Assessment and appeals

This page describes the earlier fourteen-week layout and is being rewritten; where it differs, the syllabus applies.

Read the current syllabus for weights, passing conditions and equivalent resits. Each work record contains the artifact, verification, your own explanation and one AI-use note. Reuse evidence between a practice unit and its project milestone.

The two AI-off checks and individual defense establish independent ability. Project quality cannot substitute for the independent passing requirement. A future or missing assessment is not a zero and cannot establish a completed course result.

Receive feedback before the next related acceptance. If you disagree with a judgment, identify the criterion and evidence you want reconsidered and use the teaching team's published appeal route. Suspected inconsistency requires human examination and an opportunity to respond. No writing-style detector decides authorship or guilt.

The adopting teaching team supplies actual submission dates, reassessment appointments, approved accommodations and appeal contacts before teaching. Ask the team for these arrangements if they are not yet provided.

CP1: individual diagnosis, test and repair

CP1 is a 30-minute, 10-point, individual, closed-book, AI-off paper in Week 5. It asks you to locate a reported defect in an unfamiliar ci-dashboard path, write a contract-based regression test, describe a minimal repair and protect other callers. The marking categories may be public so that the task is accessible and predictable.

The actual code excerpt, planted change, distinguishing input, line numbers, trigger observations and answer remain private teacher material. Public practice teaches the defect category and the verification method; it does not reuse a CP1 instance. The question shape may be public; the instance is not.

If the initial CP1 result is below 50%, one equivalent resit is available after feedback. It keeps the same learning outcomes, time and difficulty. The replacement is capped at 5/10, and the effective mark is the higher of the original and capped replacement. An approved invalid sitting is replaced without using the resit opportunity; a valid pass cannot be retaken for a higher mark. Both records remain. CP1 results do not change model access.

CP2: individual unfamiliar-code review

CP2 is the Week 11 individual checkpoint. It lasts 30 minutes and is completed on paper, without AI, agents, devices or reference material. You review an unfamiliar code listing, identify evidence-backed defects, and give a bounded recovery and verification argument. The paper and answer sheet are collected at the end; project work cannot substitute for this individual evidence.

After feedback, an initial CP2 mark below 5/10 permits one equivalent resit with the same construct, timing and difficulty. The replacement mark is capped at 5/10 and the original attempt remains recorded; the effective mark is the higher of the original mark and the capped replacement mark. A technical failure or an approved accommodation failure is replaced by an equivalent sitting and does not consume the resit. Ask the teaching team for the published appointment and accessibility arrangements.

Request a review by naming the rubric point and the evidence on your script. A person reviews the evidence and gives you an opportunity to respond; no writing-style or authorship detector decides the mark. The teaching team publishes the review contact and response timetable before teaching.

Oral defense

The individual oral defense is worth 20 points. Book a 15-minute slot in Week 13 or Week 14: 12 minutes are for your answers and 3 minutes are for the examiner to record the result. Bring your own work and work records. Complete the defense without AI.

The defense connects four domains: requirements and code paths; verification and evaluation; security and incident recovery; and a changed-input responsibility decision. LO3, LO6, LO7 and LO8 are required. It transfers your evidence across these domains rather than re-testing all nine learning outcomes one by one, and the project demonstration cannot replace it.

The explanation check asks what you changed, why you changed it, which inputs are most likely to fail, and what you verified yourself. Keep claims tied to the evidence in your work. If your first mark is below 50% of the defense (below 10/20), feedback is followed by one equivalent resit with the same outcomes, time and difficulty. The resit mark is capped at 50% of the defense (10/20), the effective mark is the higher of the original and capped resit marks, and both records remain.

A second calibrated examiner reviews a case near the threshold or a case with a scoring disagreement, or when you ask for a review. Recording audio or video is not a default requirement; any recording follows the published notice, permission and retention rules.