plagiarism through prompt design

Lu Xun is heavily written about online, so generic prompts invite copied analysis from study guides and summaries. Design prompts that require students to engage with a specific passage, make a claim about a particular detail, or connect the story to a course discussion. These constraints make it harder to borrow ready-made essays.

Combine this with a low-stakes in-class writing task early in the term so you have a sample of each student's natural voice. If a later submission differs dramatically, you have a basis for a conversation. This approach is more constructive than relying solely on detection software.

Collect data to improve the next term

Keep a simple tally of the most common errors and strengths across the course, such as unsupported claims about history or particularly strong readings of symbolism. This information tells you where to adjust lectures and which readings need more support. It also helps new teaching assistants understand what to expect.

At the end of the term, compare grade distributions across sections to spot inconsistencies. Large gaps may point to differences in grader interpretation that you can address before the next offering. Treating grading as a process that can be improved over time leads to fairer outcomes for students.