A generic literary analysis rubric often leaves teachers rewriting the same clarifications in the margins of every essay. All the Light We Cannot See invites specific kinds of analysis, from its braided timelines to its recurring images of sound and light. A rubric that names those opportunities helps students aim higher and helps teachers grade against the skills the unit was actually designed to teach.

Start by deciding what the assignment is really assessing. Some teachers want to see close reading of a single passage, while others want a broad argument about theme across the whole novel. Writing that purpose in one sentence at the top of the rubric prevents the common problem where students produce plot summaries because the task never told them what kind of thinking to show.

Next, choose four or five criteria rather than ten. Long rubrics slow down grading and blur the priorities, since students rarely absorb more than a handful of categories anyway. A tight rubric also works better with AI grading tools, because each criterion can be evaluated clearly and reported back to the student in language they can act on.

Criteria That Fit This Novel

Thesis and argument should sit at the top, and the descriptors should be concrete. A strong thesis might claim that Doerr uses radio and touch to show that understanding can travel across barriers that vision cannot, while a weak one simply says the book is about World War II. Describing that difference in the rubric gives students a visible model of what an arguable claim looks like.

• Thesis and argument: a specific, debatable claim about the novel rather than a summary statement
• Textual evidence: accurate details from the story, integrated smoothly into the student's sentences
• Analysis: explanation of how the evidence supports the claim, including the effect of Doerr's craft choices
• Organization: clear paragraphs that build toward a conclusion instead of repeating the same point
• Conventions: control of grammar, punctuation, and formal tone appropriate for the grade level

Writing Performance Levels That Mean Something

Performance descriptors are where many rubrics fail. Phrases like "good analysis" and "excellent analysis" differ by one adjective and give students no idea how to improve. Instead, describe observable behaviors, such as "explains why the short chapters shape the reader's sense of time" at the top level compared with "mentions the chapter length without explaining its effect" at a middle level.

Keep the language parallel across levels so that students can see what moves an essay from one band to the next. If the proficient descriptor mentions two pieces of evidence, the advanced descriptor might mention evidence from different parts of the novel that complicate each other. That kind of progression turns the rubric into a roadmap for revision instead of a set of labels.

Testing the Rubric Before You Use It

Before releasing the rubric, grade three sample essays with it, ideally one strong, one average, and one weak. If you find yourself wanting to award points that the rubric does not mention, the rubric is missing something important. If two criteria keep overlapping, combine them, because students and AI tools will both struggle to separate them cleanly.

Share the rubric with students early and, if time allows, have them score an anonymous sample essay themselves. Students who have practiced applying the criteria tend to write with those criteria in mind, and the conversations about why a score was earned can be more valuable than the score itself. It also reduces the number of grade disputes at the end of the unit.

Using the Rubric With AI Grading

A well-built rubric is the single most important input for AI-assisted grading. When the criteria and descriptors are explicit, an AI tool can apply them consistently to every essay and generate comments tied directly to each category. Vague rubrics, by contrast, produce vague feedback, no matter how capable the software is.

After the first batch, review the comments against the rubric and note where the feedback felt off. Often the fix is a small wording change in a descriptor, such as clarifying what counts as analysis rather than explanation. Over a few units, the rubric becomes a refined instrument that saves time and produces feedback students can genuinely use.