It is common for two teachers using the same rubric to give the same essay noticeably different scores. One may weigh the strength of the thesis more heavily, while the other focuses on evidence. For students, this inconsistency can feel arbitrary and unfair. Calibration sessions reduce those differences by getting graders to talk through how they interpret the criteria.

A Small Gods essay set makes a good calibration exercise because the novel invites varied interpretations. Papers might argue that Vorbis is a sincere believer, a cynical manipulator, or something in between, and each reading can be well supported. Teachers must therefore evaluate reasoning rather than agreement with a particular view. That is exactly the skill calibration helps develop.

Start by choosing three to five sample essays that span a range of quality. Remove student names and give each teacher the rubric. Ask everyone to score each paper independently before any discussion. This ensures that initial judgments are not influenced by others.

Running the Calibration Session

After independent scoring, record each teacher's scores on a shared sheet and look for differences. Focus the discussion on the criteria where the largest gaps appear. Ask each person to point to specific text in the paper that justified their score. This keeps the conversation grounded in evidence rather than impressions.

• Score each sample paper independently before discussing anything
• Record all scores in one place so differences are visible
• Discuss the criteria with the widest disagreement first
• Revise unclear rubric language based on what the discussion reveals
• Rescore a fresh paper to confirm that agreement has improved

Common Sources of Disagreement

Teachers often disagree about how to treat a paper that is well written but weak in analysis. Some reward the polish, while others penalize the lack of depth. Clarifying how much weight each criterion carries resolves much of this. Making those decisions explicit prevents later arguments.

Another source of disagreement is the treatment of unconventional interpretations. A student who offers an original reading of Brutha's role may confuse a grader who expects the standard view. Agree in advance that well-supported originality should be rewarded. This protects creative students from being penalized for taking risks.

Maintaining Consistency Over Time

Calibration is not a one-time event. Graders drift as they become tired or as they develop their own shortcuts. Build in checkpoints during a large grading effort, such as rescoring a previously graded paper to see whether results match. If they do not, pause and recalibrate.

Rubric-driven AI tools can serve as a steady reference point. Because they apply the same descriptors to every paper without fatigue, they can reveal when a human grader's scores begin to drift. Comparing the two helps identify issues early. The tool does not decide the grade but supports consistency.

Documenting Decisions

Keep notes on the decisions the group makes during calibration. Record how you decided to handle ambiguous cases and what examples illustrated each level. Future teachers can use this record to understand the standard without repeating every discussion. It also provides evidence of fairness if a grade is challenged.

Share the resulting clarified rubric with students so they benefit from the same shared understanding. When students see that teachers have worked to ensure fairness, they tend to trust the grading more. That trust is one of the most valuable outcomes of calibration. It improves the classroom climate as well as the data.