Homework 3 — Rubric
ECBS5293 — Computing for Analytical Work
Deliverable: Homework 3 — Debugging, data files, recovery, and final diagnosis note Format: hw3-submission.zip (made with git archive from your commits; includes REFLECTION.md) and the video, both uploaded to Moodle. Due Fri 9 Oct 2026, 23:59; one day late at −10%, nothing after Sat 10 Oct 23:59. Total points: 100
Overview
You receive a repo with several realistic failures. Some are code errors; some are problems in the input data files. You repair the project, verify the output, explain each failure, and — because this is the last deliverable — write a short reflection looking back across all three homeworks. The final exam comes first, on Wed 7 Oct (10:50–12:30, room D-001); the mock exam, in the final’s format with a key, is on Moodle by Wed 30 Sep. A self-check key for this homework is posted on Moodle when the late window closes (Sun 11 Oct).
How scoring works
Every criterion is scored at exactly one of its three anchor values — no in-between points. The video is scored per element.
Rubric
1. Correct fix (25 points)
| Level | Points | Criteria |
|---|---|---|
| Excellent | 25 | Every failure fixed; the pipeline runs end to end from a fresh unzip and produces the verified output; data problems fixed with loading options, not by editing the data. |
| Satisfactory | 16 | Most failures fixed; one remains, or one was fixed by hand-editing the data. |
| Needs Improvement | 6 | The pipeline does not run, or fixes mask the problems rather than resolve them. |
What we’re looking for: fixes that would survive the next data file.
2. Verification (15 points)
| Level | Points | Criteria |
|---|---|---|
| Excellent | 15 | Row counts, column names, dtypes, and missing counts checked after each load (the four-line check, left in the code); the output file exists and is what the task asked for. |
| Satisfactory | 10 | Some of those checks. |
| Needs Improvement | 4 | “It ran” with no data validation. |
What we’re looking for: the four-line sanity check, applied.
3. Diagnosis notes (15 points)
| Level | Points | Criteria |
|---|---|---|
| Excellent | 15 | One note per failure, each with all five parts, each labelled code or data with the evidence that decided it. |
| Satisfactory | 10 | Notes present but a label is wrong or evidence is thin. |
| Needs Improvement | 4 | Failures lumped together, or causes guessed. |
What we’re looking for: you can tell a wrong column name from a wrong delimiter, and you can say how you knew.
4. Video walkthrough (15 points)
Scored per element: full, half (vague or wrong), or 0 (absent). Up to 2 minutes for this homework.
| Element | Full | Half | What earns full |
|---|---|---|---|
| Symptom | 3 | 1 | Shows the pipeline running end to end and names what used to fail |
| Cause | 5 | 2 | Explains the failures, distinguishing data problems from code problems |
| Change | 3 | 1 | What you changed — loading options vs. code — and why that split |
| Verification | 4 | 2 | Shows the checks that prove the data loaded correctly |
Your voice is required: a silent recording scores 0 on all four elements — the point is you explaining. Open by naming the homework and the repo; a missed name is a clean-submission matter (criterion 6), not a video one. Excellent / Satisfactory / Needs Improvement map to 13–15 / 8–12 / 0–7.
Approved alternative formats (see the syllabus’s accessibility section — arrange in advance): a written walkthrough with annotated screenshots, or an audio-only recording, is scored on the same four elements at the same values. The “your voice” requirement applies to the standard video route only, never to an arranged accommodation.
5. End-of-course reflection (15 points)
| Level | Points | Criteria |
|---|---|---|
| Excellent | 15 | All four specifics present: a quoted sentence from a prior diagnosis note; a step in the loop you admit to skipping and what it cost; the failure class you were slowest at and a concrete plan; one thing you stopped and one thing you started doing with AI. |
| Satisfactory | 10 | Three of the four specifics; the rest is general. |
| Needs Improvement | 4 | Generalities, or a reflection that could have been written without re-reading your own work. |
What we’re looking for: honesty and specificity. 200 concrete words beat 300 padded ones; ~300 is a ceiling.
6. Clean submission and Git hygiene (10 points)
| Level | Points | Criteria |
|---|---|---|
| Excellent | 10 | GIT_LOG.txt shows a sequence of meaningful commits ending in a clean final state; nothing junk in the archive; run instructions correct. |
| Satisfactory | 6 | Minor clutter or a stale instruction, or commit messages that say nothing, or the video does not open by naming the homework and repo. |
| Needs Improvement | 2 | No history, junk in the archive, or instructions that do not work. |
What we’re looking for: before archiving, you knew exactly what changed.
7. AI use disclosure (5 points)
| Level | Points | Criteria |
|---|---|---|
| Excellent | 5 | Specific: what you asked, what it got right or wrong, how you verified. Or an honest “did not use AI.” |
| Satisfactory | 3 | Present but generic. |
| Needs Improvement | 1 | Missing, or contradicted by the submission. |
What we’re looking for: the same honesty the reflection asks for.
General Notes
- Diagnosis notes, video, and reflection together (45 points) outweigh the fix (25). This is deliberate: the course grades the diagnosis, not the tinkering.
- AI tools are allowed. You must be able to explain everything you submit, in your own words, without notes.
- Fix data-loading problems with loading options (delimiter, header, dtype, encoding, dates), not by editing the data file.
- Re-read your own three diagnosis notes before writing the reflection — that re-reading is the point, and the best exam preparation you have.
- Nothing is accepted after Sat 10 Oct 23:59: the self-check key is public from Sun 11 Oct.