Assessment Guide¶
Instructor material
This page is for instructors and teaching assistants. Students should not need it.
Assessment components¶
| Component | Source | Nature |
|---|---|---|
| Module quizzes | Self-Assessment | Recall, application, judgement |
| Lab artefacts | Labs | Practical skill, self-marked by rubric |
| Case discussions | Case Studies | Analysis and reasoning |
| Capstone | Module 13 | Integration across the whole course |
Suggested weighting¶
For a course that awards credit:
| Component | Weight | Rationale |
|---|---|---|
| Module quizzes | 20% | Regular, low-stakes checks that keep pacing honest |
| Lab artefacts | 35% | The practical skill the course exists to build |
| Case studies | 15% | Reasoning and communication |
| Capstone | 30% | Integration across all modules |
Adjust to your own programme's conventions. The distribution above deliberately weights doing over recalling.
Marking the quizzes¶
Several questions, especially in Part C of each quiz, have more than one defensible answer. The model answers say so explicitly. Mark the reasoning:
- A well-argued answer that disagrees with the model answer should score full marks.
- A correct conclusion with no reasoning should score partial marks only.
- Where a question asks for a judgement, expect the student to present both sides before taking a position.
Marking lab artefacts¶
Every lab carries a self-marking rubric. When you mark, ask the same question the rubric does:
Could a competent stranger act on this document without asking the author anything?
Common deductions:
| Observation | Interpretation |
|---|---|
| Success criteria like "users are happy" | No observable definition; revisit Module 01 |
| Empty "out of scope" section | The boundary was never decided |
| Constraints and assumptions conflated | Foundational confusion; revisit Module 02 |
| A single-point estimate with no range | Uncertainty not acknowledged; revisit Module 07 |
| Risks with no owner or trigger | A register written to be filed, not used; revisit Module 08 |
| A status report that is green throughout | Either a genuinely green project, or optimism; probe it |
Marking the capstone¶
Use the Capstone Rubric, which defines eight weighted dimensions with four levels each, and states explicitly what is not assessed — ambition, hours spent, formatting beyond professional adequacy, and any claim the artefacts do not support. The rubric is the same instrument students use for self-marking, deliberately: a rubric that differs between marker and student cannot be used to predict a result, and predictability is what makes it a learning tool.
Grade coherence, not polish. A strong capstone tells one consistent, honest story across its charter, plan, reports, and retrospective. Signals of a strong submission:
- The charter's scope and the plan's WBS agree.
- Risks in the register actually appear, or are consciously absent, in the reports.
- The status reports disclose problems when they occurred, not after resolution.
- The retrospective names specific, owned changes supported by the earlier artefacts.
- Where reality diverged from plan, the divergence is explained rather than hidden.
A weaker submission may be well formatted but internally inconsistent — for example, a charter that excludes a feature that the WBS then includes.
Academic integrity¶
Because answers are public, quizzes are unsuitable as high-stakes graded examinations. Use them for low-stakes checks and pacing, and assess the labs and capstone (which produce individual artefacts) for the grade. If you need a high-stakes examination, write a new one whose answers are not published.
Feedback that helps¶
- Point to the specific artefact section rather than giving a general score.
- Ask a question rather than issuing a verdict: "How would a stakeholder use this?" moves students further than "this is vague".
- Reward early disclosure of bad news in reports and discussions; you are trying to build a habit that the grade should not punish.
See the Course Map for how assessments align with module outcomes.