Many English departments use common assessments so that students in different sections of a course are held to the same standard. A shared essay on the Book of Job is a good candidate, since it requires close reading and argument. But a common prompt does not guarantee common scoring, and without careful calibration two teachers may give the same essay very different grades.

Score drift happens for understandable reasons. Teachers bring different expectations about what counts as strong analysis, different tolerance for errors, and different habits of generosity or severity. Left unexamined, these differences can mean that a student's grade depends more on which room they sat in than on the quality of their work.

Departments can address the problem through a structured calibration process. The process takes a few hours at the start of a grading cycle but saves much more time in disputes and regrades later. It also builds a shared professional understanding of what excellent writing about Job looks like, which benefits instruction as well as assessment.

Start With a Shared Rubric and Anchor Papers

The rubric should describe each performance level in observable terms and be agreed on by all teachers before the assessment. Anchor papers, which are real student essays that exemplify each score level, make the rubric concrete. For a Job essay, anchors might show a thesis that merely summarizes, a thesis that makes a claim but lacks support, and a thesis that is supported with well-analyzed verses.

• Draft a rubric with written descriptors for every performance level and criterion.
• Select anchor essays from a prior year or pilot, removing student names.
• Have each teacher score the anchors independently before the group meets.
• Discuss every discrepancy until the group agrees on how to apply the rubric.
• Record the decisions and update the rubric language to reflect the agreements.

Run Calibration Sessions That Produce Real Agreement

In a calibration session, each teacher shares the score they gave a sample essay and explains the evidence that led to it. The discussion often reveals that disagreements stem from different interpretations of a single descriptor, such as what counts as "sufficient" explanation of a quotation. Resolving these differences in conversation clarifies the standard for everyone.

Track agreement over time by comparing scores on a small set of double-scored essays. If two teachers differ by more than one level on a regular basis, revisit the rubric language or hold another session. Departments that monitor agreement can document improvements and demonstrate fairness to administrators and families.

Use Double Scoring Strategically

Scoring every essay twice is rarely feasible, but double scoring a sample can catch problems early. Have a second teacher score ten percent of essays from each section, selected at random, and compare results. Large differences point to either unclear rubric language or a grader who may need additional calibration.

Borderline essays, such as those that fall between two score levels, are also good candidates for second readings. Establishing a clear process for these cases protects students from the luck of the draw and gives teachers confidence in their decisions. Documenting the process helps when questions arise from students or parents.

Support Consistency With Technology

Rubric-based AI grading tools can apply the same criteria to every essay, offering a consistent baseline across sections. Teachers can compare the tool's suggested scores with their own to identify where they may be drifting. This comparison can be a useful professional learning tool, prompting conversations about expectations and standards.

As with any tool, human review remains central. Teachers should examine suggested scores, consider context, and make final decisions. When technology is used to support calibration rather than replace judgment, it can improve consistency while preserving the professional expertise that makes assessment meaningful.