Rethinking Grading in the Age of AI: Four Essential Questions for Educators

With the arrival of generative AI, instructors face a pivotal moment as traditional grading assumptions are challenged. Rather than questioning authorship after a submission, it's essential to flip the perspective: focus on what can be done to ensure trustworthiness in student work before it is assigned. Here are four crucial questions that educators should consider in the lead-up to grading, aimed at enhancing the assessment process.

1. Can I See the Thinking, or Only the Text?

The reliance on a final draft as the sole submission places too much burden on that document. In a world where writing can easily be polished by AI, asking for more than just the finished product is becoming increasingly crucial. Requesting a range of outputs, such as proposals, rough drafts with revisions, or reflection paragraphs detailing the student's thought process, creates opportunities for deeper engagement and learning. This approach fosters an environment where students can genuinely interact with their work, showcasing their evolution as writers, thinkers, and creators.

Such documentation not only sheds light on a student's progression but also helps establish authorship clarity. It allows instructors to see the steps a student has taken, making it easier to identify who wrote what and how they arrived at their conclusions. This isn't just about assessing the final piece; it’s about evaluating the entire learning journey. And this is the part most people overlook: the journey matters as much, if not more, than the destination when it comes to education.

2. Does My Rubric Still Measure What is Scarce?

Most existing grading rubrics were developed in an era when creating polished, coherent prose posed a challenge for students. Today, thanks to AI, even fluency can be easily replicated. As a result, rubrics that primarily reward organization and mechanics risk being outmoded. A simple check-box for grammar correctness or structure won’t push students to think critically. These outdated measures fail to capture the essence of what makes student work unique.

Instead, educators should recalibrate their rubrics to assess more nuanced aspects such as the strength of arguments, clarity of thought, and the quality of thoughtful revisions in response to feedback. If a rubric can be satisfied by machine-generated text, it’s time to rethink its criteria entirely. The challenge lies not just in crafting a new scoring system but in aligning it with the critical skills students need in a world where AI is an omnipresent tool. This is more significant than it looks — a well-thought-out rubric can encourage the very competencies that human learners should excel in.

3. What Does My AI Policy Ask Students to Actually Do?

Merely requiring students to tick a box about their AI use misses an opportunity for deeper engagement. Rather than offering a cursory acknowledgment of AI assistance, instructors should prompt students to document their decision-making process regarding AI use. What prompts did they input? What responses did they choose to keep or disregard, and why? Such transparency elevates the conversation from mere compliance to critical judgment, encouraging students to engage with AI outputs thoughtfully.

If you're working in this space, consider how this can lead to a richer classroom dialogue. Instead of students feeling like they’re navigating a minefield of rules and regulations surrounding AI, they start taking ownership of their choices. This also encourages a culture of honesty and analytical thinking among students, assets that are incredibly valuable in any academic or professional setting.

4. Who Decides the Hard Cases?

Whenever grading is involved, contentious cases are inevitable. Addressing these requires collaboration rather than isolated decision-making. Sharing samples with a co-instructor or colleagues provides a platform for varied insights and interpretations. Engaging in discussions about those borderline submissions can shed light on differing expectations, ensuring consistency in grading that is fair and transparent.

It’s not about scoring a piece through an automated detector or late-night second-guessing; it’s about collective engagement and discourse over contested submissions. Within half an hour, instructors can review borderline evaluations together, compare their assessments, and clarify the criteria that led to divergent views. This collegial approach not only fine-tunes the judgement process but also improves the overall fairness and accuracy of grades.

Significance and Future Outlook

What’s critical here is that these four questions don’t necessitate additional software or institutional revisions. They hinge on creating assignments that yield substantial evidence, adjusting criteria to value what remains unique to human potential, and maintaining a personal touch in grading. Even with large classes, a thoughtful, individualized response to each student can have a lasting impact far beyond a mere numerical value given by an algorithm. After all, it’s the human interaction in assessments that students will remember most.

As syllabi are finalized in the coming weeks, now is the time to address these questions before the first set of essays lands on your desk. Educational practices need to evolve just as rapidly as the technology reshaping them. The job of educators isn’t just to assess; it’s to cultivate thoughtfulness and integrity in their students, qualities that will last a lifetime.

Roger Ochse, EdD, is Professor and Honors Director Emeritus at Black Hills State University. With over forty years of experience in teaching college writing, including nineteen years online, he is currently developing the Green Pen Framework, an approach aimed at producing trustworthy assessments through thoughtfully designed assignments.