Modern classroom with digital display wall

Grade the Human,
Not the Homework

What is Grade The Human, Not the Homework?

What happens when we stop grading the homework, and start grading student understanding instead?

Grade the Human, Not the Homework is a practical framework for scalably assessing student understanding, rather than solely grading an assignment submission. This approach moves beyond grading assignments to evaluating what students actually know.

Goals of a Human-Centered Assessment Strategy

A human-centered assessment strategy still demands independent practice, but grades students based on their understanding of it, not the submission itself.

  • Students complete independent practice and submit assignments as usual, maintaining the value of authentic learning experiences.

  • Assessment focuses on student understanding rather than the submission itself, evaluating what students actually know.

  • AI use or non-use becomes trivial when we expect students to be accountable to their work, AI used or not.

Strategy Framework

Four assessment types for verifying student understanding at scale, with deployment options for every class size.

Interactive Check-ins

After independent practice and assignment submission, engage students in real-time conversations to assess their understanding through live discussions, spot-checks, or one-on-one sessions.

Literature

Spot Check

“Tell me about the main argument from the reading. Provide a counterargument.”

Business

Change a Variable

“What happens to your sales projection if interest rates increase?”

Humanities

Defend a Decision

“Why did you use the word 'marvelous' to describe the king's helmet?”

How Do I Scale This?

  • On Campus

    Rotate through 8-10 students during an in-class activity while the rest of the class continues working. Over multiple class meetings, every student gets assessed without stopping instruction.

  • Office Hours / Async

    Use office-hour mini check-ins or collect a short async discussion video. This keeps scheduling flexible while still requiring students to explain their own reasoning.

  • Audit Model

    Use staged sampling: review a smaller random subset each round, then rotate to a new subset next time. Students prepare consistently because any round could include them.

  • Strategic Use

    Reserve check-ins for high-stakes assignments where AI misuse risk is highest or where key course outcomes are assessed. You get stronger integrity coverage without adding check-ins to every task.

When You Have TAs
  • Assign TAs a rotating roster so each TA handles a predictable subset of check-ins each week.
  • Use a shared 2-3 question script so TA feedback is consistent across sections.
  • Escalate unclear or high-risk responses to the instructor for final review.

AI Policy Language for These Assignments

Use this Demonstration-Based policy as suggested syllabus language for Grade the Human assessments. Copy it into your syllabus and adapt the bracketed details to your course.

Suggested starting language. Personalize for your course. Align any integrity language with your institution's academic integrity policies.

Demonstration-Based

When the grade comes from demonstrated understanding, so AI use polices itself.

Pairs with Grade the Human, Not the Homework

Suggested syllabus language

You may choose to use AI on this assignment, or not. That's up to you. Your grade will be based on a [follow-up conversation / oral defense / in-class checkpoint] where you explain, defend, and adapt your submitted work. Because credit comes from demonstrating your own understanding, using AI without learning from it will leave you unable to succeed. Use whatever tools help you genuinely learn the material.

Browse Sample Assignments

Not sure where to start? Let us pick one for you, or open a subject below to explore examples in your field.