When several teachers in the same department teach Stone Cold, their essay marks can drift apart without anyone noticing. One teacher rewards sophisticated vocabulary, another prizes clear structure, and a third gives credit for personal engagement with the story. By the time results are compared, the same essay might earn a full band difference depending on who marked it. Consistency has to be built deliberately rather than assumed.

The first step is a shared set of exemplar essays that the whole department has discussed and agreed on. Choose three or four responses to the same question on Link or Shelter, spread across the mark range, and have every teacher mark them independently before comparing scores. The disagreements that emerge are more useful than the agreements, because they reveal which criteria are being interpreted differently.

Exam-style tasks on Stone Cold typically ask students to explore how a character is presented or how a theme develops across the novel. Because the book is short and approachable, students often produce fluent but shallow responses that sound confident. Markers need to be alert to this pattern and consistently reward analysis of methods over fluent summary, otherwise grades will inflate unevenly.

Building Shared Band Descriptors

Published mark schemes are helpful, but they are often written in language broad enough to cover many texts. A department-level translation that spells out what each band looks like for Stone Cold specifically removes a lot of guesswork. For example, a high band response might consistently link Link's wary first-person narration to his fear of trusting strangers, while a lower band response only notes that he is homeless and young.

• Agree on what counts as an analytical comment versus a descriptive one
• Decide how much credit to give for contextual references to homelessness
• Clarify whether quotations must be embedded or may be set out separately
• Set expectations for how often students should comment on structure
• Define how to treat essays that argue unusual but well-supported readings

Running Moderation Without Losing a Weekend

Moderation meetings tend to be long because teachers talk through every disagreement in detail. A leaner approach is to sample five essays per teacher, have a colleague mark them blind, and meet only to discuss those with a gap of more than one mark. This approach surfaces genuine outliers without dragging the department through every borderline case. It also respects teachers' time, which makes people more willing to participate.

Record the outcomes in a shared document that includes the essay, the agreed mark, and a two-sentence rationale. Over a few terms this becomes a library of reference points that new teachers can use to calibrate quickly. It is far more effective than handing a newcomer a mark scheme and hoping their instincts match the department's.

Where AI Fits in Department Marking

AI grading tools can serve as an extra pair of eyes during moderation. Running the same sample through a tool configured with the department's rubric offers a neutral comparison point, and large gaps between the tool's reading and a teacher's mark are worth discussing. The tool is not the authority, but it is consistent, which is precisely what human markers find difficult over a long evening.

Departments also use AI to produce first-draft feedback for the whole cohort, then ask teachers to review and personalize it. This gives every student a baseline level of detailed comment while saving staff from rewriting the same observations repeatedly. Teachers keep the final say, and students get faster, more uniform feedback than a purely manual approach would allow.

Keeping the Standard Stable Year to Year

Marking standards slip when departments rely on memory. Each autumn, revisit the exemplar bank, retire any essays that no longer feel representative, and add a few new ones from the previous cohort. A short calibration session at the start of the unit is worth more than several hours of correction after marks have already been released.

When the standard is stable, students benefit in a very practical way: a comment from one teacher means the same thing as the same comment from another. That clarity makes feedback more trustworthy and improves how students respond to it. Consistent grading is not a bureaucratic exercise; it is part of teaching fairly.