Published essays by professional writers make excellent mentor texts for studying craft. But for teaching students what their own writing should look like, student writing samples are often more effective. A student who reads a published essay may admire the writing without believing they can produce something similar. A student who reads a strong essay written by a peer in last year's class can see the goal as achievable because the writer was in the same position they are now: a student in this class, with this teacher, on a similar assignment. Student samples bridge the gap between 'what great writing looks like' and 'what great writing by someone like me looks like.' That second perception is what makes the model useful for instruction rather than just inspirational.

The most effective use of student samples is not just showing strong examples. It is showing a range: a strong sample, a developing sample, and a sample with specific, identifiable issues. When students see all three, they can locate their own writing on the continuum and understand what the next level looks like. A class that scores all three samples using the rubric before writing their own essays develops a calibrated understanding of quality that no amount of rubric reading alone can provide. They have seen what a 4 looks like, what a 3 looks like, and what a 2 looks like in actual student work. The rubric descriptors become concrete rather than abstract.

Ethical use of student samples requires clear protocols. All samples should be anonymized: remove the student's name and any identifying details. Samples from previous years are generally safer than samples from the current year because there is no risk of the author being identified by classmates. If you use a sample from a current student, get their explicit written permission and ensure they understand how the sample will be used. Never use a weak sample in a way that could embarrass or identify its author. The purpose of using student writing is to teach, not to critique individual students publicly.

Building a sample library is a long-term investment that pays off for years. At the end of each semester, save three to five strong student essays (with permission) that represent different performance levels on your rubric. Over two or three years, you will have a collection of fifteen to twenty-five samples that cover every rubric dimension at every score level. This library becomes your most versatile teaching tool: it calibrates student understanding, trains peer reviewers, supports rubric introduction, and provides models for every writing skill you teach.

How to Use Student Samples in Instruction

These five uses cover the most common instructional contexts where student samples are more effective than published texts.

• Rubric calibration: before the first graded essay, have students score two or three anonymized samples using the class rubric. Compare student scores to teacher scores. Discuss discrepancies. This builds assessment literacy.
• Before/after revision: show students a first draft and a revised version of the same essay. Ask them to identify what changed and why each change improved the essay. This teaches revision as a visible, concrete process.
• Targeted skill instruction: use a sample that demonstrates one specific skill well (strong thesis, effective evidence integration, vivid description) as the model for a mini-lesson on that skill. The specificity makes the lesson concrete.
• Common error correction: show an anonymized sample that contains a common error (comma splices, weak conclusions, evidence without analysis). Have students diagnose the problem and propose a fix. This builds editorial judgment.
• Anchor papers for scoring: establish a set of anchor papers at each performance level that the class agrees on. Students reference these anchors during self-assessment and peer review to calibrate their own evaluation.

Student Samples and AI Grading Calibration

Student writing samples serve a calibration function for AI grading tools as well as for students. Running your anchor papers through the AI tool before the first grading cycle shows you whether the AI's scoring aligns with the team's agreed-upon standards. If the AI scores a consensus-3 paper as a 4, the rubric language may need tightening for that dimension. If the AI and the team agree on all anchor papers, the rubric is clear enough for both human and machine scoring. This calibration step takes five minutes and significantly improves the accuracy of AI-assisted grading for the rest of the semester.

The sample library is one of the most valuable assets a writing teacher builds over a career. It costs nothing to create. It requires only the habit of saving strong student work (with permission) at the end of each semester. And it improves instruction, assessment, peer review, and grading calibration in ways that no textbook or published resource can replicate. If you do not yet have a sample library, start building one this semester. Future-you will be grateful.