Ask five English teachers to grade the same Pygmalion essay, and you will likely get a spread of at least a full letter grade. The difference is rarely about skill. It reflects different priorities, such as one teacher valuing organization and another valuing insight. A norming session brings those differences into the open.

The tool for the job is a set of anchor essays. These are real student papers, with names removed, that represent clear examples of each performance level. Teachers use them as reference points when they grade the rest.

Choose six to eight essays to start. Include at least one strong paper, one weak paper, and a couple that sit near the boundaries. The boundary cases generate the most useful conversation.

Plan for about ninety minutes. That is enough time to score, compare, discuss, and adjust the rubric without wearing everyone out. Provide coffee if you can.

Running the session

Have every teacher score the first essay independently and write down a level for each rubric row. Then share the scores at once, so nobody adjusts to fit the group. The differences are the point of the exercise.

• Ten minutes to review the rubric and the prompt
• Independent scoring of the first anchor essay
• Group comparison of scores and discussion of disagreements
• Repeat with two or three more essays, including boundary cases
• Ten minutes to record decisions and rubric edits

Handling disagreements productively

When scores differ, ask each teacher to point to the words in the essay that drove the decision. Often the disagreement traces to an ambiguous rubric phrase. Fix the phrase, and the disagreement usually disappears.

Sometimes the disagreement is philosophical, such as whether a creative reading of the ending deserves credit when the evidence is thin. Note those, decide as a group, and write the decision into the rubric. That prevents relitigating it later.

Adding an AI reference point

Some departments run the anchor essays through an AI feedback tool as an extra data point. With a rubric-based tool such as GraideMind, the results show how the written criteria apply to each anchor. If the tool's read differs from the group's, it prompts a useful look at the wording.

The tool does not settle the question. Teachers still decide what the standard should be. But it makes the rubric easier to test.

Following up after the workshop

Circulate a one-page summary of decisions within a day or two. Include the revised rubric and the anchor essays with agreed scores. Teachers will refer back to it during grading.

Check in midway through the grading period. A quick conversation about borderline cases keeps drift from creeping back in. It takes ten minutes and saves hours of regrading.