The Real Problem
Many teachers spend hours reading short‑answer responses and deciding how many points each answer deserves. Because the work varies in wording and detail, judgments often differ between classes or even between days, leading to student confusion and a feeling that marking is arbitrary. The effort required can also push grading into personal time, reducing work‑life balance.
Inconsistent partial credit also makes it hard to communicate expectations. When a student receives a mark that seems lower than a peer’s for a similar answer, they may question the fairness of the system. Teachers, meanwhile, must remember the rubric they created, which can be difficult when handling dozens of papers in a single session.
How Quesly Helps
Quesly lets teachers upload lesson notes, a readable TXT file, or a text‑based PDF, then generates short‑answer questions automatically. After the class submits their work, the system locks the responses and Jev, the AI assistant, proposes a mark and a brief rationale for each answer based on the teacher‑provided marking criteria.
These suggestions appear alongside the original student text, giving teachers a concrete reference point. Teachers can compare the AI’s reasoning with their own view, edit the mark if needed, and then finalize the grade. The process still requires teacher approval, but it reduces the time spent weighing each answer from scratch.
Illustrative Classroom Example (Synthetic)
Ms. Rivera teaches a unit on ecosystems. She uploads her lesson outline, which includes key concepts and a rubric that awards up to three points for depth, one point for correct terminology, and one point for clarity. Quesly creates five short‑answer prompts, and her students submit handwritten responses that are scanned into the platform.
When the submissions are locked, Jev analyses each answer and suggests a total of four points for a response that mentions “photosynthesis” and explains its role, but omits the term “chlorophyll.” The suggestion includes a note: “Term missing – deduct one point.” Ms. Rivera reviews the note, confirms the deduction, and records the final mark.
What Still Needs Teacher Judgment
The AI can only work from the criteria the teacher supplies; it cannot interpret nuance that falls outside those rules. If a student provides a creative but correct explanation using alternative terminology, the teacher must decide whether to award full credit despite the AI’s deduction for a missing keyword.
Additionally, the teacher remains responsible for ensuring that the suggested marks align with the broader learning goals of the course. The system does not replace professional pedagogical decisions, such as awarding extra credit for exceptional insight or adjusting marks for known classroom circumstances.
Practical Takeaway
Use Quesly to generate a first draft of marks for short‑answer work, then spend the saved minutes refining those suggestions rather than starting from zero. Keep the rubric visible, compare AI rationales with your own expectations, and adjust only when the reasoning reveals a legitimate discrepancy.
By treating the AI’s output as a collaborative partner rather than a final verdict, teachers can achieve more consistent partial credit, free up personal time, and give students clearer feedback about why points were awarded or deducted.
