Picture two students who wrote nearly identical essays on A Tree Grows in Brooklyn. One is in Ms. Alvarez's class and gets an 88, and the other is in Mr. Chen's class and gets a 79. Neither student did anything different, but their grades did.

This kind of gap is common when several teachers share a text and an assignment. Each reader has their own sense of what an A looks like, and small differences add up across a grade level. Students notice, and so do parents.

Calibration is the fix. It is a short, structured process where teachers score the same essays, compare, and agree on what each level means.

It does not have to be a big project. An hour or two at the start of the unit can save weeks of arguments later.

Start with a shared rubric and sample essays

Agree on one rubric before students write. Then collect five or six anonymous sample essays from a prior year or from a trial run, ideally ones that span the range from weak to strong. Each teacher scores them on their own before anyone talks.

• Choose one rubric and make sure everyone has the same version
• Pick five or six anonymous sample essays across a range of quality
• Have each teacher score the samples independently and write down their scores
• Compare scores and discuss the essays where they differ the most
• Write short notes on what each score level looks like and keep them as anchors

Focus the conversation on differences

Agreement is easy and teaches little. The useful talk happens around the essay where one teacher gave a 95 and another a 78. Ask each to point to the lines that drove their score, and the disagreement usually turns out to be about one or two specific things.

Write those decisions into the rubric. If the group agrees that a strong thesis must include a reason, add it to the descriptor. Over time the rubric becomes a record of shared judgment.

Check again mid-unit

Drift returns fast, especially under deadline pressure. Set up a second check after everyone has graded about a third of their essays. Swap a few papers, score them blind, and see whether the group is still close.

This is also a good moment to look at the distribution of scores by class. A big difference in averages is not always a problem, but it is worth a conversation.

Use a consistent first pass

One way to reduce drift is to give every teacher the same starting point. GraideMind applies the department's rubric to each essay in the same way, so scores and comments begin from a shared baseline before each teacher makes their own adjustments.

That does not replace calibration. It supports it, since the department can review where teachers changed the first-pass score and use those cases to refine the rubric further.