Build a grading calibration workbook from de-identified submissions
Overview
OpenAI's starter prompt for faculty who want to check how consistently a course is being graded. Given de-identified submissions, the rubric, grade exports, sample feedback and course outcomes, it builds a workbook that points out scores that may not match, sums up what students commonly do well or miss, and suggests rubric clarifications with evidence for each flag. It never assigns or changes grades, leaving every scoring call to faculty. Remove direct student identifiers before sharing submissions or grade exports.
Use cases
Prompt
Copied 0
Review the de-identified submissions, rubric, grading export, sample feedback, course outcomes, and calibration rules I provide. Create a calibration workbook that: - flags possible scoring inconsistencies - summarizes common strengths and gaps - identifies feedback themes - suggests rubric clarifications - cites evidence for every flag Do not assign or change grades. Route every scoring judgment to faculty review.
Can AI?
- Build a grading calibration workbook from de-identified submissionsPaste into a general LLM and build a grading calibration workbook from de-identified submissions.