Getting Started: Grading Assignments

Once students have started submitting work, the Submissions view is where you review it, apply auto-graded scores, and score anything left for manual review. This walkthrough covers both ways to get there — directly from the Lab.Computer dashboard, or via a Canvas-linked assignment — and what you actually see once you open a submission.

Reaching the Submissions view

Standalone, from the Lab.Computer dashboard: open the course and go to its Assignments tab. The Submissions column shows how many students have submitted — click that count to go straight to the list.

If the assignment is Canvas-linked instead, start from Canvas: open the assignment and click Load [Assignment] in a new window — the same button students use to launch it. Because you're signed in as the instructor, that same launch lands you on the assignment's grading page in Lab.Computer rather than the student notebook.

Either path leads to the same place: click View Submissions to see every student who has submitted, with columns for Student Name, Email, Submission Time, Score, Status, and an action to open the submission.

Opening a submission doesn't launch anything for the instructor — there's no live, editable notebook to run. What opens is a read-only HTML rendering of exactly what the student submitted, cell by cell, with each graded cell showing the points it was awarded out of the maximum.

Auto-graded vs. manually graded cells

Whether a cell needs your attention depends on how it was configured when the assignment was built. Recall from the previous article that any cell marked with Points is graded in one of two modes:

• Auto — the cell's output is checked automatically against expected output the moment a student submits. These scores populate immediately and typically don't need instructor review.
• Manual — the default mode for graded cells. These are left at their configured point value until an instructor opens the submission and assigns a score.

In practice, most assignments mix both: auto-graded cells for objective checks (does the function return the right value), and manually graded cells for anything that requires judgment (code quality, written explanations, open-ended analysis).

Manual grading and overrides

Each graded cell in the rendering carries its own Point field next to the max, along with Save, No Credit, and Full Credit buttons, and a comment box for feedback to the student. For a Manual cell, enter a score and click Save. Auto-graded scores can be overridden the same way — for example, if a functionally correct answer used an approach the auto-checker didn't anticipate.

The submission's total score, shown at the top of the page, is the sum of all cell scores and updates as you grade — you don't need to grade every cell in one sitting.

Grading Canvas-launched submissions

For assignments launched from Canvas, the review and grading workflow is identical to what's described above — the same rendering, the same Auto/Manual mechanics, the same per-cell controls. The one difference: the resulting score is also passed back to the linked Canvas gradebook column via LTI, so students see their grade in Canvas without the instructor re-entering it there. See Canvas Assignment Setup and Launch for how that gradebook link is configured.

A note on academic integrity

Because each submission preserves the student's actual notebook — not just a final score — instructors can review the work itself, not only the output, when something looks off. Proctoring, if enabled for the assignment, adds an additional layer of session-level oversight on top of this.

Putting it together

Across this series: an account gets you in, a course gives you a place to teach, an assignment gives students something to do, and grading closes the loop. Whether your path runs standalone through Lab.Computer or through your school's Canvas instance, the underlying environment students work in — and the way you build and grade assignments — is exactly the same.