Bulk Grading Assist
Grade a whole class at once — upload a batch of submissions, let the machine evaluate them, and export the results.
Bulk Grading Assist lets you grade a whole class's work at once. You upload a batch of student submissions, the Feedback Machine evaluates each one against your criteria, and you review and export the scores and feedback.
Grade a batch
Start a bulk import
From your Feedback Machine's menu, select Bulk Import. Upload a .zip containing your students' .docx and/or .pdf files — each file becomes its own submission (only .docx and .pdf files are processed). A single import accepts up to 1 GiB of files (with individual files up to 30 MB). Then select Start Bulk Import, and keep the browser tab open until all submissions are created.
Let it evaluate
Feedback Machines creates a submission for each file and evaluates it against your criteria. Evaluations run in the background — up to 200 at a time — so large batches process steadily without you waiting on each one; when the system isn't busy, a full class of around 1,000 submissions can finish in as little as an hour. You'll see live progress for each submission. Any that run into trouble are retried automatically, and a row that needs something from you says so: Replace files on the row if its file couldn't be read, or Retry at the top of the import for anything else.
Review results
The import lists every submission with its score and status. Open any one to read its full feedback and criteria breakdown — including anything hidden from students. To read the batch by rubric part rather than one submission at a time, and to approve the grades, select Review Results to open the machine's Results page on this import.
Export
Open the export page with the import's Export Results button, with Export on the Results page, or from the machine's menu under Bulk Export — all lead to the same page. Coming from an import or from Results, it opens on the set you were just looking at; otherwise it opens on your most recent import whenever that import is newer than any student submission. The export scope dropdown at the top sets which submissions are included: every submission, students only, each student's latest or highest-scoring submission, or one specific import listed under By Import. The page shows how many submissions the current scope will export.
If any evaluated submission in the scope hasn't been approved yet, the page says how many. Nothing is held back — the export includes every submission in the scope — so approve first, or delete the rows you don't want from the file.
When the scope is a Canvas import, the page is built around getting the results into Canvas:
Scores for Canvas Gradebook — in Canvas, export your gradebook (Actions, then Export) and upload that CSV here. Canvas requires the SIS columns from your export, which Feedback Machines doesn't store. Choose the Canvas assignment the scores belong to; Feedback Machines writes the scores into that assignment's column so Canvas imports the file without asking you to map it (Actions, then Import). If the assignment's points possible differ from the machine's, the page says so with both numbers and offers to scale the scores to match. After the export it reports how many students received a score and names any in your gradebook who didn't. Choosing New assignment instead adds a column Canvas will ask you to confirm on import, with the machine's points possible filled in.
Feedback for Canvas SpeedGrader — a zip with one PDF per submission, named after the student's original file, to attach as a comment on their submission in SpeedGrader. Each PDF carries what the student sees in their final results: the score, who approved it and when (if approved), the key takeaways as you edited them, and the rubric breakdown with each part's summary.
Below those, under Other exports, are two files for your own records that Canvas can't import:
Export Evaluations (CSV) — one row per submission: the score, each rubric part's score and summary, the key takeaways and the overall feedback summaries, with separate First Name and Last Name columns alongside each submitter's email and the Canvas student ID. It is also the record of the review: whether each submission was approved, by whom and when; how long it was reviewed and by which reviewers; when it was evaluated and whether that evaluation still matches the current criteria; and who rewrote the key takeaways or a part summary. Review time counts while a submission is open on the Results page, pausing after twenty idle minutes. For student scopes it adds each student's number of attempts, total edits and score change from their first submission.
Export Submissions (ZIP) — the original student files.
For any other scope, only these two exports are offered.
Add or replace a submission in an import
An import doesn't have to be final. On the import's page you can bring in a student the batch missed, and give a failed row another set of files.
Add Submission takes up to the machine's Files Per Submission
.docxor.pdffiles (30 MB each), evaluated together as one submission, plus the student's name. The new submission joins the import and is evaluated like the rest.Replace files appears on any row that ended in error. It opens the same form, already filled in for that student, and re-queues their row with the files you choose. The replaced submission is kept as a record and marked as replaced; the row and every export follow the new one.
On a Canvas import, filenames carry the student's identity. Files named the way Canvas exports them (student_canvasid_submissionid_filename) match the student automatically and fill in their name, which is what lets their score land in the Canvas gradebook export. Files named any other way are accepted after a prompt explaining that the submission won't be matched to a Canvas student — so for a student whose files failed, reach for Replace files on their row, which keeps the Canvas details the row already has. Files naming two different students are refused.
If the student you're adding already has a row in the import, Feedback Machines tells you what state that row is in and asks you to confirm before the new files take its place. An evaluation already on that row — and any review of it — is set aside rather than carried over.
Nothing is deleted when a row is replaced or retried. Each row shows once, with its current evaluation, on the Results page and in every export.
Refine and re-evaluate
Bulk grading is iterative. As you review the results, you'll often spot evaluations you'd have graded differently — that's expected, and bringing the machine into agreement with your judgment is the core of the workflow. A few rounds is normal.
The place to do it is the machine's Results page. There you can read the whole batch by rubric part, pull the submissions you disagree with straight into the Adjust chat, preview the corrected evaluations, and apply the change across every submission at once — then approve. See Review, Adjust & Approve Results.
Last updated