A poetry survey built around a single anthology is often taught by several instructors in the same term. Each may choose different poems, write different prompts, and apply slightly different standards when grading. Students notice these differences quickly, and complaints about uneven grading tend to surface around midterms and final papers.

Poetry makes this problem harder than it is in many other subjects. There is no single correct answer, so graders must judge the quality of an argument and its evidence. Reasonable instructors can disagree on whether a particular reading is persuasive, and those disagreements show up in scores.

Departments do not need to eliminate this variation entirely, but they should limit differences that stem from unclear criteria rather than genuine judgment. Shared rubrics, calibration practices, and consistent feedback tools reduce the portion of inconsistency that is avoidable. The goal is fairness for students, not uniformity of opinion.

Start With a Shared Rubric

The foundation of consistency is a rubric all instructors agree to use for common assignments. It should define the major criteria, such as thesis, evidence, analysis of form, and writing quality, and describe what each performance level looks like. Instructors can still add their own comments, but the core scoring structure stays the same.

• Agree on the major criteria and their relative weights
• Write descriptors that describe observable features of the writing
• Collect sample essays at each performance level for reference
• Hold a short calibration session before each major assignment
• Review score distributions across sections after grading

Calibration Without a Heavy Time Commitment

Traditional calibration involves several instructors grading the same essays and discussing differences, which is valuable but time-consuming. A lighter version has each instructor score three sample essays independently, then compare results in a brief meeting. Even this modest practice reveals where criteria are being read differently.

AI grading tools can supplement calibration by applying the shared rubric to every essay in every section. Instructors can then compare the tool's scores with their own to find cases where interpretation diverges. This provides an ongoing check on consistency rather than a one-time event at the start of the term.

Preserving Instructor Autonomy

Faculty are understandably wary of any system that seems to dictate how they grade. The best approach positions shared tools as a baseline, with instructors free to override scores and add commentary. This respects professional judgment while still giving students a more predictable experience.

Departments should also be transparent about how the tool is used and how its output is reviewed. When faculty understand that they remain the final decision makers, adoption tends to go more smoothly. Clear communication early in the process prevents resistance later on.

Measuring Whether Consistency Improves

After implementing shared practices, departments should look at the data. Comparing average scores and score ranges across sections shows whether the gaps have narrowed. Student complaints and grade appeals also provide useful signals about whether the changes are working.

If large gaps remain, the cause is often unclear rubric language rather than instructor disagreement. Revising the descriptors and repeating the calibration exercise typically resolves the issue. Over a few terms, the department builds a durable grading framework that supports both fairness and good teaching.