Learn/Guide
Process-Over-Product Grading Policy Template
A process-over-product policy tells students and faculty what evidence counts when AI is available: drafts, inquiry logs, verification notes, and defense — not only the final file. Use this template to align integrity statements, rubrics, and tooling.
Published 2026-07-23
What should a process-over-product policy declare?
The shift is explicit: learning evidence includes the path to the answer. Final submissions remain important as syntheses, but they are insufficient alone when generative AI can produce them cheaply.
This matches Edudojo's tagline — grade the process, not the submission. See process over product philosophy for the underlying rationale.
What sample language can departments adapt?
Start with a short preamble your governance body can approve, then add course-level AI use rules (allowed, restricted, or forbidden by task).
"Our department assesses student learning through visible process and verified reasoning. When AI tools are used, students must document inquiry, revision, and verification. Submitting a polished artifact without demonstrable engagement in the learning process may be treated as incomplete work, regardless of surface quality."
What artifacts should students submit?
Required artifacts make the policy operable — students know what to produce, and graders know what to review.
- Process log or platform export from a guided workspace
- At least one intermediate draft with instructor or peer feedback addressed
- Verification appendix: sources checked, AI claims rejected or confirmed
- Short reflection: what the student learned vs what AI supplied
- Defense component where stakes are high — oral, live Q&A, or timed explanation
How should you weight grades?
| Component | Suggested range | Notes |
|---|---|---|
| Process and inquiry | 25–40% | Prompt quality, iteration |
| Verification and literacy | 15–25% | Red Team / source checks |
| Final synthesis | 25–35% | Still graded for quality |
| Defense / reflection | 10–20% | Mastery check |
Ranges are starting points — calibrate to discipline norms.
How do you align tools, integrity, and appeals?
State that process evidence may be reviewed in integrity investigations — replacing detector scores as primary evidence. Provide appeal paths when platform logs contradict student narrative. Privacy practices: privacy policy.
Policies work best when tools reduce faculty burden. Edudojo's Socratic Engine captures process inside the assignment. Pilot status: Parishkar College MVP in testing; classroom not yet started — details.