Textual evidence is the backbone of literary analysis, yet it is one of the hardest skills to grade consistently. Students include quotations because they have been told to, but many of those quotations are either too long, unrelated to the claim, or left without explanation. A rubric that asks only whether the essay "includes evidence" cannot tell the difference between a thoughtful integration and a decorative one.

Equal Rites provides plenty of material for practicing this skill, from short exchanges between Granny Weatherwax and Esk to descriptive passages about Unseen University. Students have to choose which moments best support their argument, which is itself a valuable analytical decision. Teachers can grade not just whether evidence appears but whether the student chose well and handled it responsibly.

Three qualities matter most: relevance, integration, and explanation. Relevance means the evidence actually supports the claim being made. Integration means the quotation is introduced and woven into the student's own sentence, and explanation means the student says what the evidence shows rather than leaving the reader to guess.

A Scale for Evidence Quality

It helps to define levels of evidence use in concrete terms so that scoring does not depend on mood. At the lowest level, the essay makes claims with no support. At the highest, every claim is backed by a carefully chosen passage that is introduced smoothly and followed by analysis that connects it to the thesis.

• No evidence: claims about the book are made without reference to the text.
• Vague evidence: the essay refers to events in general terms with no specific detail.
• Dropped evidence: quotations appear but are not introduced or explained.
• Integrated evidence: quotations are introduced and linked to the claim with some explanation.
• Purposeful evidence: carefully chosen passages are analyzed in depth and drive the argument forward.

Common Problems and How to Respond

The most common problem is the dropped quotation, where a line from the novel appears in the middle of a paragraph with no lead-in and no commentary. A helpful comment is to ask the student what the quotation proves and to add a sentence that says so. This single revision often improves a paragraph more than any other.

Another frequent issue is evidence that is accurate but irrelevant, such as a quotation about Esk's appearance used to support a point about her determination. Gently pointing out the mismatch and suggesting a more fitting scene teaches students to match evidence to claims. It also shows them that choosing evidence is an act of interpretation rather than a mechanical step.

Paraphrase Versus Direct Quotation

Students often assume that every piece of evidence must be a direct quotation, but strong essays frequently paraphrase events and quote only the lines whose exact wording matters. In a novel as playful as Equal Rites, a line might deserve quotation because the phrasing itself is part of the joke. Teaching students to decide between paraphrase and quotation improves both concision and analysis.

Your rubric can acknowledge this by rewarding a mix of strategies rather than a fixed number of quotations. Setting a minimum count may encourage padding, while asking for "well-chosen evidence" gives students room to make smart decisions. Make sure to include a note about citation format so that students know how to reference page numbers or chapters consistently.

Grading Evidence Efficiently

One efficient method is to underline every piece of evidence in the essay and write a short code beside it: R for relevant, I for integrated, and E for explained. At a glance, you can see whether the essay is evidence-rich but analysis-poor or the reverse. This approach turns a subjective impression into a pattern you can describe to the student.

AI grading tools can perform a similar scan, flagging quotations that are not followed by explanation and drafting a comment for each. Teachers can then confirm the flags and add judgment about the quality of the choice of evidence, which still requires human reading. The result is a faster process that preserves the part of grading that depends on professional expertise.