Students talk, and when two classes write the same essay on The Woman's Hour, they quickly compare scores. If one section's average is a full letter grade higher than another's, the conversation turns to fairness. The difference may reflect real performance, but it may also reflect different grading standards. Departments that care about equity need ways to detect and reduce this variation.

Inconsistency comes from several sources. Graders interpret descriptors differently, drift in strictness as they work through a stack, and are influenced by factors like handwriting or familiarity with the student. Even one teacher grading alone can score the same essay differently on different days. Recognizing that these effects are normal is the first step toward managing them.

The most powerful remedy is calibration. Before grading, all teachers score the same few essays independently, then compare results and discuss disagreements. The conversation exposes differing interpretations of the rubric and leads to shared standards. Thirty to forty-five minutes spent this way can improve consistency across an entire unit.

Anchor papers and exemplars

Anchor papers are sample essays that illustrate each score level on the rubric. Having a strong example of a four, a three, and a two removes much ambiguity. Graders can refer back to anchors during scoring to recalibrate when they sense drift. Students also benefit from seeing these examples, since they clarify expectations.

• Calibrate with shared sample essays before grading begins
• Select anchor papers that represent each level of the rubric
• Recheck anchors midway through a large grading session
• Have a second reader rescore a random sample of papers
• Compare section averages and investigate large gaps

Guarding against drift

Drift occurs when a grader's standards shift as fatigue sets in or as they encounter unusually strong or weak papers. A simple safeguard is to periodically return to an anchor paper and check that it would receive the same score. Another is to grade in a different order, so that no section is always read last. These small habits reduce systematic bias.

Mixing the order of papers also helps. If a teacher grades all of the strongest students first, subsequent papers may seem weaker by comparison. Randomizing the stack evens out the effect.

Using second readers and audits

A light audit by a second reader can reveal inconsistencies. Choose ten papers at random from each section and have another teacher score them. Differences larger than one level suggest a need for discussion. The goal is not to punish anyone but to align understanding.

Technology as a consistency check

Software that applies the same rubric to every essay can serve as a reference point. When a teacher's score differs sharply from the tool's suggestion, it prompts a second look, which may reveal an error or a legitimate reason for the difference. Because the tool does not tire or favor particular students, it can be a useful check on human variability. The teacher still makes the final decision, but the comparison improves reliability.

Record and share what you learn. Notes on common disagreements and resolutions help new teachers and improve the rubric over time.