Grading

Answer grouping

Fourteen students wrote x = 5/3. Grade it once.

On most questions, large groups of students give literally the same answer. Answer grouping clusters matching answers so you grade each cluster in a single pass instead of fourteen identical times.

How it works

For the current question, InteGrade reads the work inside each submission's question region and clusters submissions whose answers match. You review a cluster, apply rubric items once, and the grading lands on every submission in the cluster.

Screenshotanswer groups for a question, one cluster expanded

Staying in control

  • Open any cluster to see every member's actual work before applying grading. A cluster is a suggestion, not a verdict.
  • Pull any submission out of a cluster to grade it individually. Same final answer with suspect work is exactly the case to pull out.
  • Ungrouped or one-off answers are graded normally, one at a time.

Where it shines

Grouping pays off most on questions with short, well-defined answers: numeric results, simplified expressions, multiple choice, true/false. Long free-form derivations cluster less because the full work rarely matches exactly.