How to grade a learner's text assessment — for short answer, long form, or any question that needs human judgment.
Most text assessment question types auto-grade (multiple choice, multi-answer, dropdown, exact-match single input). Long form questions and open-ended single input always require manual grading, and that's what this article covers.
Manual grading is a real time investment, especially at cohort scale. Before you start grading, make sure the assessment has a clear rubric or evaluation criteria so different graders score consistently — see Assessment Grader Settings for grader pool and rubric design.
How You'll Know There's Grading to Do
There are two ways graders see new submissions:
Email and Slack notifications. If your org has notifications enabled, graders get an email (and/or a Slack message) when an assessment is ready to grade. The notification includes a link straight to the submission.
Manager Dashboard. All assessments awaiting grading are visible in the Manager section. Use this as the central queue for grading work.
How to Grade a Text Assessment
1. Review the assessment. Open the submission and read through the question(s) that require grading. Pay attention to anything that hints at the learner's understanding beyond the surface answer.
2. Score the question(s). Use the sliding scale to award points. Apply the rubric (if you have one) consistently across learners.
3. Add written feedback. Click the Feedback icon next to the question to leave a comment. Specific feedback ("Strong on X, missed Y") is more useful than generic ("Good job").
4. Save the feedback. Click Save to attach the feedback to the question. The learner sees both the score and the feedback when they review their final grade.
5. Submit when complete. Once all questions are graded, click Submit. The learner is notified that their assessment has been graded.
What Makes Good Manual Grading
Apply the rubric, not your own bar. If two graders are scoring the same program with different mental standards, the assessment loses its meaning. Stick to the rubric criteria, especially on borderline submissions.
Calibrate before grading at scale. If you're one of several graders, run a quick calibration with the team — score 2–3 sample submissions, compare scores, discuss differences. Surfaces standard drift before it lands in production scores.
Write feedback the learner can use. "Needs improvement" doesn't tell the learner what to do next. "Your answer covered the why but not the how — try adding a concrete example" gives them a path forward.
Grade promptly. Submissions that sit for weeks create distrust in the program. A defined grading SLA (e.g., 5 business days) — paired with the notification — keeps the queue moving.
Configuration Pitfalls
Grading Without the Rubric. If the assessment doesn't have a defined rubric and several graders are involved, scores will diverge. Pause grading and define the rubric before continuing.
Generic Feedback. "Good work" or "Try again" doesn't help the learner improve. Tie feedback to the specific criterion or question.
Submitting Without Reviewing. The Submit button releases the grade and feedback to the learner. Double-check scores and feedback wording before clicking — once submitted, edits require coordination with admin.
Grading on Surface Patterns. Long-form responses that look thorough but miss the actual question get higher scores than they should from a quick read. Read for substance, not length.
Ignoring the Queue. Submissions that wait become a backlog and erode trust. If the queue is consistently behind, the grader pool is too small or the SLA needs adjustment — see Assessment Grader Settings.
Where This Fits
You're here because a learner's text assessment is waiting on your grade. The rubric framework lives in Assessment Grader Settings. For video grading, see Grade a Video Assessment.
See Also
- Assessment Grader Settings — grader pool design and SLA.
- Grade a Video Assessment — for grading video submissions.
- View Assessment Responses — for admin-level review across a cohort.
- Resetting Attempts on an Assessment — when a learner needs another attempt.