A rubric designed around a specific novel is far more useful than a generic literary analysis template. The Machine-Gunners offers particular features to assess, including Westall's use of a close third-person view on Chas, the contrast between the children's games and wartime reality, and the dialect that gives Garmouth its voice. A tailored rubric tells students what skilled reading of this book looks like before they start writing.

Start by deciding how many criteria you can realistically score. Four or five rows is usually the limit before a rubric becomes a checklist that nobody reads closely. Each row should describe a skill you actually taught, so students are assessed on what they had a chance to practice.

Then write descriptors in plain language, avoiding phrases like "demonstrates sophisticated understanding" that mean different things to different readers. Describe what the writing does instead, such as "explains how a specific word choice shapes the reader's view of a character." Descriptors that name observable behaviors are easier to apply fairly.

Core Criteria to Include

Most teachers find that five criteria cover the work: thesis and argument, textual evidence, analysis of the author's craft, historical and thematic context, and control of written expression. Weighting these differently can reflect your priorities, for instance giving analysis more points than conventions in a unit focused on close reading. Be transparent about the weights so students know where to put their effort.

• Thesis and argument: a debatable claim about the novel sustained across the essay.
• Evidence: carefully chosen quotations or references integrated into sentences.
• Craft analysis: explanation of how Westall's language, structure, or perspective creates meaning.
• Context: purposeful use of the Blitz and home front setting to illuminate the text.
• Expression: clear paragraphs, varied sentences, and accurate conventions.

Writing Descriptors That Separate Levels

The most difficult task in rubric writing is distinguishing neighboring levels. A useful trick is to describe the middle level first, using the typical student essay as a model, and then ask what one step up would add and what one step down would lose. This keeps the language honest and anchored in how real essays behave.

Include a few concrete markers at each level when possible. At a high level, evidence might be described as connected to the claim and explained in at least two sentences, while at a lower level it might be described as present but left unexplained. These markers help colleagues score similarly and give students language to use during self-assessment.

Testing the Rubric Before You Use It

Before releasing a rubric to students, apply it to three sample essays at different quality levels. If you hesitate between two scores for a clear reason, the descriptors probably need revising. Sharing the draft with a colleague and comparing scores on the same essay can reveal ambiguity that you cannot see on your own.

Once the rubric is stable, it can power both teacher scoring and AI-assisted feedback, since a well-written rubric is exactly the kind of input that produces specific, usable comments. Revisit it after each unit and adjust language based on where students misunderstood expectations. A rubric is a working document that improves with each cohort.